Introduction: The Time Course of Phonological Activation in Visual Word Recognition: An ERP Investigation
What follows is the abstract and introduction to my thesis, submitted March 10, 2008.
The present study used event-related potentials (ERPs) to investigate the role and time course of phonology in visual word recognition during silent reading. Participants (N = 15) performed a lexical decision on target words (e.g., animal) that were preceded by one of three masked primes: exemplars of the target word (e.g., bear), words that were homophonic to the exemplar, referred to as homophone foils (e.g., bare); or words that were orthographically similar to the target word, referred to as spelling controls (e.g., beer). Three ERP components were of particular interest: the N250, the classic N400, and the P600. No differences between homophones and controls were found, although varying effects emerged in each of the time windows examined. The results suggest that the selection of a word’s meaning does not depend on its phonological form, but rather that phonology plays a role after its selection.
The Time Course of Phonological Activation in Visual Word Recognition: An ERP Investigation
Language is an ability that adapted in humans over time and remains unique to humans to this day. The ability to read is crucial to success in modern society; from the relatively simple task of following directions or instructions, to gaining complex knowledge from books, journals and the internet, reading is a tool we use every day. Few can achieve an education without this ability; an education that, in turn, can lead to a well-paying job and numerous opportunities one could not otherwise afford. It seems, then, that the development of reading progresses in our lifetime from learning to read, to reading to learn.
Learning to read involves associating the visual form of a word (i.e., its orthography, symbolized as O) to its meaning (i.e., semantics, symbolized as S). According to the dual-route theory of reading (Coltheart, Rastle, Perry, Langdon, & Ziegler, 2001), this process can happen along one of two pathways: the direct route, in which a reader achieves meaning directly from the orthography (O –> S), and the indirect or phonological route, in which phonology (e.g., the way a word sounds, symbolized as P) acts as a mediator between orthography and meaning (O –> P –> S). The triangle model, based on connectionist principles, proposes a single mechanism for visual word recognition in which a distributed pattern of activation across phonology, orthography, and semantic units determines word meaning (Harm & Seidenberg, 2004; Seidenberg & McClelland, 1989). Computational models show that the triangle model is more accurate when the direct and phonological routes act cooperatively, with the division of labour between each route depending on factors such as word frequency and homophony (Newman, Haigh, & Jared, in press).
The goal of visual word recognition research has traditionally been to determine the extent to which a reader relies on orthography and phonology in order to activate a word’s meaning. Understanding how meaning is activated during reading has strong implications in early reading education and therapy for reading disabilities. For instance, differing views as to whether we rely more on the direct versus the phonological route when reading, led to in-depth debates about whether children should be taught to read via a whole-word approach or a phonetic approach (Castle, Riach, & Nicholson, 1994; Gathercole & Baddeley, 1989; Jeynes & Littell, 2000; Manning & Kamii, 2000; Vellutino, 1991). A general consensus has emerged that the phonological route is pivotal to beginning readers (see Adams, 1990). Ehri (1992) suggests that one logical function of the phonological route is to strengthen direct connections between print and meaning, which would consequently diminish one’s reliance on phonology as reading skill improves. Several studies have offered support for this view by suggesting that a direct approach to reading is the most common finding in studies with adults. This view is often referred to as a weak phonological theory because it argues that the extent of phonology’s influence on visual word recognition is dependent on various mediating factors, such as reading skill, homophony, and word frequency (Coltheart et al., 1991; Jared, Levy, & Rayner, 1999). Having said that, there are still proponents of so-called strong phonology theories of reading, which contend that phonology is activated during reading regardless of skill level and psycholinguistic factors, such as word frequency and homophony (Frost, 1998; Lukatela & Turvey, 1994). The current study will use electrophysiological measures during a visual word recognition task in order to distinguish these competing theories. Before describing the goals of the current study, however, it is important to summarize previous work examining the role of phonology in reading.
Questions surrounding the role of phonology in reading have been traditionally studied through the use of homophones and/or pseudohomophones. Homophones are words with identical pronunciations, but which differ in spelling, derivation and meaning (e.g., bear/bare). Pseudohomophones are nonword stimuli that share their phonological representation, but not their orthographic representation with real words (e.g., brane). The rationale for using homophonous items is as follows: if phonology mediates word recognition, then presentation of a homophone (e.g., bare) or a pseudohomophone (e.g., brane) will result in the semantic activation of the unseen homophone mate (e.g., bear) or the pseudohomophone’s base word (e.g., brain), respectively. If it can be shown that the meaning of the unseen homophone mate (or the pseudohomophone’s base word) has been activated, then we can infer that phonology has mediated the computation of meaning.
Evidence for homophone effects has been observed in studies employing the homophone error paradigm. In this paradigm, the member of a homophone pair that is appropriate to the experimental context (e.g., bear) is replaced by its partner, the homophone error word (or homophone foil; e.g., bare). If participants fail to notice the homophone foil, then the inference is made that the phonological representation of the homophone foil resulted in activation of the unseen correct homophone. Due to the high degree of orthographic overlap between many homophones, it is essential for a spelling control word to be present in experiments employing the homophone error paradigm. A well-developed spelling control word should be as orthographically similar to the appropriate homophone as the homophone error word, and should also approximate the homophone error in word frequency. For example, for the homophone pair, bear-bare, an appropriate spelling control word would be the word beer. Thus, if participants fail to notice the homophone error, then the effect can be attributed to the shared phonological representations between the correct homophone and the homophone foil, and not to their orthographic similarity.
Van Orden (1987) manipulated orthographic similarity and homophony (homophone errors or ‘foils’ and spelling controls, which represented the key trials) in a series of semantic categorization experiments in which participants were asked to determine whether a presented word was part of a specified category. He found that participants made significantly more false-positive errors to the phonemically-similar homophone foils (e.g., rows) than to their spelling controls (e.g., robs); similarly, there were more errors to the nonword foils (e.g., roze) than to their spelling controls (e.g., rone). This finding was referred to as a phonological interference effect because the sound of the foil interfered with participants’ decision-making processes, leading them to falsely judge a homophone foil as being an exemplar of the prompted category (e.g., rose in the category Flowers). Furthermore, Van Orden, Pennington, & Stone (1990) determined that the size of the homophone effect was dependent on the frequency of the unseen exemplar, such that the effect was larger if the exemplar was lower in frequency than its presented homophone mate. More importantly, however, Van Orden et al. still observed a homophone effect for high frequency words. Their findings led to the development of the spell check theory. According to this theory, when more than one meaning is activated by presentation of a homophone, we must retrieve the appropriate orthographic representation of the presented word (e.g., its spelling) in order to resolve the competition. When the appropriate homophone is a low frequency word, however, its spelling is more difficult to retrieve, leading to an increase in false-positive errors. The homophone effect can therefore be defined as an increase in processing or reaction time, and/or in false-positive errors, associated with the presentation of a homophone error.
Van Orden’s (1987) results were called into question by Jared and Seidenberg (1991), who claimed that his categories were overly specific, allowing participants to predict category exemplars even before they were presented. Thus, phonological activation could occur due to contextual priming, and not to the generation of a phonological representation directly from the printed word form. By proposing the use of general categories (e.g., living thing and object) in their semantic categorization study, Jared and Seidenberg (1991) examined whether the observed homophone effects in Van Orden (1987) and Van Orden et al. (1990) were exaggerated. They hypothesized that broad categories would reduce the probability that participants could generate semantic candidates proper to stimulus presentation. This reduction of phonological effects was precisely what the authors found; although homophone effects were still observed for low frequency homophone foils with low frequency exemplars, they failed to observe homophone effects for high frequency words when broad semantic categories were employed. On the basis of these results, Jared & Seidenberg (1991) argued that phonological activation of meaning occurs only for low frequency words, and that the direct route is favoured for high frequency words.
The impact of phonology on reading for meaning has also been studied in tasks that closely resemble reading, such as proofreading and sentence verification tasks. Jared, Levy and Rayner (1999) took into account predictability and frequency within their proofreading task, which they administered to both high- and low-skilled readers. Their study involved monitoring eye movements while participants read passages or sentences in which an appropriate homophone was replaced by a homophone foil or a spelling control. The only reliable evidence for phonological activation of meaning was found for low frequency homophone foils replacing low frequency correct exemplars; these findings were consistent with those of Jared & Seidenberg (1991). Furthermore, Jared et al. (1999) found that reading skill modulated phonological influences on reading for meaning. Good readers produced an orthographic pattern of eye movements; they had longer gaze durations for homophone foils compared to correct homophones and similar gaze durations for homophone foils and spelling controls. In contrast, the gaze durations for poor readers followed a phonological pattern; they had shorter gaze durations for homophone foils compared to spelling controls, and their gaze durations for homophone foils and correct homophones were similar.
A lexical decision task (LDT) involving homophonous items is often used to study the role of phonology in reading. In this task, the time it takes for a person to decide the lexical nature of items presented (whether or not it is a real word) is measured. It has been argued that this shallower level of processing, compared to reading for meaning, does not require the reader to understand the word in its context, allowing for an early phase in the word-recognition process to be revealed (Pexman, Lupker, & Jared, 2001). If phonology actively contributes to word recognition, then one would predict that low-frequency homophones would delay ‘word’ responses, assuming a spelling check has been employed to select between alternative meanings. Similarly, pseudohomophones (e.g., brane) would be expected to delay ‘nonword’ responses, since the spell check must operate to discriminate between the read word and the nonword foil. Pexman et al. conducted a LDT involving both homophones varying in frequency, and pseudohomophones varying in orthographic familiarity. In their study, they matched the word frequency of homophones to spelling control words (e.g., pain/pear), and matched the bodies of pseudohomophones and pseudowords (nonwords, e.g., RADE, CADE) for orthographic similarity. In addition, they varied the orthographic familiarity of the nonword (pseudohomophone and pseudoword) stimuli. In the wordlike context, the nonwords had orthographic representations that were similar to real English words (e.g., rade, cade). However, in the non-wordlike context, the nonwords had orthographic representations that were atypical of English words (e.g., golph, tolph). The authors expected that the atypicality of the nonwords in the non-wordlike context would allow participants to make their lexical decision based on a superficial analysis of orthographic factors, and thereby, reduce any influence of phonology on word recognition. The results revealed two main influential factors modulating homophone effects. First, adding support to past research, homophone frequency influenced the speed and accuracy of lexical decisions. More specifically, homophone effects were largest for low frequency homophones with higher frequency mates. Second, homophone and pseudohomophone effects were larger when nonwords had a familiar orthographic representations. When pseudohomophones and pseudowords had atypical orthographic representations, no evidence for homophone or pseudohomophone effects were observed. In summary, Pexman et al. concluded that readers do activate phonology during visual word recognition, but that the extent of phonological activation is influenced by word frequency and orthographic familiarity.
While the homophone error paradigm has provided support for a weak phonological theory of word recognition, findings from studies using fast priming procedures (e.g., prime is presented too quickly to permit overt identification) have been cited as evidence in support of the strong phonological theory of word recognition (Lukatela & Turvey, 1994). The fast priming procedure involves presenting a prime for a duration that is too quick to permit its overt identification, followed immediately by presentation of a related or unrelated target. In this task, participants are required to respond only to the target; however, their processing of the target is assumed to be covertly affected by a related prime. So-called phonological priming effects occur when reaction times to targets (e.g., clip) are faster for phonologically related primes (e.g., klip) compared to orthographically related primes (e.g., clep). Evidence for phonological priming has been found with brief prime exposures (e.g., 14 ms), leading to the conclusion that phonology is generated automatically and provides the primary code by which meaning is accessed (Lukatela & Turvey, 1994). However, based on findings in which phonological priming lagged behind orthographic priming, the former occurring only with longer prime exposures, some theorists argue that orthography provides the primary constraint on word recognition (Ferrand & Grainger, 1994). Since the current study employs a fast priming procedure, a brief review of the literature in this area is in order.
Lukatela and Turvey (1994) conducted a seminal study in which participants were required to name out loud a target word (e.g., frog) that was preceded by an associate (e.g., TOAD), a word (e.g., TOWED) or nonword (e.g., TODE) homophonic with the associate, or an orthographic control (e.g., TOLD) of the associate. At brief (e.g., 50 ms) stimulus onset asynchronies (SOAs), the associate prime and the homophonic primes (e.g., TOWED and TODE) produced equal associative priming; however, no priming was observed for the orthographic controls. At longer SOAs, homophonic words (e.g., TOWED) led to associative priming whereas homophonic nonwords did not. In addition, at longer SOAs, associative priming in orthographic controls was observed for the words (e.g.., TOLD) but not for the nonwords. Based on these results, Lukatela and Turvey (1994) concluded that phonology provides the initial code by which a word’s lexical representation is accessed. They acknowledged the responsibility of orthographic structure in reducing lexical competition following activation by the word’s phonological code. In other words, the way a word sounds is important in activating its corresponding lexical representation, whereas the way a word looks is important in selecting the word’s activated representation from other, simultaneously activated representations (e.g., choosing between TOWED and TOAD).
Lukatela and Turvey’s (1994) choice of a naming task has been criticized on grounds that naming invariably activates phonological output codes, which could have produced exaggerated phonological priming effects (Holyk & Pexman, 2004). Ferrand and Grainger (1994) conducted a similar study, but within the context of a lexical decision task. These authors manipulated prime duration (the length of time for which a prime word is presented to the participant) from 14-57 ms, to examine the possibility that orthographic and phonological processes follow different time courses, with orthographic information being accessed earlier than phonological information. They employed a masked form priming procedure, which involves the presentation of a fully visible word either before (forward masking) and/or after (backward masking) the prime word. The purpose of masking is to overwrite the sensory representation of the prime and block elaborate processing. The results of their study showed clear evidence of orthographic priming, but no evidence of phonological priming with a 29 ms prime duration exposures. Phonological priming did develop, but only with longer prime exposures (> 67 ms). These results conflict with those of Lukatela and Turvey (1994) and suggest that the difference in task, naming versus LDT, may play a role in the timing of phonological activation.
Ziegler, Ferrand, Jacobs, Rey, and Grainger (2000) employed an incremental priming technique, which adds a parametric manipulation of prime duration to the traditional design of a fast masked priming study. In the two previous studies mentioned, prime duration was manipulated, but not within the same list of stimuli. The advantage of the parametric manipulation is that additional information on the time course and nature of priming effects can be obtained. Zeigler et al’s (2000) findings replicated those of Ferrand and Grainger (1994); priming in the orthographic condition occurred with brief prime durations (29 ms), while phonological priming did not occur until prime durations were increased. Interestingly, as prime duration increased, orthographic priming did not continue to increase, whereas phonological priming continued to increase steadily with increasing prime duration. The researchers concluded that orthographic information is activated at the earliest stages of word recognition, while phonological priming emerges at a later stage.
Conflicting results suggest the possibility that the experimental methods employed in past studies lacked the necessary sensitivity to answer such a question. Evidence for rapid activation of phonological codes has been obtained repeatedly with the masked priming paradigm at brief prime durations (Lukatela & Turvey, 1994; Rastle & Brysbaert, 2006), but it is possible that longer prime durations cannot be properly studied without the more sensitive measures provided by event-related brain potentials (ERPs). The exquisite temporal resolution and continuous index of neural activity provided by ERP measures can be used to clarify the sequencing and timing of phonological effects, overcoming the limitations of the ‘end-point’ explanatory approach necessarily utilized in behavioral studies. The use of ERP measures to examine the impact of phonology in reading is a relatively new approach.
Historically, ERP research on reading has concentrated on semantic processing, focusing on one ERP response in particular, the N400. This component is defined as a negative-going brain wave at 400 ms post-stimulus presentation, and the response was first demonstrated by manipulating the terminal words of visually presented, contextually constrained sentences (Kutas & Hillyard, 1980). It was found that the N400 was larger to words that violated semantic expectations (e.g., He spread the warm bread with socks) than to semantically appropriate words (e.g., He spread the warm bread with butter). These early results have been supported by numerous studies (Bentin, 1987; Holcomb & Neville, 1990; Bentin, Kutas, & Hillyard, 1993; Connolly, Phillips & Forbes, 1995), with the general consensus being that the N400 reflects an index of semantic congruity.
A few ERP studies have investigated the impact of phonology in the activation of word meanings using the logic of the homophone error paradigm. The rationale of these studies was that if participants detect a semantically inappropriate word, then an N400 would be elicited. Spelling control words (e.g., beer) would be expected to produce a larger N400 response than would correct target words (e.g., bear), since the former are semantically inappropriate to the context (e.g., animal). The critical question was whether the N400 to homophone foils (e.g., bare) would be smaller than that observed for spelling controls, indicating that meanings were activated by phonology, or whether the N400 to homophone foils would be similar in amplitude to that observed for spelling controls, indicating that meanings were activated directly by orthography.
Ziegler, Benraiss, and Besson (1999) used an ERP study to investigate the role of phonology in a semantic categorization task modeled after Van Orden (1987). They expected that a direct O –> S pathway to meaning would result in a larger N400 to homophones (e.g., meet) than to exemplars (e.g., meat) of a semantic category (e.g., food). Orthographic control words (e.g., maid) were also used, and these stimuli were matched to the homophones for word frequency and number of letters. Behaviourally, it was found that participants made significantly more errors in rejecting homophone foils that sounded like members of the category than in rejecting orthographic control words; they also took longer to correctly reject the homophones. In the ERP data, however, results were more difficult to interpret. Ziegler et al. (1999) analyzed the latency bands in three epochs: 0-300 ms, 300-800 ms, and 800-1600 ms. The results were similar across the two earlier latency periods: homophone foils and orthographic controls did not differ in N400 amplitude, suggesting that readers relied on the direct route in accessing the meaning of the presented word. However, in the later latency period (800-1600 ms), homophone foils and orthographic controls diverged, and the homophone amplitudes closely resembled those of category exemplars (more positive-going). Ziegler et al. (1999) interpreted their results as providing evidence for a direct orthographic path to meaning, with phonology exerting an influence on meaning selection at a later stage, well after the direct activation of the appropriate meaning.
The first ERP study to examine the impact of phonology using a masked priming paradigm was conducted by Grainger, Kiyonaga, and Holcomb (2006). They employed the masked priming paradigm and kept prime duration constant at 50 ms. Based on previous ERP studies, they expected their manipulation of phonology and orthography to modulate the N250 component, which is hypothesized to reflect sublexical processing during visual word recognition, and the N400 component, which – as discussed – is thought to reflect semantic processing. Their central prediction was that the N250 component should be modified by both orthographic primes and phonological primes, and that the onsets of these two effects should be different. Results were analyzed in three latency bands: 150-250 ms, 250-350 ms, and 350-550 ms. In the earliest epoch, there was a significant effect of priming for the orthographic control condition at posterior sites, but no significant phonological effects. In the second epoch, a significant priming effect of pseudohomophones was seen reliably only at anterior sites, but none was observed for the orthographic condition. Finally, in the latest epoch (where the bulk of activity resembling the classic N400 component was observed), priming effects were significant for pseudohomophones across the scalp, while there was no significant effect for the orthographic condition once again. When it comes down to the time course of phonology, these results placed orthographic effects as emerging about 50 ms earlier than phonological effects. Grainger et al. concluded that their data provided clear evidence for fast phonological priming in conditions in which all possible contamination factors are eliminated, and provides an upper boundary for the time at which phonology starts to have an influence (approximately 250 ms post-target onset).
One of the objectives of the current research was to clarify the role of phonology in skilled reading using ERP measures. Grainger et al. (2006) failed to include an appropriate homophone in their study to compare results with those of the homophonic foil condition; the present study made use of both appropriate homophones and homophone foils, as well as an orthographic control condition. A masked priming paradigm was used in conjunction with a lexical decision task in order to examine the time course of phonology, with prime duration kept constant at 67 ms (see Ferrand & Grainger, 1993). Only low frequency homophones were used in the present study, based on work showing that homophone effects are largest when optimal selection competition is provided (a low frequency or unfamiliar exemplar is harder to identify as correct when put next to a low frequency error word than is a high frequency exemplar). It was expected that a homophone effect would be observed through a decrease in the N400 component in the homophone error primes relative to the spelling control words. Also of interest to the present study were the N250 component and the P600 component, said to be triggered when the brain encounters an unexpected linguistic item instead of the one that was expected (a post-lexical effect; Vissers, Chwilla, & Kolk, 2006). It was expected that no homophone effects would be observed in the N250 component, consistent with Grainger et al.’s (2006) observations.
The following is a proposal I wrote for an Advanced Cognition lab during my Psychology B.A. No actual ERP/EEG research was completed; all discussed research and anticipated results are based on peer-reviewed literature. Please do not submit any part of this work as your own.
ERP examination of background noise while learning: Repercussions on subsequent episodic memory performance
At ______ University, not unlike other academic settings, students are faced with a number of short-lived opportunities and often fall into poor studying habits. A brief observation of the main floor of the campus library demonstrates a social atmosphere that might distract from focus on class material: music, clacking keyboards, and loud conversations among friends discussing their plans for the evening.
Just how destructive this multi-tasking might be to a student’s grades depends on many factors, such as the amount of time these same students spend on the infamous “______ Advantage” laptops, which provide them with constant access to media. To understand how different noise sources affect memory is indeed important practical knowledge, especially true to those who invest heavily in an education; would students weigh their priorities more carefully if it could be shown scientifically and directly that applied focus pays off? The emphasis of the present study therefore is to examine, in an economically-valid laboratory setting, the flexibility and accuracy of students’ episodic memory retrieval based on encoding processes performed with the presence of auditory distractors.
The irrelevant speech effect (ISE) is the impairment of recall when irrelevant auditory material is presented together with the items to be memorized, competing for available processing resources. Numerous studies have demonstrated that white noise has no effect on to-be-remembered information, while nonvocal music shows a moderate ISE, vocal musical shows a greater one, and irrelevant speech (IS) shows the greatest ISE, regardless of the meaningfulness or loudness of the speech (Martin-Loeches, Scheweinberger, & Sommer, 1997; Klatte, Lee, & Hellbruck, 2002; Enmarker, 2004; Norris, Baddeley, & Page, 2004; Kopp, Schroger, & Lipka, 2006). However, meaningful IS is more disruptive than meaningless IS for reading comprehension and proof reading, and it has been shown to cause impairments on attention, as well as episodic and semantic memory (Enmarker, 2004).
Behavioural studies have focused on sorting between competing theories regarding the memory processes which regulate the ISE, of which few will be discussed. Baddeley’s model of working memory (WM) explains this effect through interference in the phonological loop, when irrelevant spoken material competes for attentional resources with the information being intentionally processed (1986, as cited in Martin-Loeches, Schweinberger, & Sommer, 1997; Klatte, Lee, & Hellbruck, 2002). The effect will theoretically be larger when the irrelevant material is phonologically similar to the target items. On the other hand, the changing-state hypothesis, as proposed by Jones in his object-oriented episodic record (O-OER) model, is the most promising theory when it comes to explaining the ISE, according to study results (Martin-Loeches, Schweinberger, & Sommer, 1997; Klatte, Lee, & Hellbruck, 2002). This theory suggests that IS is based on acoustic differences (changing-states) rather than phonological similarity between target and nontarget items (same-state); the higher the phonological variability, the stronger the perceived effect.
Klatte, Lee and Hellbruck (2002) examined the assumption that ISE occurs under the phonological loop model as discussed above, and four experiments were performed. In Experiment 1, participants practiced articulatory suppression (AS) during retention periods, a process in which rehearsal is prevented by the repetition of a word (i.e. “the”). This task is often given to participants by an experimenter because it is believed that visually presented items must be translated into a phonological code by means of subvocal rehearsal; in line with this assumptions are consistent findings that AS abolishes the ISE when items are presented visually, but not when presentation is auditory (Klatte, Lee, & Hellbruck, 2002; Norris, Baddeley, & Page, 2004).
The study design included both visual and auditory presentation, and background sounds (either irrelevant speech or broadband sound, which does not disrupt serial recall performance) were presented continuously. No significant difference was found between the visual and auditory groups, and performance was poorer overall in the IS conditions. In the visual presentation AS condition, the ISE was completely abolished, as was expected; in the auditory presentation AS condition, however, the ISE withstood suppression. This was in line with the assumption that speech gains obligatory access to the phonological store while visually presented verbal material must be translated into phonological code through subvocal rehearsal, which was successfully prevented by AS. The results of Experiment 1 set the basis of the subsequent experiments by confirmed Baddeley’s hypothesis of differential access to the phonological store for heard and read materials; visual material must be articulated in order to become susceptible to IS.
Experiment 2 explored recent findings that heard items can be retained in short-term memory without the help of serial recall for at least 10 seconds. The authors therefore predicted an ISE after a retention interval which exceeded the duration of the phonological trace originally postulated by Baddeley. Their results indeed contradicted the assumption of rapid temporal decay of phonological traces without rehearsal for auditory presentation, as the ISE remained even with AS and a longer retention interval.
Their results contradicted the idea that phonological traces decay rapidly when AS is used to prevent rehearsal. Experiment 3 was virtually identical and served mainly to confirm the findings of Experiment 2.
Finally, in Experiment 4, visual presentation was used and AS was confined to the retention interval as opposed to throughout the entire task. Participants were therefore free to rehearse during presentation of the stimuli. According to the phonological loop model, the ISE should withstand the suppression because rehearsal was permitted; the results showed that the ISE was able to survive a 10-second retention interval filled with AS, which seems to suggest that there is a long-enduring phonological trace which can be maintained for longer periods than initially imagined.
Overall, Klatte, Lee, and Hellbruck’s findings are in line with the assumption that auditory material gains direct access into the phonological store while visual stimuli must first be translated; however, they contradict the standard phonological loop model by demonstrating phonological traces which exist beyond a 10-second filled interval that does not allow rehearsal for maintenance. Suppression during presentation of the stimuli did abolish the ISE, but left it unaffected when initial processing occurred without suppression and with the distraction of irrelevant background noise. Their explanation for these events included the straightforward notion that competing auditory input leaves less capacity for effective rehearsal of learned material. The changing-state effect was also found in the study, where sequences of different syllables produced more disruption in memory than repetitions of single syllables, which contradicts Baddeley’s model.
Enmarker (2004) studied the effects of meaningful IS and road traffic noise on attention, episodic memory, and semantic memory in teachers and students. Their interest was based on findings that background noise causes detrimental influences on long-term memory performance and attention in children and young adults. A literature review led to the assumption that difficult tasks would suffer more from noise exposure (especially when it is unpredictable) because it is assumed that attentional resources are allocated toward the distraction, leaving less processing ability for the task at hand. This should be especially true when the concurrent cognitive processes utilize the same type of processing. Enmarker therefore hypothesized that verbal episodic memory should be impaired by meaningful IS more so than road traffic noise, though the competing auditory input would still lead traffic noise to impair attention.
It was also hypothesized that, because recognition tasks provide more environmental cues about items than recall tasks, the difficulty of the retrieval task should be based at least in part on availability of retrieval cues as well as how well the information was encoded; in other words, performance on recognition tasks was still expected to be impaired, but less so than in recall tasks.
Age-dependent factors were also considered in Enmarker’s study, because it is possible that, if memory ability plays a role in efficient processing, memory performance becomes more susceptible to noise interference as a person ages. Teachers were chosen as the main subjects in order to achieve a matched group in terms of education, while maintaining a subject base known for exposure to a demanding sound environment. The teachers were assigned to one of three independent groups in which they were exposed to either meaningful IS, road traffic noise, or silence. Conditions were carefully controlled so that noise assignment during encoding was the only true variable; for instance, each participant was run through at the same time of day, in a climate chamber with controlled air temperature and light level. Furthermore, in order to prevent semantic knowledge from confounding the results, the text read by participants had all real names and words replaced by imaginary ones. Text comprehension was then tested with 8 cued recall question and 12 multiple choice recognition questions; demographic variables were also collected, as well as questions concerning hearing and health status, and ratings about perceived effort used and degree of difficulty of the text.
Enmarker’s findings were that students aged 18-20 years outperformed both young and old teachers alike, which led to the proposal that episodic memory probably has its peak in early adulthood, maybe around 20-25 years. However, there was no interaction between age and noise, contrary to expectations; this was incompatible with the assumption that memory performance declines when attention is divided at the time of encoding relative to conditions of full attention. In line with the prediction, meaningful IS and road traffic both impaired cued recall of the text; the relative impact of the two noise sources did not differ for this task, likely because both of these had similar levels of acoustic variation. However, road traffic noise did not impair the recognition task whereas the meaningful IS did (though less so than in the recall task). There were no significant differences in the ratings of perceived difficulty of the text or effort used between the groups; hearing, attention, affect and annoyance did not mediate these noise effects. Enmarker concluded that, in order to develop a more elaborated understanding of noise impact on memory systems and attention, research must examine semantic content and acoustic variation independently.
Norris, Baddeley, and Page (2004) performed five experiments to examine the retroactive effects of IS on recall from short-term memory (STM). They also explored the assumption, recently challenged, that IS has the same effect on memory as AS, based on the fact that both impair recall by setting up competing streams in memory. Following a theoretical review they proposed that there may be no need for simultaneous presentation of IS with to-be-remembered items to achieve the ISE; this would once again be inconsistent with theories that predict the absence of an ISE when rehearsal is prevented. Previous findings provide evidence that IS presented before a list has no effect, while there is a significant effect of IS presented during the second half of a list or subsequently (during a retention period). Therefore, the authors concentrated on the conditions under which each of the models of STM (WM, O-OER, the feature model, and temporal distinctiveness theory or TDT) predicts retroactive effects of IS.
All conditions in Norris, Baddeley and Page’s experiments were designed to minimize the possibility of rehearsal, based on broad research concerning the processing times necessary in order for all desired effects to be achieved. In Experiment 1, the IS (excerpts of speech in a foreign language) was presented over headphones during the retention interval between presented lists and recall tasks; an arithmetic task prevented rehearsal. This avoidance of rehearsal was promoted as much as possible: participants were informed that recall tests would not be graded unless the simple arithmetic problem was done correctly.
As expected, it was found that performance was reliably better under conditions with no IS; the effect of interest was also found, when performance did not reliably differ between conditions in which IS was presented simultaneous to the list and those in which IS was presented after the list, during the retention period. Both of these conditions reliably differed from the condition in which IS was presented throughout the entire trial. Experiments 2-5 built on these results by attempting to demonstrate similar results with an overt task (digit-reading) rather than covert task (such as the standard procedure, AS) to prevent rehearsal.
In Experiment 2, the authors examined the common belief that the phonological store decays after a period of 3-6 seconds. Even after longer delays were used, a reliable retroactive IS effect was still produced. This suggests either that the phonological loop is not the locus of the ISE, or that IS affects an alternative system that short-term information is transferred to after longer delays. This experiment demonstrated that the ISE could be detected after as many as 12 seconds.
Experiment 3 had the same results as the prior when IS was presented simultaneous with list presentation, and retention length was added as a condition. It was thought that a short retention period would encourage use of the phonological loop, while a longer interval might make use of the proposed alternative short-term phonological store; participants were unaware of the length of the upcoming retention interval to avoid rehearsal style becoming a confound, if rehearsal was able to be used at all. It was noted that, when measured as a proportion of the number of errors at each retention interval, the ISE was greater for the shorter intervals, which suggests that the phonological loop does play a reliable role.
Experiment 4 directly address the question of whether the long-lasting effect of IS was contingent on the stimulus list having been in the phonological store at some point. AS during list presentation would prevent the visual information from being translated into the phonological store, which would force participants to rely on some alternative store; if IS influences this alternative store directly, a retroactive ISE should still be observed. The results confirmed that the ISE does depend on the information having first been encoded in the phonological store, though it does not necessarily need to be retrieved from it; in other words, whatever alternative store supports the delayed retroactive effects of IS does not seem to be sensitive to the presence of IS.
Experiment 5 served to push participants to their performance limits by using a more demanding digit-reading task and a shorter stimulus presentation period, in order to minimize the possibilities of rehearsal. This final experiment simply provided further confirmation of a retroactive IS effect. Overall, Norris, Baddeley, and Page were able to demonstrate that IS has retroactive effects on material already in memory in 4 of their 5 experiments. They were able to infer that rehearsal was not a factor based on consistently poor results in performance after a filled retention period of 9 seconds; they also based this assumption on the fact that, when asked, no participants reported being able to rehearse the list in Experiment 5. We can conclude from this study that the ISE is not mediated by its influence on rehearsal, and that it only emerges when items are first encoded into the phonological store.
Neuroimaging studies on the ISE have been scarce; Martin-Loeches, Schweinberger, & Sommer (1997) claim to be the first ERP study to directly manipulate IS. They proposed that the effect of an experimental manipulation on the ERP can be directly measured as a “difference wave,” subtracting the ERP obtained under a condition in which the experimental effect in question is absent or small from the ERP recorded when it is present. This procedure eliminates the activity of elements equally involved in both conditions, leaving the effects or activity produced by the experimental manipulation. Participants (N = 12, M = 26 years) with no neurological or hearing complaints completed the experiment in a sound-attenuated chamber. They were provided with a fixation point in order to minimize eye movements (which would cause recording artifacts) and a chinrest in order to stabilize head position. Furthermore, since rehearsal was allowed, they were asked to reduce movements of the mouth and tongue as much as possible and encouraged to use covert rehearsal methods. They were given instructions in writing as a standardization procedure. Presentation of stimuli and IS were auditory; subjects were advised to attend to the male voice (target) and ignore the female voice (nontarget or IS). Memorization should be achieved in order of presentation.
Only the results concerning Condition 2-1 are of interest to the present study, as this was the condition that introduced irrelevant speech. Subjects in this condition showed a long-lasting activity with positive values over frontal electrodes, and negative values over posterior and occipital electrodes (all were bilaterally distributed). Significant condition effects were present for the time windows between 200 and 700 msec, while the long-lasting activity of IS alone was most reliable in the 200- to 300-msec window. IS was also associated with a significant increase of error rates. Martin-Loeches, Schweinberger, and Sommer stated that their data suggest that the ISE is based on phonological variability within the irrelevant stream more than on phonological similarity between relevant and irrelevant material; however, it does appear from conditions involving phonological similarity that phonologically similar material requires more processing than does dissimilar material, increasing processing demands. Overall, these findings are consistent with the changing-state hypothesis and once again contradict the phonological loop model.
Finally, Kopp, Schroger, and Lipka (2006) used EEG to further investigate the effects of IS. Their literature review pinpointed certain structures that are related to short-term rehearsal: the premotor cortex, supplementary motor cortex, left prefrontal cortex, and cerebellar regions. Older PET studies have associated some neural structures that are involved in the ISE specifically; these studies showed distributed suppression of components of the verbal working memory network, particularly in left frontal and temporal brain regions. However, these PET studies and other imaging methods are limited when it comes to temporal dynamics of neural activity; this leads to an incomplete picture when it comes to memory processes, as oscillatory activity (particularly of the theta and gamma rhythms) are shown to be closely related to encoding, rehearsal and retrieval. Initial evidence was found that the neural basis of the ISE is the reduction of long-lasting synchronization of gamma activity in the underlying phonological rehearsal network; this study aimed to find these patterns under conditions of spoken and written recall. The authors predicted a reduction of gamma synchronization at left fronto-central sites from quiet to IS in the spoken conditions.
Participants of Kopp, Schroger, and Lipka’s study (N = 21, aged 18-32) all had normal or corrected-to-normal vision, and baseline working memory capacity was measured before the experiment through a digit span task. Word lists were matched in word frequency and semantic relatedness, and the trials were designed in a way that prevented participants from establishing elaborated rehearsal strategies. Participants were interviewed after the experiment about their rehearsal strategies, and subjective ratings on task difficulty were collected. All participants reported using a phonological strategy, and stated that any attempts at more complex strategies were given up during the practice trials when the item presentation rate proved too fast; in other words, attempts to prevent elaborated rehearsal strategies were successful. A main effect was found for distraction, F(1,20) = 26.3, p < 0.001, with a pronounced decline of performance in speech compared to quiet conditions. Gamma coherence did in fact decrease at central and left-frontal electrode combinations during the retention interval in the spoken recall condition; with written recall, however, no decrease was found from quiet to IS conditions.
It was proposed that an effect was found in spoken recall because articulating the first few items may have interfered with the remaining items in STM. Kopp, Schroger, and Lipka (2006) concluded that a possible future direction of distraction by IS is to explore its interaction with recall mode. Randomized spoken and written recall would have an effect on rehearsal methods since participants would not be able to anticipate the most efficient method.
As the theoretical explanations for the ISE are narrowed down and the mechanics of its processes are clarified, it becomes more appropriate to examine specific effects and interactions within the STM paradigm. Because the focus of the present study is on how IS affects learning in students, the information gained from behavioural and neuroimaging studies will be applied or adapted to a study which imitates student life as carefully as possible while experimentally controlling for extraneous variables. The conditions of interest in this study are presentation modality (visual, which would mimic independent study; and auditory, which resembles lecture-style learning), distraction (white noise, vocal music, and meaningful IS), and test type (recognition and recall). The design will employ intentional learning and allow covert rehearsal (covert in order to avoid movement artifacts), because students are aware that they will be tested in an academic setting. Finally, because it has been shown that simultaneous IS and IS during a retention interval do not differ significant, IS in this study will be continuous (presented throughout both text presentation and the interval).
The present study could have strong theoretical implications for the understanding of memory and attentional resources, but it would also be of significance to students who may not know how to make the most efficient use of their study time. Divided attention has shown a larger detrimental effect during encoding than during retrieval; therefore, emphasis in noise research has been on the encoding phase rather than on retrieval (Enmarker, 2004), and thus all memory tasks in this study will be performed in silence.
Hypotheses will be based on two main assumptions: first, that the difficulty of the retrieval task is affected by how well the information has been encoded, and also on the availability of retrieval cues (Enmarker, 2004); secondly, that the quality of encoding is affected by how large resources could be allocated to the task, as a result of resource competition and availability. It is hypothesized that participants in the white noise condition will show more activity in the brain areas associated with short-term rehearsal: the premotor cortex, supplementary motor cortex, left prefrontal cortex, and cerebellar regions. Subsequent memory for participants in the noise conditions (exposed to vocal music or meaningful IS) is expected to show increased negativity in the brain regions previously linked to the ISE (particularly in left frontal and temporal brain regions), which would support the idea that these memories were not as deeply encoded therefore still register as fairly unfamiliar to the brain. It is also expected that participants in the noise conditions will only be able to express learned information in a limited manner; recall task accuracy will be significantly lower for the music and IS groups than for participants in the white noise condition. It is expected that there will be little discrepancy between the three groups in terms of recognition task accuracy, with the IS condition reflecting the worst performance.
Participants (N = 16, 18-24 years of age) will be ______ University Psychology students in their third or fourth year of study in order to match participants for education level and estimate baseline intelligence. They will be right-handed, monolingual, and native speakers of English. They will be screened for problematic neurological histories, and all will have normal or corrected-to-normal vision and hearing. Participants will be paid or receive course credit in exchange for participation.
Instructions will be given in writing in order to standardize the procedure. Stimuli will consist of one practice paragraph and three true paragraphs adapted from a history textbook, with all real words and names replaced by imaginary ones to prevent semantic memory activation. Each paragraph will be composed of four sentences, all matched in length and overall approximate time-to-read (approximately one minute). Stimuli will be divided into three trials, and a partial Latin square will balance text order and distraction order so that each level appears once per participant, as well as memory test type order (either recall or recognition will appear twice for each participant). Ratings of perceived effort used and perceived test difficulty will be collected from participants after each trial. Demographic variables and GPA will also be collected in order to provide information regarding background similarities between participants and individual rough baseline intelligence levels; this information will not be analyzed and will remain confidential.
The experiment will be conducted in a sound-attenuated chamber, with both light level and temperature (21°C) controlled. Subjects will be wearing headphones, and will be seated 2 feet from a 19” laptop monitor and keyboard, with a chinrest to stabilize head position. In the visual presentation condition (N = 8), sentences will be presented on the monitor one word at a time to avoid excessive eye movements and recording artifacts, and the beginning and conclusion of sentences will be signaled with a fixation point, which will remain on screen for 100 msec after the participant pushes a key. Subjects will control the presentation rate by hitting the J key (allows them to move forward by a word) or the F key (allows them to move forward by an entire sentence); the paragraph is looped, so once the last sentence is complete, the first sentence reappears. In the auditory presentation condition (N = 8), the participant will not have the same control because a student would not be able to replay a lecture. The target file will be played slowly once over headphones in a male voice.
In both conditions, the distraction condition (nontargets) will be played over headphones; participants will be told to ignore these during the trials. White noise of constant amplitude will be played for the no-distractor condition. A popular “Top 20” song, rated in terms of lyrical content and acoustic variability, will be pulled from the radio for the vocal music condition, as this would best represent the type of music that the majority of students would listen to in their own time. The IS condition will be composed of excerpts of meaningful speech in a female voice; this will alternate with the male target speech in the headphones so that it is heard equally from each earpiece. All three conditions will be matched for frequency and loudness.
The recognition and recall memory tests will be presented on the monitor upon the completion of each trial. They will be randomized so that participants do not anticipate question type and rehearse accordingly. Recognition memory tests will appear in the form of a multiple choice question; first, the question will appear, and the participant presses any key to view each possible answer individually before they must make their response (one out of four possible answers is correct). Responses are made by selecting the screen that displays the correct answer and hitting the Enter button. The cued recall memory task is in the form of fill-in-the-blank questions, also presented one at a time. The participant must type in the appropriate word as a response; they may move among questions and return to enter a response at a later time. Both recognition and recall test questions come in groupings of three for their respective trials.
The ERP recording will be recorded from 64 sites on the scalp. The electrooculogram (OEG) will also be obtained. In order to measure tongue and mouth movements, an additional electrode will be located in the position most sensitive to subvocal articulation, which will be determined during the practice trial to accustom participants to the EEG equipment. Each epoch will begin 100 msec prior to target onset (it will begin when the subject presses a key at any fixation point in between sentences), and will end upon the appearance of the next fixation point. A baseline will be calculated by averaging the pretarget recordings, and the ERP will be digitally filtered from .1-20Hz to reduce noise. Ocular artifacts will be corrected, and trials with uncorrectable EOG artifacts or other facial activity (+/- 75mV) will be excluded from analysis. Finally, the signal will be averaged for each condition to increase the signal-to-noise ratio, at approximately 40 trials per condition. The dependent measures to be analyzed are the amplitude of the component and the latency of the component; the independent measures are distraction (3 levels), presentation modality (2 levels), and test type (2 levels).
Participants will be randomly assigned to either the visual or auditory presentation modality before they arrive to be tested individually in the study. Upon arrival, they will receive a standardized instruction sheet explaining the aim of the study and what they are expected to do over the course of the EEG recordings. Once informed consent has been attained, facial areas will be cleaned in preparation for the electrode cap. The cap will take approximately 15 minutes to set up, during which time the participants will be permitted and encouraged to review the instruction sheet. Upon completion of the set-up, a short practice trial will familiarize the participant with the task demands and finalize appropriate facial electrode locations. The experiment will not proceed until the participant fully understands the instructions.
Depending on presentation modality, participants will either notify the experimenter when they are prepared to hear the first trial (the auditory condition), or push any key to proceed from the fixation point to the first trial (the visual condition). Participants in the visual trial will be given three minutes to review the paragraph, while the run time for the auditory presentation will take approximately one minute (the same as the estimated time-to-read each paragraph in the visual condition). A retention period of 10 seconds will follow the end of target presentation. Participants in the noise conditions will hear this continuously from the initial target onset (the first sentence) until the end of the retention period. On the monitor, a 3-2-1 countdown of the last three seconds of the retention period will notify the participant that the test is about to begin.
Both the recognition and recall tests are only three questions long, and since the ISE is fairly short-lived, only a minute will be provided to complete the test. Blank answers will be scored as incorrect. All recall tests will be reviewed for spelling errors in case the program scores these as incorrect. Each trial will therefore take from 2-5 minutes.
Upon completion of the final trial, participants will be unhooked from the EEG and the predictions of the researcher, as well as previous results, will be described. A debriefing sheet will be provided, and the opportunity to be contacted with the results upon completion of analyses will be offered to the participant. Overall, each individual session should take no more than 50 minutes.
A 3×2 repeated measures ANOVA will be performed on the data from the visual presentation condition (see Figure 1). It is expected that performance on memory tasks will show a significant main effect of Distraction, with the white noise condition demonstrating the highest response accuracy overall and the IS condition demonstrating the lowest accuracy, with the music condition close behind IS. It is also expected that a significant main effect for Test Type will demonstrate that memory performance in recognition tasks is superior to performance in recall tasks. Furthermore, a Distraction x Test Type interaction is expected; t-tests will demonstrate that, while performance from recognition to recall tasks decreases for both music and IS conditions, there will be not be a significant difference between recognition and recall test results in the white noise condition.
A 3×2 repeated measures ANOVA will also be performed on the data from the auditory presentation condition (see Figure 2). Identical results to the above are expected, with greater discrepancies between the white noise and IS conditions due to the fact that cognitive processes are now competing for the same resources.
ERP recordings will demonstrate general increased negativity over brain areas associated with STM and verbal working memory.
Importance of Anticipated Results
ERP examinations of the ISE have barely made their way into the literature; this report supports recent behavioural literature while integrating neuroimaging data that may eventually help us understand not only how, but where in the brain and why these effects occur in memory processes. While the temporal advantages of EEG make ERPs worthwhile, future studies may focus on using the modern imaging technique of fMRI to locate the areas of increased and decreased activation associated with IS. This could better narrow the search, so to speak; otherwise, analyses of the huge amounts of data provided by ERP studies are not much better than fishing in the brain.
Enmarker, I. (2004). The effects of meaningful irrelevant speech and road traffic noise on teachers’ attention, episodic and semantic memory. Scandinavian Journal of Psychology, 45(5), 393-405.
Klatte, M., Lee, N. & Hellbruck, J. (2002). Effects of irrelevant speech and articulatory suppression on serial recall of heard and read materials. Psychologische Beitrage, 44(2), 166-186.
Kopp, F., Schroger, E., & Lipka, S. (2006). Synchronized brain activity during rehearsal and short-term memory disruption by irrelevant speech is affected by recall mode. International Journal of Psychophysiology, 61, 188-203.
Martin-Loeches, M., Schweinberger, S.R., & Sommer, W. (1997). The phonological loop model of working memory: An ERP study of irrelevant speech and phonological similarity effects. Memory & Cognition, 25(4), 471-483.
Norris, D., Baddeley, A.D. & Page, M.P.A. (2004). Retroactive effects of irrelevant speech on serial recall from short-term memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 30(5), 1093-1105.
Figure 1: Anticipated results
Figure 2: Anticipated Results
The following is a paper I wrote for my advanced Cognition lab during my Psychology B.A. It was the basis of research I was later to use for my Honours Thesis (not yet posted on the blog). Please do not copy or submit any of this article as your own work.
Homophone Effects and Pseudohomophone Effects in Visual Word Recognition
The current study examines homophone and pseudohomophone effects in visual word recognition to see whether word frequency, phonological nature and orthographical nature significantly influence the time it takes to process the meaning of a word, as proposed by Pexman, Lupker and Jared (2001). Participants (N=22) saw lists of stimuli that included both homophones and pseudohomophones and varied across the three aforementioned factors, and they completed a Lexical Decision Task. Data was analyzed separately for the Word and Nonword using Repeated Measures ANOVAs and Paired t-tests for the interactions. All hypotheses were supported with statistically significant results, where larger homophone and pseudohomophone effects were observed for low frequency words with higher frequency mates, orthographically similar (wordlike) words, and phonologically similar words (pseudohomophones).
Homophone Effects and Pseudohomophone Effects in Visual Word Recognition
There is a debate within the research on visual word recognition as to whether the orthographical or phonological representation of a word is used to compute its meaning. While it is accepted that phonology is influential, there is the question of whether our use of phonology is strategic, or whether it can sometimes hinder our performance when we do not mean to use it. These effects are commonly studied through the use of homophones and pseudohomophones. Homophones are two or more words that share the same phonology but possess different orthographies and meaning (e.g., made and maid). Pseudohomophones are nonwords that have the identical phonology of a real word (e.g., golph). The rationale for using these is that, if phonology mediates word recognition, we would expect the presentation of a homophone to result in the activation of its mate, and the presentation of a pseudohomophone to result in the activation of the pseudohomophone’s base word (e.g., golf).
A homophone effect can be defined as a slowing in the time it takes to process the meaning of a word due to various influences on phonological processing. This is often measured using a lexical decision task (LDT), in which a person makes a decision regarding whether an item is a word or a nonword. The rationale for using LDT to study homophone effects is that, if phonology is being activated pre-lexically, we could predict that homophones would delay word responses while pseudohomophones would delay nonword responses. Many factors have been shown to influence the extent to which homophone effects are observed. These include the word frequency within the English language, the reader skill (in which low-skilled readers rely more on phonology or the indirect route of reading), and the orthographic familiarity of the word (whether it resembles the typical structure of an English word).
Van Orden (1987) performed a series of semantic categorization studies. Participants were asked to determine whether a presented word was part of a category; words were paired according to orthographic similarity as a spelling control so that any effects must be due to phonology. In the category “Flowers,” Van Orden noted that there were more errors to the foil ‘rows’ than to its spelling control ‘robs’; similarly, there were more errors to the foil ‘roze’ than to its spelling control ‘rone’. This was referred to as phonological interference, to signify that the sound of the foil was interfering with participants’ decision making processes.
Jared and Seidenberg (1991) examined phonology in reading and the influence of word frequency. They concluded that homophone effects in semantic categorization studies are largest for low frequency homophones, because people tend to use a phonologically mediated route to process low frequency (LF) words, while they use a direct route (orthography à meaning) for high frequency (HF) words. Jared, Levy, and Rayner (1999) found similar results: while proofreading, participants made more errors when judging whether a LF word was used appropriately in the context of a sentence than if the word had a HF in the English language, but this homophone effect was mainly seen in low-skilled readers. Skilled readers used a more direct root (they relied more on the orthographic representation), where less skilled readers rely more on phonology.
Pexman, Lupker and Jared (2001) discussed factors that modulate the size of homophone effects in lexical decision over six experiments. They used rigorous stimulus controls by matching the word frequency of homophones and controls and matching the bodies (orthographic similarity) of pseudowords and pseudohomophones (RADE, CADE). Words were presented in either the context of a pseudoword or a pseudohomophone (but not both). The results of one experiment showed that, in the pseudoword context, a homophone effect was found for LF homophones only; in the pseudohomophone context, the effect was seen for both low- and high-frequency homophones. Another experiment looked at words that were orthographically atypical of the English language, and it was reasoned that homophone effects should be reduced if participants made use of the orthographic information – indeed, in the pseudoword context, no homophone effects were found (they were found only for LF homophones in the pseudohomophone context). It was concluded that there are three main influential factors in homophone effects: homophone frequency, orthographic nature, and phonological nature. Homophone effects should be larger for LF homophones with higher frequency mates, for wordlike (orthography typical of English words) as opposed to non-wordlike words, and for pseudohomophones (words that sound like real words) rather than pseudowords (nonwords). Furthermore, they asserted that people use phonology to make judgments for LF words, and orthography for HF words.
The present study is an adaptation of Pexman, Lupker and Jared’s (2001) study with two major changes: participants will view both pseudowords and pseudohomophones (in order to study orthographic typicality), with likeness as a within-subjects factor as opposed to a between-subjects factor. There are two main questions that will be examined: do effects of orthographic typicality extend to a within subject’s design, and are pseudohomophone effects dependent on orthographic typicality and list context? Three main hypotheses will be tested: homophone effects will be larger (in other words, reaction times will be slower) in a wordlike context than a non-wordlike context, larger for low frequency words with higher frequency mates than for high frequency words, and larger for pseudohomophones than for pseudowords.
Participants were a mixture of undergraduate and graduate students from the University of Western Ontario (N = 22, 9 females, mean age = 26.1 years). All participants were right-handed, spoke English as their first language, and had minimal exposure to a second language. They had no history of visual or auditory impairments, neurological or psychiatric disorders, or reported learning or reading problems. No compensation was received for participation. Data was collected by researchers at UWO and analyzed in the present paper by those at ______ University.
Two sets of stimuli were used: word stimuli and nonword stimuli. The word stimuli, composed of 80 items, included 40 homophone pairs (20 high frequency and 20 low frequency) and 40 controls matched in frequency. The nonword stimuli, composed of 160 items, included 80 pseudohomophones (40 wordlike and 40 non-wordlike) and 80 pseudowords, also equally split into word-likeness (see Appendix A for stimuli examples). Four lists were created from the total set of stimuli, and two were presented to subjects: one list was a Wordlike (WL) context, and the other was a Non-Wordlike (NW) context. Both pseudohomophones and pseudowords were presented in a single list. The order of list context was counterbalanced across subjects. Furthermore, each word was counterbalanced such that it appeared in both contexts. Each word and nonword was presented only once to a participant.
The above resulted in 12 list conditions: wordlike context high frequency homophone (WLHFHom), wordlike context low frequency homophone (WLLFHom), wordlike context high frequency control word (WLHFCon), wordlike context low frequency control word (WLLFCon), wordlike context pseudoword (WLPsW), wordlike context pseudohomophone (WLPsH), non-wordlike context high frequency homophone (NWHFHom), non-wordlike context low frequency homophone (NWLFHom), non-wordlike context high frequency control word (NWHFCon), non-wordlike context low frequency control word (NWLFCon), non-wordlike context pseudoword (NWPsW), and non-wordlike context pseudohomophone (NWPsH).
Subjects were seated one meter from a 19-inch computer monitor. Stimuli were presented in black 12 pt. Arial font in uppercase letters against a white background. Subjects were required to indicate via a button press if the presented letter string was a word (right button) or a nonword (left button), and they were instructed to respond as quickly and accurately as possible. They completed 10 practice trials with verbal feedback prior to beginning the experiment. For the actual experiment, subjects viewed a fixation cross for 500ms and then a blank screen for another 500ms. A stimulus then appeared and remained on screen until the participant responded with their lexical decision. The process started over with another fixation cross, and the series continued to repeat with a new stimulus each time.
All data were analyzed in SPSS 14.0 (see Appendix B for full output). A 2x2x2 Repeated Measures Analysis of Variance with context, word type, and frequency as the within subject variables was conducted on the Words list condition. A significant effect of context was found, F(1,21) = 28.9, p < 0.001, where wordlike words (M = 797.213, SD = 31.784) had slower response times than non-wordlike words (M = 588.976, SD = 15.569). A significant effect of word type was also found, F(1,21) = 28.1, p < 0.001, where homophones (M = 713.455, SD = 16.680) had slower response times than control words (M = 672.733, SD = 15.923). Finally, a significant effect of frequency was found, F(1,21) = 85.2, p < 0.001, where HF words (M = 642.164, SD = 13.473) had faster response times than LF words (M = 744.025, SD = 19.536).
A significant interaction was found between context and word type at F(1,21) = 14.09, p < 0.002 (see Figure 1). A paired t-test on this interaction found a significant value of t(21) = 4.859, p < 0.001 for the comparison between wordlike homophones (M = 829.468, SD = 148.748) and wordlike control words (M = 764.957, SD = 155.763). A significant t-value, t(21) = 3.612, p < 0.01, was also found for the comparison between non-wordlike homophones (M = 597.442, SD = 72.466) and non-wordlike control words (M = 580.510, SD = 75.201). These t-tests indicate that participants were faster to respond to control words than homophones; this difference was less pronounced in the non-wordlike conditions, though participants took less time overall to respond to these.
A significant interaction was also found between context and frequency at F(1,21) = 13.56, p < 0.002 (see Figure 2). A paired t-test on this interaction found a significant value of t(21) = -6.852, p < 0.001 for the comparison between wordlike HF words (M = 725.421, SD = 120.838) and wordlike LF words (M = 869.004, SD = 186.222). A significant t-value, t(21) = 4.454, p < 0.001, was also found for the comparison between non-wordlike HF words (M = 633.919, SD = 67.292) and non-wordlike LF words (M = 558.907, SD = 65.765). These t-tests indicate that participants were faster to respond to high frequency rather than low frequency words; again, they took less time overall to respond to stimuli in the non-wordlike context.
Finally, a significant interaction was found between word type and frequency at F(1,21) = 15.28, p < 0.002 (see Figure 3). A paired t-test on this interaction found a significant value of t(21) = -2.632, p < 0.02 for the comparison between HF control words (M = 633.919, SD = 67.292) and HF homophones (M = 650.409, SD = 62.378). A significant t-value, t(21) = -5.209, p < 0.001, was also found for the comparison between LF control words (M = 711.548, SD = 87.545) and LF homophones (M = 776.502, SD = 104.112). These t-tests indicate that participants were faster overall making decisions about control words. Participants were also much faster to respond to high frequency words than low frequency words, as shown above.
A 2×2 Repeated Measures ANOVA with context and pseudotype as the within subject variables was conducted on the Nonwords list condition. A significant effect of context was found, F(1,21) = 28.18, p < 0.001, where wordlike nonwords (M = 852.872, SD = 36.877) had slower response times than non-wordlike nonwords (M = 614.843, SD = 16.809). A significant effect of pseudotype was also found, F(1,21) = 59.96, p < 0.001, where PsW (M = 716.531, SD = 17.943) had faster response times than PsH (M = 751.184, SD = 18.031). A significant interaction was found between context and pseudotype at F(1,21) = 33.35, p < 0.001 (see Figure 4). A paired t-test on this interaction found a significant value of t(21) = -7.437, p < 0.001 for the comparison between wordlike PsW (M = 819.566, SD = 171.782) and wordlike PsH (M = 886.178, SD = 176.661), indicating that participants were faster to respond to pseudowords than pseudohomophones. A significant difference was not found between pseudowords and pseudohomophones in the non-wordlike condition.
These results supported all of the hypotheses. Though there was not a three-way interaction, two-way interactions were found between all of the variables, showing that some combination of factors influences the observed effects in different ways, and main effects were consistently found. Paired t-tests clarified the interactions to reveal the specific combinations or levels that produced these effects: homophone and pseudohomophone effects were observed through slower reaction times to pseudohomophones than to pseudowords, to low frequency words than to high frequency words, and to words in a wordlike as opposed to those in a non-wordlike context. This is a powerful replication of Pexman, Lupker and Jared’s (2001) results, considering it was done in a within-subjects design and also extended to a context where subjects were exposed to both homophones and pseudohomophones.
There were not any outstanding limitations to the present study. Improvements on the Pexman, Lupker and Jared (2001) study design increased the power but maintained the well-matched and counterbalanced aspects of the stimuli, and a large number of data points were collected. These results, like any that clarify our understanding of visual word recognition, have implications in education and rehabilitation for people who are learning how to read or correcting a reading deficit. They also provide the opportunity further theoretical developmental about how we compute meaning from words, and lend direction to future research questions.
Jared, D., Levy, B.A., & Rayner, K. (1999). The role of phonology in the activation of word meanings during reading: Evidence from proofreading and eye movements. Journal of Experimental Psychology: General, 128(3), 219-264.
Jared, D., & Seidenberg, M.S. (1991). Does word identification proceed from spelling to sound to meaning? Journal of Experimental Psychology: General, 120(4), 358-394.
Pexman, P.M., Lupker, S.J., & Jared, D. (2001). Homophone effects in lexical decision. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27(1), 139-156.
Van Orden, G.C. (1987). A ROWS is a ROSE: Spelling, sound, and reading. Memory and Cognition, 15(3), 181-198.
The following is a paper I wrote for an advanced Cognition class during my Psychology B.A. Please do not submit this paper as your own work.
Qualitative nature of encoding and memory retrieval: Levels of processing
The current study examines the effects of level of processing on subsequent memory recognition and recall. Undergraduate students (N = 13) participated in a classroom task to test memory for stimuli processed according to case, rhyme or semantics. They made judgments concerning whether a target word was an appropriate response to a previously viewed question, and filled out first a recall then a recognition test. It was hypothesized that, for both tests, memory for words in the semantic condition would be greatest, followed by those in the rhyme condition and finally those in the case condition. Two Repeated Measures ANOVAs found main effects for question type, and Paired t-tests supported the hypotheses by demonstrating the predicted relationships.
Levels of Processing
Memory has been an area of intense research focus for decades because it plays a role in just about everything we do, though it tends to go unnoticed until a failure occurs. Numerous processes (mainly attention, encoding, rehearsal and retrieval) must be studied in order to achieve even a basic understanding of the complexity of memory. The question of levels of processing, which concerns encoding and retrieval processes, is important because it could have extensive theoretical and educational implications. Some relevant findings have included the idea that it is the operations carried out on material that determine retention, rather than a simple intention to learn (though attention is of course necessary). Also, different encoding operations are important to different retrieval conditions, a finding which presents the factor of context into learning processes. Knowledge about these mechanisms could change the way teachers present difficult material for the first time, or the way students prepare for an upcoming final. To understand the basic operations of learning and memory would indeed have significant implications in our school systems.
Many theorists have attempted to explain the process of remembering based on a model of the constituents of memory discussed above. Craik and Lockhart (1972) presented the depth of processing view of memory, focusing on encoding operations themselves. They claimed that a memory trace is the record resulting from the perceptual and cognitive analyses carried out on the stimulus. Preliminary stages of perception concern physical or sensory features, whereas the later stages are more concerned with matching the input against past learning to extract meaning; therefore, greater depth implies a greater degree of semantic analysis (meaningful events are better retained due to deeper processing). It was also argued that memory trace may be less or more elaborate and durable depending on the number and qualitative nature of the perceptual analyses carried out on the stimulus, as well as the degree of attention and time devoted to analyzing the stimulus (deeper processing is assumed to require more time, unless the material is of high familiarity or frequency). Craik and Lockhart concluded from their review that an important goal of future research would be to specify the memorial consequences of various types of perceptual operations.
Schulman (1974) used an incidental learning task to determine whether subjects’ semantic judgments about congruity (the appropriate use of a word) led to memorial discrepancies due to differences in processing. Consistent results across the conditions of his study led to what is called the principle of congruity: congruous queries about words yield better memory for the words than incongruous queries. He also found that when false alarms occurred in a recall test, the intrusions were most often semantically related to the word to be recalled; this is evidence that conceptual storage processes are widely involved in encoding. Schulman’s concluding argument was that the words of an incongruous query are encoded as unconnected semantic units, while the words of a congruous query are relationally encoded.
Craik and Tulving (1975) claimed that it is the qualitative nature of the encoding achieved that is important for memory; they concurred with past findings that semantic processing is superior to structural processing. They performed ten experiments within the levels of processing framework proposed by Craik and Lockhart (1972), examining the plausibility of basic notions of the framework, in order to rule out alternative explanations. They attempted to achieve a better characterization of depth of processing while continuing to critically question the model. Subjects were induced to process words on different levels by answering questions that required judgments about typescript, rhyme, or categorization, and then taking part in either a recognition or recall test (which was unexpected in most cases). Time to make decisions was also measured to see if it was correlated with depth of processing. Craik and Tulving found that, consistent with previous findings, deeper encodings took longer to process on average, and there were two significant predictors of high levels of performance on a subsequent memory test: deeper levels of encoding and congruity, as established by Schulman (1974).
The present study is based on Craik and Tulving’s (1975) Experiment 9, in which an intentional learning condition tested the raw occurrence of observable depth of processing effects. Based on prior studies, it is hypothesized that participants will show greater memory traces for words encoded through semantic judgments than those based on phonology (rhyme); furthermore, memory for words encoded through phonological judgments will be superior to memory for words processed orthographically (case). It is also expected that recognition tests will yield better memory than recall tests due to context effects.
Participants were members of the _____ University community (N = 13); most were students (age range 19-21). Six students participated out of a third year Cognition lab in order to provide their own data for analysis. These students then each conducted the experiment on a peer (age range 17-22) to provide additional data for the study.
A Microsoft Office PowerPoint presentation was used to display the stimuli. There were 48 trials presented, in question-answer format. Each question remained on screen for five seconds, followed by a word for one second. The next question appeared after five seconds to allow for response time. Each trial led to one of three encoding types: a case judgment (“S1: Is the word in capital letters? S2: flour”), a rhyme judgment (“S1: Does ______ rhyme with each? S2: SPEECH”) or a semantic judgment (“S1: Is a _____ a type of weapon? S2: knife”). Target stimuli were counterbalanced in terms of question type, case (half of the words were written in uppercase), and congruity (half of the words were congruous, or ‘yes’ judgments). All participants made decisions about the same sequence of words.
Three separate forms were used to collect data for the memory tasks. The first was used during the initial presentation of the stimuli to aid in the encoding process and attain accuracy values, to ensure the participant understood directions (see Appendix A). The second was a recall test, in which 48 spaces to recall words seen were provided (see Appendix B). The third handout was a recognition test with 132 items, 48 of which were old words (see Appendix C). The recognition task was presented last so not to confound the recall results.
Participants were informed that the purpose of the experiment was to examine an element of memory. They were instructed that, if the target word shown was an appropriate answer to the question posed, the participant should circle ‘yes’ on the sheet provided to them; otherwise, they should circle ‘no’. They would be given five seconds to make that decision before the next question was presented. Participants were also requested not to write any words down during the course of the experiment, as they would be given a series of memory tasks following the completion of the trials.
Once participants had been fully instructed, the decision forms were distributed and the PowerPoint presentation began. Upon completion of the trials, the forms were set aside (but not yet collected). Participants were instructed that they would have ten minutes to complete a recall test, and these forms were distributed; upon completion, they were set aside with those from the encoding process. Participants were instructed that the final task was a recognition test, and that they would again be given ten minutes to complete it. They were to circle exactly 48 words from the list given that they believed had been shown during the initial PowerPoint presentation. At the end of this ten-minute period, all participants inserted their question response and recall sheets into the centre of the recognition test sheets, and the bundles were collected by the experimenters.
All data were analyzed in SPSS (see Appendix D for full output). A 1×3 Repeated Measures Analysis of Variance on the recognition test results found a significant effect of question type, F(2,24) = 19.25, p > 0.001, where proportion recognition of words in the semantics condition was highest (M = .807, SD = .12), followed by the rhyme condition (M = .625, SD = .16), and finally the case condition (M = .462, SD = .14). A second 1×3 Repeated Measures ANOVA on the recall test results also found a significant effect of question type, F(2,24) = 55.69, p > 0.001, with the same general pattern: words in the semantic condition were more often recalled (M = .327, SD = .11), followed by the rhyme condition (M = .125, SD = .08), and finally the case condition (M = .091, SD = .08). When recognition and recall standard error means were plotted, it was apparent that participants had better memory for words in the recognition task (see Figure 1).
Paired sample t-tests were conducted on both recognition and recall data. A significant t-value was found between case recognition and rhyme recognition, t(12) = -2.67, p < 0.05. The comparison between rhyme recognition and semantic recognition was also significant at t(12) = -3.29, p < 0.01. In the recall condition, only the comparison between rhyme and semantics was significant at t(12) = -12.57, p < 0.001.
Results supported the hypotheses and were consistent with Craik and Tulving’s (1975) results. An effect of question type was found in both recognition and recall conditions, with semantic encoding trials consistently resulting in the highest mean proportion of remembered words, and those processed at a structural level associated with the worst memory performance. Recognition memory was superior overall to recall memory. In other words, participants were highly successful remembering words that they had processed in terms of category or concept when they had to pick it out of a list of words; their performance was poor, however, in the case where a word was processed structurally (according to case) and they were required to name it in free recall.
The results of this study have implications in education and theories of information processing. The more we understand about the way we learn, the more we can apply this knowledge to everyday behaviour. There were some limitations in the present study, however. It is possible that there was a time confound, since the recognition test occurred later in time than the recall test; perhaps memory traces for words would have been stronger had at least ten minutes not occurred since the initial presentation of the stimuli. Another confound could have been the fact that this study used intentional learning conditions; the experimenter loses some control in the sense that the participant is instructed to learn and may use an unknown coding strategy. However, because the stimuli were presented in close temporal proximity, any efforts at maintenance rehearsal would have likely been disrupted by competing tasks of working memory. Because the participant had to attend to judgments about the next word at a regular interval, this decreases the likelihood of the intentional learning approach becoming problematic. With regards to the results’ implications in education, an intentional learning task would have more economic validity because students learn with the assumption that they will be tested.
Future studies should attempt to replicate these results with a greater number of participants in a between-subjects fashion, with participants either completing the recognition or the recall test; this could increase the power if indeed the extra time between the initial stimulus presentation and the subsequent recognition test was a limiting factor in the present study.
Craik, F.I.M. & Lockhart, R.S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11, 671-684.
Craik, F.I.M. & Tulving, E. (1975). Depth of processing and the retention of words in episodic memory. Journal of Experimental Psychology: General, 104(3), 268-294.
Schulman, A.I. (1974). Memory for words recently classified. Memory & Cognition, 2, 47-52.
The following is a paper on the Capgras Delusion written for my Neuropsychology class in the final year of my Psychology B.A.
Doctor, There’s an Impostor in My Bed! Perceptual Anomalies in the Capgras Delusion
Imagine that you are eating lunch with your mother when you realize something about her is different. It is not a physical alteration, such as a new haircut or glasses, so much as the uneasy feeling you get while looking at her. As the weeks go by you continue to treat her with suspicion; you carefully inspect her behaviour and notice that she is responding strangely. The obvious truth finally sinks in—she is not your real mother! Though practically identical, this woman is clearly an impostor!
Welcome to the life of a Capgras syndrome patient, where anything from the people or pets to the places and objects you hold most dear may suddenly feel so unfamiliar that a delusion emerges: that of their substitution or duplication, perhaps by an actor, a robot, or simply, innocently, their identical twin figure.
The first well-documented case of this syndrome was an article by Capgras and Reboul-Lachaux (1923, as cited in Doran, 1990). Their patient was a 53-year-old woman who had the illusion that people in her life were being replaced by identical doubles; at one point, she came to believe that she herself had been duplicated. This “illusion of doubles” came to be coined ‘Capgras Syndrome’, and is now considered the prototype of various delusional anomalies categorized as misidentification syndromes.
Symptoms of Capgras tend to be monothematic and circumscribed. In other words, the impostor delusion tends to stand out against an otherwise normal belief system—“an island of irrationality within a sea of rationality” (Stone and Young, 1997). General trends within these patients, as summarized by Stone and Young, include difficulty in certain facial processing tasks, lack of interest concerning the fate of the “original” duplicated person, suspiciousness and occasional hostility (but mainly agreeableness) toward the double, and little recasting of beliefs to accommodate the delusional one. There is no evidence of overall reasoning deficits among these patients, especially considering that most feel it necessary to provide evidence for their claim.
Though Capgras typically accompanies a diagnosis of paranoid schizophrenia and is most often encountered in patients of middle or old age, over one third of cases occur in conjunction with traumatic brain lesion (Ramachandran and Blakeslee, 1998), especially injury to regions of the brain implicated in the processing of faces. As discussed in a detailed review by Stone and Young (1997), the delusions are often linked to lesions in the parieto-temporal regions of the right hemisphere.
Specifically, facial recognition is thought to be processed along two neural routes of the visual system, as proposed by Bauer (1984, as cited in Stone and Young, 1997): the ventral route for overt recognition or identification, and the dorsal route to the limbic system for affective significance. Cognitive dissonance resulting from a dysfunction in this second pathway is hypothesized to be at the root of Capgras syndrome. Brain imaging studies have provided evidence of dysfunctional connections among frontal cortex, multimodal association areas and paralimbic structures, resulting in cognitive-perceptual-affective dissonance (Papageorgiou et al., 2003); this suggestion of dissonance is consistent with the popular hypothesis that Capgras patients are able to recognize someone but not experience the appropriate affective responses. Since prosopagnosia, an inability to recognize faces, is thought to result from damage to the ventral route, these two syndromes are often considered mirror images of one another (Young and Ellis, 1990, as cited in Stone and Young, 1997). Finally, worsening of the condition is promoted by the vicious circle of confirmation bias: the patient looks for strange behaviours in others, and in response to this unusual behaviour, others act strangely toward the patient. This makes it difficult to break from a cycle of self-fulfilling prophecy.
Most reported cases of Capgras were diagnosed before modern diagnostic criteria and current neuroimaging methods; therefore, it should come as no surprise that the previously widely accepted explanation for the delusion was proposed by the psychodynamic field of psychology. It was believed that a patient created a double according to a defense mechanism due to guilt or conflicting emotions, most notably the Oedipus or Electra complex. Since a person would experience internal dissonance at discovering sexual feelings toward a parent or family member, it was proposed that they created a double in order to replace the object of their emotions and relieve this tension. However, current technology caused a shift from the psychodynamic foundation to viewing Capgras as a disorder of the central nervous system (Doran, 1990); advances in brain imaging clearly show some neurological damage in the majority of cases. Furthermore, the fact that Capgras delusions exist for pets, among other things, further weakens the psychodynamic approach.
Though psychological factors cannot be ruled out of the etiology, there seems to be an undeniable organic or physiological basis to misidentification syndromes. Stone and Young (1997) discuss two factors that partially explain the existence of Capgras delusions: a perceptual deficit and a reasoning bias. The Perceptual Deficit Account, similar to Fine, Craigie and Gold’s (2005) Explanation Account, claims that the delusion is the sufferer’s attempt to account for an anomalous perceptual experience. This account is incomplete on its own because the questions of why the impostor belief persists or that other sensory modalities do not override the faulty interpretation of the visual experience remain to be answered.
The Reasoning Bias Account claims that the original misperception is triggered by an organic process but that the delusion itself is determined by the psychology of the patient; in other words, the delusion is caused by a misinterpretation of an unusual experience. Delusional patients have abnormal reasoning processes: they tend to request less information before making a decision and jumping to conclusions (Huq et al., 1988, as cited in Stone and Young, 1997) and they have a tendency to attribute even hypothetical negative events to external causes (Candido and Romney, 1990, as cited in Stone and Young, 1997). Therefore, they are experiencing biases (rather than deficits) by using available information in a certain way through otherwise normal processes; for instance, by attributing their lack of affect to external consequences, they may be mistaking changes in the self for changes in others.
Both accounts are insufficient to determine why Capgras patients maintain these implausible beliefs despite available contradicting evidence. Stone and Young (1997) explore a philosophical model of belief formation to determine how patients might validate their decisions. Maher (1988, as cited in Stone and Young, 1997) claimed that delusional beliefs are formed the same way as normal and scientific ones: through a balance of observational adequacy and conservationism. It was proposed therefore that Capgras patients rely too heavily on observational adequacy (they have a bias toward first-hand experiences or the evidence of their senses), but maintain the integrity of their own personal beliefs by isolating this delusional belief, which is why the delusion tends to be circumscribed.
When patients feel that they must further validate their claims or beliefs, they have tended to refer to slight physical differences between the original and the impostor. Though this exploration is a step closer to resolving the dilemma of why the impostor belief is not rejected, it cannot explain the consistency of the impostor belief over other potential rationalizations. As put by Fine, Craigie, and Gold (2005):
The delusional belief is just one of several hypotheses that are observationally adequate… Many neurologic conditions result in direct perceptual abnormalities (e.g., acquired color blindness), yet we would not expect a corresponding delusion (e.g., that the world has been drained of color) as a default.
Furthermore, hypothesizing an abnormality in a patient’s belief acceptance system insinuates general patient susceptibility to any number of beliefs when an experience conflicts with prior knowledge. It is yet unknown whether Capgras patients experience subtler but similarly strange delusions.
The possibility that Capgras patients experience widespread emotional impairments, and not simply those pertaining to their loved ones, has been explored. This accompanies the suggestion that the misinterpretation of sensory evidence results from the fact that there is a larger discrepancy between, say, the affective response you should have to your mother as opposed to the delivery boy, so the missing affect toward the mother would be more noticeable. It has indeed been noted that patients often report persistent loss of affective response and generalized feelings of unfamiliarity, and they often seem to lack appropriate accompanying emotion (Christodoulou, 1977, as cited in Stone and Young, 1997). This could also explain why patients show little concern regarding the whereabouts or wellbeing of the original that was replaced.
Though this idea (that all emotions have been reduced due to some injury to the amygdale) seems intuitive, a case study by Hirstein and Ramachandran (1997, as cited in Ramachandran and Blakeslee, 1998) showed that this is not necessarily the case. Ramachandran hypothesized that his Capgras patient would not register a change in skin conductance when looking at pictures of his family intermixed with those of strangers. Control participants had a large response to their parents as expected, whereas the delusional patient showed no increased response; however, at times there would be a blip after a long delay, as if the brain were doing a double take. This showed that the patient was not responding emotionally to his parents. The experimenter kept track of the patient’s other emotions over a period of months to determine whether there was a global disturbance in emotion, but his responses were appropriate and diverse in all other situations. It was hypothesized that the large discrepancy between expected and actual emotion when the patient should have been experiencing a “warm fuzzy” feeling is the cause of the rationalization or delusion, and supports the Perceptual Deficit/Explanation Account. Ramachandran also considered that Cotard’s syndrome, where a person believes they are dead, is an exaggerated form of Capgras, where all affect is diminished or absent.
This finding is undermined by the argument that patients with similar perceptually anomalous experiences do not necessarily form Capgras syndrome; Coltheart et al. (2000, as cited in Fine, Craigie, and Gold, 2005) reported that patients with bilateral ventromedial frontal damage (an area thought to be important for integrating somatic state information) also failed to show increased autonomic responsiveness to familiar faces, though these patients did not develop delusional beliefs. Therefore, a lack of autonomic responses is itself insufficient to produce the delusion; there must be other determining factors (Ellis and Lewis, 2001).
Fine et al. (2005) also explore the Expression Account, where delusional beliefs are expressions of the content of a particular disordered experience. In other words, the disordered experience is directly that of replacement by an impostor, so the account eliminates the problem of why the impostor belief is invariably constructed. Dalgarrondo, Fijuisawa, and Banzato (2002) performed a case study on a 26-year-old blind woman with Capgras syndrome based on the sensory modalities of touch, smell, and taste. Neurological examinations, MRI, and SPECT were all normal. This case reinforces the argument that facial recognition impairment is an insufficient explanation for Capgras delusions, suggesting that psychological factors such as ambivalence and fear of losing someone may play a role. The authors believe that an anomalous perceptual process is the result of a delusional transformation of the world, rather than the cause. This is consistent with the Expression Account presented above; however, such an account must then explain how such a perceptual belief arises.
A case study by Dietl et al. (2003) presents another anomaly to the current hypotheses of Capgras delusions. A 59-year-old woman with paranoid psychosis developed Capgras regarding her daughter and a close friend. A few days on antipsychotic treatment with haloperidol resolved it concerning her daughter but the delusion regarding her friend persisted. A year later, she began experiencing delusions of plans for self-substitution, but did not expect this person to be physically similar to herself. Her delusions about her daughter returned and she eventually reached the conclusion that this girl may not be her real daughter but was a pleasant person nonetheless, so did not confront her with suspicions any longer. Several inconsistencies with current hypotheses include the fact that the woman did not expect that her substitute would physically resemble her, and that a delusion regarding her daughter’s substitute occurred in that person’s absence. Therefore, this patient was not simply experiencing a face processing problem. Furthermore, the patient did not develop Capgras delusions for her husband, but rather reported that his facial features were altered; this suggests that additional cognitive processes were malfunctioning.
One insight that this case study presents is the idea that Capgras syndrome can be maintained without current visual input, which suggests that a reduced feeling of familiarity can become more general and independent; this is consistent with reports of blind patients with Capgras. It is therefore illustrated that additional factors must be integrated into the current model of Capgras etiology. Caution must be used when interpreting these results, however, due to the fact that the patient suffered from a paranoid disorder.
There are no conclusive treatments currently available for Capgras syndrome. Sometimes remission occurs naturally and sometimes the disorder persists; in many cases antipsychotic medication or neuroleptic therapy is provided. When remission does occur, it is frequently due to the removal or modification of a concomitant neurological condition; MacCallum (1973, as cited in Doran, 1990) hypothesized that organic conditions cause a change in a patient’s affects and perceptions, thus presenting Capgras syndrome.
In conclusion, no existing account or theory of Capgras syndrome is sufficiently explanatory of the phenomenon. The most credible and supported position to date is summarized neatly by Stone and Young (1997):
People voicing [the Capgras] delusion suffer an impairment that leads to faces being perceived as drained of their normal affective significance, and an additional reasoning bias that leads them to put greater weight on forming beliefs that are observationally adequate rather than beliefs that are a conservative extension of their existing stock.
Scientific evaluation of the processes in Capgras syndrome has many implications for our understanding of normal object- and face-recognition processes and the nature of beliefs, teaching us important principles about brain function. For instance, the fact that objects can be duplicated shows that items with which we are familiar or are particularly fond of may enjoy special cognitive status; also, we can deem our perceptions to be governed by comparisons between states of an object rather than absolute values (Ellis and Lewis, 2001).
The fact that Capgras patients do not recognize a loved one due to their impairments suggests that affective responses could be a key part of the ability to recognize a face; perhaps “affect and cognition are under the control of separate and partially independent systems” (Zajonc, 1980, as cited in Stone and Young, 1997). The fact that scientific studies have supported the dissociation hypothesis makes it no longer possible to interpret face recognition as proceeding in a strictly sequential fashion, but rather as a feeding process into an integrative device. Future studies should look into this possible attribution process, as this knowledge is critical for further understanding of Capgras syndrome.
Dalgalarrondo, P., Fujisawa, G., & Banzato, C.E.M. (2002). Capgras syndrome and blindness: Against the prosopagnosia hypothesis. Canadian Journal of Psychiatry, 47(4), 387-388.
Dietl, T., Herr, A., Brunner, H., & Friess, E. (2003). Capgras syndrome – out of sight, out of mind? Acta Psychiatry Scandinavica, 108, 460-463.
Doran, J.M. (1990). The Capgras syndrome: Neurological/neuropsychological perspectives. Neuropsychology, 4, 29-42.
Ellis, H.D., & Lewis, M.B. (2001). Capgras delusion: A window on face recognition. TRENDS in Cognitive Sciences, 5(4), 149-156.
Fine, C., Craigie, J., & Gold, I. (2005). Damned if you do; Damned if you don’t: The impasse in cognitive accounts of the Capgras delusion. PPP, 12(2), 143-151.
Papageorgiou, C., Ventouras, E., Lykouras, L., Uzunoglu, N., & Christodoulou, G.N. (2003). Psychophysiological evidence for altered information processing in delusional misidentification syndromes. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 27, 365-372.
Ramachandan, V.S., & BlakeBlakeslee, S. (1998). Phantoms in the Brain. New York: HarperCollins.
Stone, T., & Young, A.W. (1997). Delusions and brain injury: The philosophy and psychology of belief. Mind & Language, 12(3/4), 327-364.
The following is an observation lab I completed with a partner for third year Research in my Psychology B.A. Personal information and details have been excluded; if you require this information for referential purposes, please contact me for use.
Extraversion and Gender Differences in College Males and Females
Extraversion affects the public’s perception of a person and their wages, social status and level of personal fulfillment. If levels of extraversion differ among males and females, can it not be said that gender has some greater bearing on one’s personal life than being simply their sex? That is, men or women may be treated differently in certain social circles due to their extraverted or introverted tendencies.
Recent research by Nyhus and Pons (2005) states that wages and types of employment can be directly affected by the Big Five personality types. This includes extraversion, which they define as a preference for human contact and awareness, along with the want to encourage others (Nyhus & Pons, 2005). Their study discusses interesting information not pertinent to their topic; for instance, there was found to be a strong gender difference with test subjects of the age of 10 when specific traits were compared to their wages at age 26 (Feinstein, 2000, as cited in Nyhus & Pons, 2005). It has also been documented that men in the workplace are punished for withdrawal, whereas women are punished for aggression (Osborne, 2000, as cited in Nyhus & Pons, 2005). These two traits are closely related to extraversion.
Further research by Rubinstein (2005) suggests there is no significant difference between males and females with regards to extraversion. However, some of the behaviours measured were associated more strongly with men than women and vice versa (Rubinstein, 2005). This suggests that there is either no relationship between gender and extraversion, or that different methods of research are required.
In the present study, natural observation will be used on a frequency-event scale. Trace behaviours were measured due to the difficulty in finding strong behaviours associated with personality (Salgado, 1997, as cited in Rubinstein, 2005). The hypothesis proposed is that males will score higher on a scale of trace extraversion than females.
Disconfirming this hypothesis will require that there are no differences observed, or that females will score higher in extraversion. Extraversion is defined in terms of scoring two or more on an event sampling scale that measures three behaviours: smiling, gesticulation (“hand-talkers”), and clarity of speech. A smile is defined as upper molar (teeth showing). Gesticulation includes leaning toward the assistant (demonstrating outgoing body language), using visual representation to clarify needs (i.e. pointing), and hand-talking. Clarity of speech is determined on the basis of whether the observers are able to hear the library patron from their position (see Appendix B for photograph).
Observed participants were 40 patrons of the _____ _______ Library (21 males and 19 females) who approached the Access Desk, where most questions are directed and check-outs are processed. Most of these patrons ranged in age from approximately 18-24 and were students of ______ University, though a few adults (estimated in their mid-40’s) were included in the observations. Results cannot be generalized to certain groups that were excluded from observations, including: friends or acquaintances of the observers, people who appeared to know the desk assistant personally, people who had flamboyant style (such as bright pink hair), or people who had a friend present. The purpose of this was to control for extraneous variables to our data; specifically, experimenter bias (preconceived notions or formed judgments), friendliness due to familiarity, and peer pressure that might create changes in interaction.
The data was collected by means of event sampling and frequency coding on the basis of nominal categories (extraverted or not). Coding sheets were developed that recorded gender, the three events (behaviours) being observed, whether the patron qualified as ‘extraverted’, and comments, if necessary (see Appendix C). Inter-rater reliability was calculated using percent agreement and Cohen’s Kappa (calculations can be found on Appendix D).
The observers were positioned in the waiting area at the entrance of _______ Library, with a direct view of the Access Desk. All observations were recorded over two days, tracking patrons served by one particular assistant. This controlled for differences in observed behaviour due to potential differing interaction styles of library employees. It also ensured that observations were recorded in the same order for reliability purposes. The coding sheet used can be found in Appendix C.
All calculations were completed by hand and Figures 1 through 4 were drawn with Microsoft Excel (see Appendix A).
The results indicate that, according to the created measure, 40% of the overall sample was extraverted, thus the mode of the sample itself was ‘not extraverted’. This is illustrated in Figure 1. Out of the total amount of extraverted samples that were taken, 56% were female, with the remaining 44% being male. This is illustrated in Figure 2. The sample proportion of female (pf=.48) and male (pm=.33) extraverts further points out the differences in each respective group. This is illustrated in Figure 3. Through gathering information on the coding sheet, observed differences in the amount of time females and males “upper molar smiled,” “spoke clearly,” and “gesticulated” were recorded. Females smiled and spoke more clearly while males were observed to gesticulate more often. These differences are demonstrated in Figure 4.
The present experiment observed and compared the prevalence of extraversion among males and females. It was hypothesized that men would demonstrate more of the extraverted behaviours, as measured by a frequency event-sampling scale, than women.
Previous research on the relationship between gender and extraversion has generally been inconclusive with insignificant results. Rubinstein (2005) found that, while there were no significant differences, certain behaviours scored differently between the genders. With similar results, the present study’s hypothesis was not supported.
An interesting finding that we could expand upon was how many more females smiled than males. This suggests some differing characteristics that full under the label of extraversion. To a lesser degree, females tended to speak more clearly than their male counterparts. Males, on the other hand, gesticulated more often then females, though the difference is negligible.
Some of the reasons why our sample proportion of extraverts was so low (pE=.40) could be attributed to the need-based atmosphere of the library, as opposed to a more social environment such as the ____, ______ University’s on-campus bar. Students with a higher need for academic success might spend less time engaging in social, extraverted activities and more time at the library, therefore our sample may not be representational of the overall student population.
Nyhus, E. K., & Pons, E. (2005). The effects of personality on earnings. Journal of Economic Psychology, 26(3), 363-384.
Rubinstein, G. (2005). The big five among male and female students of different faculties. Personality & Individual Differences, 38(7), 1495-1503.
The following is a paper I wrote for my Sexuality class during the completion of my Psych B.A. If you intend to reference it, please contact me for documentation specifics.
Feminists Against Feminists: The Pornography Debate
There is no doubt that feminists, a potent subgroup of the population, are often involved in social and legal controversy. When someone thinks of feminism, certain words typically come to mind: angry, activist, equal rights, and man-hater, for instance. Though all feminists certainly have the common goal of liberating women, they can differ greatly on the visions they hold as necessarily leading toward that goal; no ongoing debate makes this fact more obvious than the sex wars. This essay will focus on the feminist-versus-feminist battle over pornography, from its very purpose and existence to its potential in the liberation (or not) of women’s sex and sexuality.
For the sake of this paper, pornography will be distinguished from erotica, in which males and females are portrayed as equals in sexually arousing material; pornography, on the other hand, is often violent, degrading or dehumanizing. Both men and women are consumers of pornography; however, the industry is undeniably male-oriented, and the most common users of pornography have been found to be traditionally gender-typed, masculine males and androgynous females (Hyde, DeLamater and Byers, 2006). Despite the fact that controversial issues constantly surround it, pornography is a booming industry, bringing in between $4 and $10 billion per year in the United States (Hyde, DeLamater and Byers, 2006).
The pornography debate has been raging for over forty years and still remains a focal point of discussion. Feminists have had a strong influence on academic interest in sexuality since the 1970’s, when Catharine MacKinnon and Andrea Dworkin proposed an ordinance to outlaw objectionable pornographic images (those which could be proven as causing harm to an individual, whether male or female). This resulted in the first large divide among feminists, when feminists against the anti-porn movement (representing the anti-anti-porn side), such as Ellen Willis (1983), argued that pornography could be experienced by women as sexually liberating (as cited in Chancer, 2000). The two basic positions being argued to this day are dichotomous, emphasizing either the sexual colonization and victimization of women (anti-porn, sexual-subordination, or “victim feminists” side) or the sexual repression and passivity of women (anti-anti-porn, sexual-pleasure, or “power feminists” side) (Russo, 1987; Chancer, 2000). The resolution of this debate is of importance beyond the direct goals of feminism; not only has the recurrent divide unintentionally weakened the movement from within (Chancer, 2000), but it threatens to undermine the progress of other subgroups which have struggled for acceptance in society by denouncing their freedom of speech where sexually explicit material is concerned (gay and lesbian communities, women).
Schaeffer (2001) explored the works of MacKinnon, a feminist intent on changing the social conditions that prevent women from being independent individuals. According to her, a stereotype becomes what is real because women are damaged by it. She defines a true feminist as one who sees the male point of view as fundamental to the male power to create the world in its own image, rather than viewing sexism as a myth to be corrected. Sexism is a conditioned and socialized reality. Among other prominent issues at the time, pornography was identified as a crucial feminist issue because it harms women in three ways: in the making of it, in the perpetuation of the objectification and dehumanization of women, and in the conditioning of users to experience sexual thrills from the degradation of others, thus shaping sexual behaviour. MacKinnon said that, “to the extent pornography succeeds in constructing social reality, it becomes invisible as harm” (1992, as cited in Schaeffer, 2001). This debate is necessary in order to raise questions about the harm done to women and how it could be rectified without the conflict extending beyond the individual to the state.
Hyde, DeLamater and Byers (2006) describe the four basic reasons why some feminists object to pornography. First, pornography debases women and often portrays them as subordinates to dominant males. Second, pornography associates sex with violence, thus contributing to social attitudes that reflect desensitization to violence toward women. Third, pornography shows unequal power relationships between men and women. Fourth, the structure of the pornography industry is such that a significant number of workers within seem to be physically, sexually, and emotionally abused. In these ways, pornography is a likely perpetrator of traditional gender roles. The images presented in contemporary pornography reflect the desires and fantasies of a male-dominated society. Double standards of sexual morality condemn females for certain sexual activities that are socially permitted for men. In a review of men behind the media, Watkins and Emerson (2000) showed that screenwriting is dominated by males, which leads to the proposition that male direction of the camera is so influential that women are socialized into identifying and complying with the very patriarchal values that marginalize them.
Both feminist positions in this debate define sexuality as being socially constructed; the difference is in how they conceptualize sexual roles and practices in relation to power, and how they approach the interconnections between sexuality, sexual violence, male domination/female subordination and the pornographic industry (Russo, 1987). Anti-pornography feminists make connections between pornography and the negative treatment of women; their opposition agrees that pornography is sexist and male-dominated, but focus instead on the expression of women’s sexual impulses and desires, viewing pornography as a vehicle to achieving control over their own sexuality. In some cases, this is shown to be true; Shellrude (2001) reported on the writings of prostitutes who reported that their experience had taught them to have confidence in themselves, their bodies, and their sexuality. Furthermore, these sex workers advised readers to look beyond the demeaning objectification of sex to take charge of their own sexuality. While the anti-porn feminists agree that self-representation is a key part of liberating women from male sexuality, they believe that the intertwining of fantasy and reality in pornography only serve as a medium to condone violence and degradation of women (Russo, 1987).
Anti-pornography feminists tend to represent traditionally feminist attitudes, and characterize their critics as male-identified and anti-feminist. They are uncomfortable with pornography because they identify its function as the perpetuation of male control. It is not the act of sex itself that is discouraged by proponents of this group, but rather the acceptance by women to be victimized by sex in such a way as is demonstrated in pornography. One anti-pornography feminist was quoted as saying that “if you do not agree that pornography is wrong, then you are not a feminist” (as cited in Russo, 1987). They see women who are proponents of porn as perpetuating the view that females are sexual objects by nature, and as helping to reproduce gender inequality.
The main issues for anti-anti-porn feminists are first amendment freedom and the right to sexual pleasure. The function of pornography is to introduce women to, and allow them to enjoy, different ideologies of sex that undermine the traditional values which had previously constrained them, such as purely procreative sex. The basis of their disagreement with anti-porn feminists is that identifying pornography as the enemy will cause shame and guilt among women who had otherwise come to terms with their sexual feelings, repressing women under the guise of feminism itself. This would only perpetrate the victim stereotype that women are innocent virgins who possess less sexual aggression than men (Russo, 1987). Feminists who do not believe in the censorship of pornography believe that if traditional restrictions about sex are removed then women can truly be sexually emancipated (Shellrude, 2001); in this perspective, individual sexual defiance is valued just as much as critiques of sexist institutions.
Baumeister and Twenge (2002) performed a thorough literature review to examine and attempt to understand the origins of suppressed female sexuality. It was assumed that socializing influences such as parents, schools, peer groups and legal forces cooperated to alienate women from their sexuality. Two competing hypotheses were explored: either men (male control theory) or women (female control theory) act as the main source of suppression on female sexuality. Other factors were considered, such as the greater cost of sexual mistakes for women (pregnancy), and the idea that women naturally have a lesser sex drive. Before the findings are discussed, these theories will be explored in more depth.
Male control theory is supported by the fact that men have held superior political and social power throughout most of history, therefore they could benefit from the suppression of female power in order to maintain their current positions. However, there are other explanations to this theory. Evolutionary theory, for instance, says that men control women because they require certainty about paternity; stifling their mate’s sexual desire is a small cost for ensuring that they are the legitimate father of their child. Feminist theory states that patriarchal social arrangements reflect the view that women are possessions that must be managed to prevent social chaos, because in fact women have a stronger natural sexual desire and are “insatiable.”
Female control theory is less instinctive because of the inferior positions females have held in the past; however, social exchange theory sets a promising stage to explain the reason that women would attempt to control the sexuality of other women. According to this theory, men desire sex, and so sex becomes a bargaining chip for which women compete to attain desired resources. Women therefore punish other women who make sex too easily available because this lowers every woman’s value (the author uses an analogy of the man who does not buy a cow because he already gets free milk). Pornography, as an outlet for a portion of a male’s desire for sex, threatens a woman’s negotiating power, so she feels the need to stifle its influence.
Female control theory predicts that women would be particularly opposed to pornography; evidence for this is long-standing, and contemporary results continue to demonstrate a fair amount of female opposition to alternative sexual gratification for men. In fact, all empirical evidence in this review repeatedly favoured the female control theory; mothers and female peers were found to be the main sources that teach adolescent girls to refrain from sexual activity, and women tended to support the double standard more than men. In other words, women were the main supporters of a moral system that condemns acts by women more severely than identical acts by men. Male control theory was overwhelmingly contradicted, to the point where any male influence found usually pushed in the opposite direction of predictions (i.e. a boyfriend who pushed toward more sexual activity). What these findings mean for women is that jealousy and competition only results in the suppression of their peers; this could be a valuable lesson for feminists who compete against other feminists, thereby distancing themselves from their goals. The positive side is that Western women have vastly reduced the gender gap, hence their reliance on sex to yield an exchange is lower than in the past, and may decrease favourably in the future.
The passion involved in this debate indicates the ever-important role of sex and sexuality in our society. Because of the divisions regarding views on pornography, legal restrictions on sexually explicit material are not the solution. The first step is to outlaw general statements regarding the experiences of women, and start with the experiences of individuals themselves. Shellrude (2001) argues that if people stopped denying that sexuality is a key part of getting what they want, then the stigmatization associated with sex could be reduced. Chancer (2000) suggests that feminists should consider why men are still largely in control of pornography’s profits and production. Indeed, the development of a feminist erotica has not only been proposed but has been put into action as female-oriented pornographic videos (directed by women for women) attempt to equalize power relationships between men and women. Hyde, DeLamater and Byers (2006) report on a study in which it was found that both male and female university students responded positively to arousing videos designed for women, while only the males responded in this way to videos intended for men. This is an encouraging step forward for the maturation of female sexuality, and may very well succeed in more female sexual freedom and less perceived victimization.
If anything has been demonstrated by this discussion, it is that the division between feminists on this issue is only serving to distract their energy from true progress, resulting in the ultimate weakness of the social movement. Feminists must focus instead on their commonalities, and achieve a middle-ground point of view. In other words, they must target sexism in the pornography industry itself rather than arguing over whether individuals are proponents of it; this debate has led to the illusion of a “false dichotomy,” as proposed by Chancer (2000). Struggle over sexual oppression and repression is only weighed down by the baggage of labels, when feminists could use their shared interests and mutual respect for individual diversities to produce social change. As concluded by Shellrude (2001), the best forward-thinking feminism is one in which a woman’s sex, sexuality, and libido work together with her intellect to transform the world.
Baumeister, R.F., & Twenge, J.M. (2002). Cultural suppression of female sexuality. Review of General Psychology, 6(2), 166-203.
Chancer, L.S. (2000). From pornography to sadomasochism: Reconciling feminist differences. Annals of the American Academy of Political and Social Science, 571, 77-88.
Hyde, J.S., DeLamater, J.D., & Byers, E.S. (2006). Understanding Human Sexuality. (3rd Can. Ed.). Toronto: McGraw-Hill Ryerson.
Russo, A. (1987). Conflicts and contradictions among feminists over issues of pornography and sexual freedom. Women’s Studies International Forum, 10(2), 103-112.
Schaeffer, D. (2001). Feminism and liberalism reconsidered: The case of Catharine MacKinnon. The American Political Science Review, 95(3), 699-708.
Shellrude, K. (2001). Coming between the lines: A fresh look at the writings of anti-porn and whore feminists. Canadian Woman Studies, 20/21(4/1), 41-45.
Watkins, S.C., & Emerson, R.A. (2000). Feminist media criticism and feminist media practices. Annals of the American Academy of Political and Social Science, 571, 151-166.
The following is a study my lab partner and I ran in our 3rd year Psychology Research class. Please contact me if you have any questions or want further information; I am excluding some personal details from our report.
How Perceived Gender Affects the Performance and Attitude of Male Online Gamers
The current study examined whether male video game players show more aggression toward a perceived male or female opponent. Undergraduate students (N = 8) volunteered to participate in a study using the popular game Counter-Strike™. Participants played two one-on-one matches against unseen opponents. The opponent’s aliases implied that one opponent was male and one was female, but both were actually the same male confederate. Participants filled out pre- and post-match questionnaires, recording mood and perceived skill levels of both themselves and their opponents. This data was combined with screen shots of game scoring (with aggression measured in number of kills), and summarized in a table. Our hypothesis predicted that perceived male-on-female matches would have lower ‘kill scores’ than perceived male-on-male matches, and this was not supported; however, a 2×2 factorial design showed a pattern that may have supported an interaction between perceived gender of opponent and gender order with a greater sample size.
How Perceived Gender Affects the Performance and Attitude of Male Online Gamers
As more children and adolescents overlook passive activities like watching television for more active participant roles in cognitively-challenging video games, one may be concerned about how these increasingly realistic games are socializing the players. After all, an increased potential for hostile or aggressive behavior following arousing violent video game play has been suggested and evidenced (Ballard & Wiest, 1996, as cited in Ballard & Lineberger, 1999). As well, it became impossible to turn a blind eye on this issue following an exceptionally memorable event: Eric Harris and Dylan Klebold launched an attack on classmates and teachers at Columbine High School on April 20, 1999, murdering 13, wounding another 23, and ultimately committing suicide. Following the horrific attack, the media and its shocked audience needed something to blame, and when it was discovered that the boys were avid players of a violent video game called Doom™ (licensed by the U.S. military to train soldiers to effectively kill), many scientific and psychological questions were raised.
In order to examine the effects of video game play, Mazur, Susman and Edelbrock (1997) conducted a correlational study on video game users, measuring cortisol and testosterone responses. Fifty-two volunteers (32 females, 28 males) with a mean age of 20 years played against same-sex partners in video ping-pong. Participants provided saliva samples before and after matches. Additionally, they were asked to describe their feelings and evaluate their own and their opponents’ performances after each match. Mazur et al. found an anticipatory testosterone rise in male participants before the match, and this testosterone increase is thought to fuel aggressiveness. Female testosterone levels generally decreased throughout the experiment. Considering the total lack of violence in video ping-pong, it is interesting that gender effects still played a significant role.
Brown, Hall, Holtzer, Brown and Brown (1997) wanted to evaluate the actual performance of both male and female gamers using the same game as Mazur et al. (1997) (pong), but removed the factor of interpersonal competition and had participants play against the machine. In three separate experiments, equal numbers of males and females participated, and it was found each time that, regardless of previous video game experience, men performed better than women. Both men and women profited from experience, however, which was demonstrated by fairly equal levels of improvement. One interesting finding resulted from the third experiment, where it was hypothesized that the presence of an audience would facilitate pong performance for men but not for women. In fact, both genders showed significantly poorer performance in the presence of a female audience.
Following these findings, Ballard and Lineberger (1999) wanted to examine the effect of confederate gender on aggressive behavior following video game play. One hundred and nineteen males played 15 minutes of either a nonviolent game (NBA Jam™) or one of three levels of a violent game (Mortal Kombat™). They were then placed in a teacher/learner paradigm with either a male or female confederate. They found that participants rewarded male (but not female) confederates significantly more under the nonviolent condition, and participants under the violent condition punished confederates more stringently as game violence increased, especially if the confederate was female. This is important in the sense that it could mirror real-life events of violence against women. Films that are violent, but do not have any sexual content, were found to produce significantly more male aggression against females than do neutral or erotic films (Donnerstein, 1983, as cited in Ballard & Lineberger, 1999). As mentioned above, video games are an even more active medium of participation than watching television or a film; it is possible that males who frequently play violent video games would integrate negative stereotypes of women and play them out in actual society.
Anderson and Dill (2000) conducted two studies, one correlation and one experimental, examining the effects of violent video games on aggression-related variables. Their correlation questionnaires measured amount of exposure to video game violence and amount of time games had been played by college students prior to the experiment. Life-event information was also obtained. It was found that time spent playing video games in general was positively related to delinquent behaviours, and the effect was stronger with violent video games. Furthermore, video game violence was negatively, but not significantly, related to GPA, while time spent playing games in general was significantly and negatively correlated with GPA. These results would clearly raise concerns for parents whose children are frequently involved in video game play, if this indeed had a significant effect on one’s life events.
Anderson and Dill’s (2000) experimental study examined laboratory exposure to a graphically violent video game to test whether it increased aggressive thoughts and behaviour. A large sample size of college students (N = 210, 104 females, 106 males) became familiar with either a blatantly violent video game (Wolfenstein 3D™, where the goal is to eventually kill Adolph Hitler), or a nonviolent video game (something comparable to Tetris). Both were deemed engaging games in pilot studies, which is important to the internal validity of the experiment; some weaknesses in previous research studies examining video game play were caused by lack of control for effects due to level of arousal as opposed to exposure to violence. In a reaction time competition following the game play, participants who had played the violent video game set more intense punishments for opponents who lost a trial. There was also a main effect of gender, where aggression scores for men were M = 3.45 and M = -4.60 for women. Playing the violent video game seemed to have primed aggressive thoughts, but it was unclear whether these applied more strongly to female opponents.
Bartholow and Anderson (2002) also used the reaction time task, but organized it such that the confederate (opponent) was always female. This study engaged undergraduate students who were not habitual game players but were familiar with both games involved (Mortal Kombat™ as the violent game, and PGA Tournament Golf™ as the nonviolent one). Participants played one of the two games for a 10-minute period before competing with a confederate in a competition which allowed for provocation and retaliation. Punishment levels set for an opponent without fear of retaliation served as the measure of aggression.
What Bartholow and Anderson (2002) found was that participants who played the violent game later showed more aggression toward the confederate than did participants who played the nonviolent game, with a larger effect for men than for women. This supported their hypothesis, but conflicting results raised a new question: did women display a different, more subtle style of aggression than males? Though not the topic of focus, it was noted that women set longer (but less intense) sound blasts as punishments than their male counterparts, who tended to set louder, shorter blasts of sound.
Given the ambiguity seen in past experiments regarding the effect of one’s opponent’s gender, the present study examines how male video game players react to an opponent when they are perceived to be male or female. Since both the Mazur et al. (1997) and Brown et al. (1997) studies focused on same-sex and computerized-opponent matches, perhaps the present study could indicate a decrease in male performance for male-on-female matches. Actual performance level is controlled by having each participant play against the same male confederate; in this sense, any variation among the ratings of the opponent’s skill level can be assumed to be based on preconceived gender notions in relation to video games, rather than the actual skill of the confederate. Prior experience playing the game is somewhat controlled because volunteers will be recruited based on their prior experience. We hypothesize that males will score lower in the perceived male-on-female condition than in the perceived male-on-male condition. This hypothesis will not be supported if there is no difference in scores between the groups, or if male-on-female scores are higher than male-on-male scores.
Participants were eight male undergraduate students at ______ University with an age range of 18-24 years. They responded to ads citing a study involving “Counter-Strike™ 1.6 LAN” and volunteered to take part in two one-on-one matches using the online computer game and their first names as aliases. No compensation was offered for participation.
All participants brought their laptops, power adaptors, LAN cords, mouse (optional), and a personal copy of the online multiplayer game “Counter-Strike™ 1.6 LAN” (CS). CS was created by a developer named Valve Software and is published through an online distribution system called Steam. This multiplayer game pits two teams against each other: the terrorists and counter-terrorists (CTs). The terrorists are the ‘bad guys’ and the counter-terrorists are the ‘good guys’ who complete objectives, sometimes within set time frames. Players can purchase weapons and equipment that are realistically modeled after their real-world counterparts. Teamwork is typical; after each round, team members are allocated money based on both an individual and team performance.
Consent forms were developed for participants to read and sign acknowledging the terms and conditions of this study prior to survey administration (see Appendix A). Following completion of the full study, all participants were collectively debriefed and given debriefing forms (see Appendix B), discussing the true study questions and giving an assurance of confidentiality, as well as the opportunity to ask research-related questions. One questionnaire was administered before the matches began in order to collect general data (see Appendix C); additionally, two questionnaires were administered at the end of each match, presenting questions related to performance and post-match mood (see Appendix D).
Walkie-talkies were used to communicate between the researchers. SPSS 14 was used to analyze the data (see Appendix E).
The researchers set up two rooms in _____ _______ Library at opposite ends of the building. Participants were each allotted a half-hour time slot, and were greeted by one of the researchers at the library entrance. They were informed that the other researcher was waiting in a different room with their opponent (who was supposedly matched to their experience level), but that their opponent had to leave early so they would actually get to play two different people in two separate 10-minute matches. They were told to not text chat with the other player, so as to not affect their performance or reveal information concerning their identity. They were also instructed to use their first name as their game alias, the assumption being that their opponent was doing the same. The experimenters then gave participants a consent form to sign and explained that they would also fill out a number of questionnaires throughout the study, in order to measure their emotions before and after the games.
Once participants were informed, they filled out a pre-match questionnaire in the form of a 7-point Likert scale that asked general questions such as how lucky they considered themselves to be, whether they made tactical game decisions, whether they were experienced in Counter-Strike™, and what their mood was at the moment. Additionally, two questionnaires were administered after every match, and included similar questions referring to their performance and satisfaction with the most recent game, as well as the performance and perceived skill of their opponent. Participants played one match against a perceived female opponent, and one against a perceived male opponent. Opponent gender order was a counterbalanced variable.
Matches were timed by the observing experimenter, and walkie-talkies were used to communicate the beginning, one-minute mark, and end of matches (to cue the engaged confederate to stop playing). Screen shots were taken of the final scores by the confederate, who also noted kill scores and ‘opponent gender’ order on a separate sheet.
When the final questionnaires were filled out, participants received an explanation of why they could not yet be fully debriefed, and given the opportunity to ask questions. They were also asked to keep their experience personal until the debriefing. A meeting was scheduled shortly after the last participant had completed the study, and participants were debriefed collectively. Participants already had a history together (weekly game nights) and had recommended the study to one another, so there was no ethical issue with having their participation revealed to one another. At the collective meeting, the deception concerning the experiment’s true purpose was revealed, and analysis results were explained to the group. A question-and-answer period was offered. Those who could not make it to the meeting were contacted personally by the experimenters and offered a full explanation of previous findings, deceptions used, the study’s actual hypothesis, and the results.
A 2×2 factorial table (see Table) appeared to show a small main effect for opponent gender, with male-on-male games producing two more “kills” in total than male-on-female games (57 and 55, respectively). There seemed to be a larger main effect for game order, which represented a certain amount of learning or practice: in Game 1, 49 kills were scored, and in Game 2, 63 kills were scored. There also seemed to be an interaction found between game order and gender: when participants thought they were playing a male in the Game 2, their score improved drastically (from 21 in Game 1 to 36). However, regardless of game order, ‘kill scores’ against female opponents hardly changed across games (28 and 27) although the score from Game 1 was higher than that of male-on-male scoring. A 2×2 Repeated Measures ANOVA was performed on the data using SPSS (see Appendix E). No main effects or interactions were found. Perceived gender of opponent was insignificant at F(1,6) = 0.045, p > 0.1. Gender order (the order in which a participant played a ‘male’ or ‘female’) was also insignificant at F(1,6) = 1.171, p > 0.1. Finally, the interaction between perceived gender of opponent and gender order was not found to be significant, at F(1,6) = 2.194, p > 0.1. These results are not consistent with the hypothesis that male-on-male scores would be larger than male-on-female scores.
The data from the questionnaires presented a few basic patterns (see Appendix F). Self-ratings on the pre-match questionnaire concerning general luck and tactical decisions were good predictors of self-ratings of those same factors on the post-match questionnaires. Most participants felt that their performance in the games accurately represented their usual level of play. All participants rated their opponents as less lucky than themselves; the majority rated their female opponents as less lucky than their male opponents. Male opponents and female opponents were generally given similar ratings regarding performance and apparent experience.
There has been a boom of research done on the relationship between violent video games and aggression in recent years, especially since the events at Columbine High School in 1999. Some studies have suggested that there is a fundamental difference between males and females when it comes to aggression in games, even when those games are not violent (Mazur et al., 1997). It has also been shown that, after being exposed to violent video games, males tend to be more aggressive toward females than toward males (Ballard et al., 1999). However, the present study explores the idea that males consider females to be less experienced and less capable in their world of online gaming.
Contrary to expectations, aggression, which was operationally defined as “tabulated kills,” did not drop significantly when participants played against what was perceived as a ‘female’ opponent, and so the experimental hypothesis was not supported based on the ANOVA test. However, the scores as described on the 2×2 factorial chart indicate that there are factors affecting the gamers’ performance when confronted with a perceived target gender. The general consensus is that males are more aggressive toward other males, but there were odd effects demonstrated in the present study. What is interesting is that ‘kill scores’ stayed the same against ‘female’ opponents regardless of game order. This could be due to a conscious or unconscious effort to go easy on the ‘female’ opponent. The effect could also be described in terms of shock value, whereas the male does not expect to encounter a female in his second match and re-evaluates his game style, performing once again at his first-game level. With a greater sample size, this pattern would have become more apparent and may have proven significant.
Some interesting patterns of behaviours and attitudes were noted within the questionnaires. The following may be explained by mere modesty, but participants almost always rated their opponents as less lucky than themselves. There was an interesting relationship demonstrated, which can be explained based on the following premise: assuming that performance is stable across matches (which, based on the confederate’s results, it is), it can be said that performance not attributed to luck must be attributed to skill. In the questionnaires, participants always rated their female opponents as less lucky than their male opponents. However, the anomaly is here: even though the kill scores against participants were virtually identical, and females were rated less lucky, they were not rated as more experienced or as having performed better! There is a discrepancy present – if the male and female were the same player in reality, and the male player was rated as more lucky, does this mean that, in the eyes of a male competitor, females who perform equal to males are proportionally more advanced (or “less lucky”) for their gender group? Perhaps more open-ended questions and a follow-up questionnaire could be generated to explore this topic further.
Another factor that could be explored in future studies is the gender of the researcher who observes the participant, an effect which might account for certain aspects of a participants’ performance. Brown et al. (1997) found that both males and females display significantly lower performance in the presence of a female audience. The present study only made use of a female observer. It is possible that participants did not perform at their usual level in the matches, although the questionnaire responses stated most often that they did, and the observer made sure to face away from their laptop screens. Another possibility was that, to avoid embarrassment, the participant may have gone easier on the ‘female’ participant so not to be judged as aggressive toward females by the observer. To account for this possible factor, a 2x2x2 factorial design could be used with a new “observer gender” independent variable to control for observer effects. To collect a more thorough data set without exploring audience gender, a third level under ‘gender’ could be explored, which would be an ambiguously-named opponent. It would be left to the participant to interpret the gender of this opponent as they wished, and it would be interesting to compare this category of results with the straightforward male-female distinction. Another option of interest is if females were included as participants in the study, if possible to recruit enough experienced players, in order to see if gender differences exist on that end as well. Only with further studies can these effects be analyzed.
Certain deceptions were necessary for this study, considering the lack of resources available to the experimenters. The questionnaires were not a true source of analysis, as participants were led to believe by their mere frequency; they were administered in the attempt to discover the presence of overlooked relationships. Secondly, emotions were not a topic of interest without their link to gender, but this description was as close as the experimenters could come to the actual hypothesis without creating bias or demand characteristics. There was a concern that if participants knew certain aspects of their behaviour were being studied, they might subconsciously alter those behaviours. For instance, if they believed the hypothesis was that males wouldn’t treat women as equal opponents, they might overcompensate by playing extra hard against the female players. If they thought the topic of study was aggression and gender, they might try to take a more defensive strategy against females. The biggest indirect deception was that participants were led to believe they were playing two different people of two different genders. The assumption was that since they were required to use their first name as an alias, they would expect the same honesty from their opponent. This factor was necessary to ensure consistency of their opponent’s performance across matches, to rule out the fact that participants might rate an opponent lower in performance because they were in fact a less skilled player.
Some of the above deceptions could have been remedied by developing a ‘cover story’ that did not lead to suspicions or compensations in behaviour, but that retained a fair degree of honesty with the participant. However, validity could have been retained, in the case where participants suspected the very fact that there were female players in the study, by having them meet a ‘female’ opponent before the match (also a confederate, of course, and not their true opponent). If time constraints were not an issue, a female could be trained in the game and be the one to play all the matches, while either a male or female ran the participants through in a different room. There are many possibilities, though all come with a different set of benefits and sacrifices.
Improvements to the general design of the study could have existed with better facilities. Better is defined as being more adaptable to the needs of the study; in the present study, where rooms were booked out from a library, the greatest rearrangement possible was moving the tables a certain way. Many inconveniences could have been avoided if two neighbor rooms were used to host the study, but the confederate’s room would have to be windowless to conceal his or her identity. A double blind study could have further ensured accurate results, but based on the technicalities and restrictions of Counter-Strike™, this process may have caused more complications to the process (ie. not allowing the confederate to see their own alias on the screen). Furthermore, it was expressed by several participants that Counter-Strike™ is most popularly used as a multi-player game, so perhaps a contemporary game could be used in a future study that is better adjusted for one-on-one match play.
Finally, the data from this experiment could have been explored more thoroughly with a greater sample size. Future studies would aim for 20 participants in a within-subjects design and upwards for between-subjects in order to achieve significance. This would involve a greater amount and longer period of advertising, and perhaps some direct benefit or reward for participation. For now, the results presented by this study remain a promising and contemporary examination of the effects of gender on motivation and aggression.
Anderson, C. A., & Dill, K. E. (2000). Video games and aggressive thoughts, feelings, and behavior in the laboratory and in life. Journal of Personality and Social Psychology, 78(4), 772-790.
Ballard, M. E., & Lineberger, R. (1999). Video game violence and confederate gender: Effects on reward and punishment given by college males. Sex Roles, 41, 541-558.
Bartholow, B. D. & Anderson, C. A. (2002). Effects of violent video games on aggressive behavior: Potential sex differences. Journal of Experimental Social Psychology, 38, 283-290.
Brown, M. R., Hall, L. R., Holtzer, R., Brown, S. L., & Brown, N. L. (1997). Gender and video game performance. Sex Roles, 36, 793-812.
Mazur, A., Susman, E. J., & Edelbrock, S. (1997). Sex differences in testosterone response to a video game contest. Evolution and Human Behavior, 18, 317-326.
Mice were subjected to stress-inducing conditions after being injected with an isolated incense compound, and it was found to soothe them in a manner similar to antidepressant and antianxiety drugs:
Further investigation revealed that the compound, named incensole acetate, can regulate the flow of ions in and out of neurons in a manner similar to the way antidepressant drugs work. This incense chemical may be the key to a new class of antidepressants and shed light on the molecular workings of the brain and emotion.
A neuroscience experiment on rats showed the scaaary brain changes involved in S.A.D. (seasonal affective disorder), or depression associated with light deprivation:
The animals not only exhibited depressive behavior but also suffered damage in brain regions known to be underactive in humans during depression. The researchers observed neurons that produce norepinephrine, dopamine and serotonin—common neurotransmitters involved in emotion, pleasure and cognition—in the process of dying. This neuronal death, which was accompanied in some areas by compromised synaptic connections, may be the mechanism underlying the darkness-related blues of seasonal affective disorder.
Good thing I’m staying in a sunnier room now – but wouldn’t hurt to invest in more lamps :)