A lot of past research has been conducted to study the effects of pseudohomophones on people’s reaction time. The first people to conduct an experiment and provide evidence supporting the role of phonological mediation in visual word recognition were Rubenstein et al (1971). At the other end of the line, critics like Baron (1973) argued against the focus on phonology alone during reading and brought about visual mediation. The present study was conducted to measure participants’ reaction time for the pseudohomophones and compare the difference in results with nonwords. The findings indicate that the reaction time for the pseudohomophones was the longest, thus supporting the initial hypothesis, and past studies as well. Applications of the results could have various uses, and these are discussed keeping in mind certain possible limitations.
A pseudohomophone is a string of letters that looks and sounds like a real world, but is not really a word included in the English dictionary. For example, ‘crain’ is a pseudohomophone as it sounds like a real word ‘crane’ but isn’t really qualified to be a real word by itself. The pseudohomophone effect occurs when it takes longer time to distinguish between a real world and a pseudohomophone than the time taken to distinguish between a real word and a nonword, and this was seen in Rubenstein et al’s study (1971). They showed the participants a word that was either a real word, a nonword, or a pseudohomophone for two seconds and asked them to respond as to what they had seen (i.e. whether a real word or a nonword). The results of this study supported the pseudohomophone effect and Rubenstein (1971) suggested that this happens because individuals use a ‘phonological code’ while deciphering words. According to this code, the phenomes (the smallest units of sound that a word can be broken to) are processed to see if the word is phonologically correct i.e. for pronunciation. If the word presented is identical phonologically, then it is searched for a match visually. This means, the pseudohomophones pass through the phonological check but later get rejected at the visual check, whereas non-words hardy make it through the phonological check. This results in the pseudohomophones passing the brain’s phonological code and only getting rejected after being processed visually, therefore taking a person a longer time to respond when deciding if a pseudohomophone is a real word or not.
Martin (1982) suggested that the previous tests done by Rubenstein (1971) on pseudohomophones were flawed as they do not consider the fact that there are more than two categories of nonwords. He proposed another type of nonword, orthographically legal nonwords, which only look like real words but don’t sound like them. In his experiments, Martin (1982) showed that when orthographically legal nonwords were used in experiments, the pseudohomophone effect disappeared.
Baron (1973) emerged to be a very strong critic of Rubenstein et al. He proposed the visual coding model. He suggested that the findings of any experiment that used nonwords as stimuli were not a concise manifestation of the actual processes that take place while reading. Moreover, the presence of nonwords could have an effect on a person’s way of reading i.e. nonwords could have been likely to persuade the brain to rely on phonological cues. Baron developed the Phrase Evaluation Task to test his hypothesis which was based on the visual coding mode. The results of his study showed that participants rejected stimuli that were meaningless much faster, as compared to the other stimuli. This led him to dispute that participants mainly use visual information to make a decision, and will rely on phonology only after processing visual stimuli.
Coltheart et al (1977) proposed a different model called the dual-route model suggesting that both phonological and orthographic (visual) checks hold equal importance. This model proposes that humans also use visual cues and if these cues are familiar we may accept it as a real word without checking its sound. However, in situations where the word is not familiar, phonological checks are done much earlier in processing. Coltheart et al (1977) also proposed that after a word has been checked both phonologically and visually, if no recognition is made within a certain period of time, the brain will exclude the word out as a real one.
Taking into consideration the previous studies mentioned, this present study hypothesizes that there will be a significant difference between the conditions, with pseudohomophones taking longer to reject.
The design used for this experiment is independent groups design. It consists of two variables: the independent variable and the dependent variable. The independent variable was the type of non-word used. It consists of three levels: Condition A used pseudohomophones, Condition B used pronounceable non-words, and Condition C used unpronounceable non-words. The dependant variable was the median response time.
The participants of the study consisted of 64 second-year (male and female) psychology students of Middlesex University Dubai, undertaking the PSY2004 module.
The experiment was carried out in the psychology lab on the computers, with superlab software.
The participants were randomly allocated to three separate groups. They were all given 18 practice trials that were randomly presented, consisting of 9 real words, 3 pseudohomophones, 3 pronounceable non-words and 3 unpronounceable non-words. Each practice trial was timed and feedback on accuracy was provided during the practice trials.
For the experimental trials, the participants were provided with 50 stimuli which were divided into 25 real words, plus 25 of the three conditions i.e. either 25 pseudohomophones (Condition A), or 25 pronounceable non-words (Condition B), or 25 unpronounceable non-words (Condition C). The participants were asked to press the ‘z’ key on the keyboard every time they judged the stimulus to be real, and press the ‘/’ key every time they judged the stimulus to be fake. They were not given any feedback on accuracy during the experimental trials. The number of errors made by the participants and the median response time for the non-words were recorded.
The median response time for each participant were calculated for the correct responses only. Any participant with more than 20% errors has been excluded from the analysis. This is probably because an increased number of results that lie over a 20% error or higher will skew the distribution unfavorably.
The data of the 64 participants were collected and assessed using ANOVA, the median scores are reported in the tables below.
Descriptive Statistics of the groups:
A – pseudohomophone
The ANOVA showed that there is a significance between the medians of the three conditions as
F (2,61) =8.927, p<0.05
The findings of this study reveal that the reaction time was the shortest for Group C (unpronounceable nonwords) and the longest for Group A i.e. pseudohomophones. Between Group A and Group B, there was a very little difference in the results as compared to the massive difference between Group C and the other two conditions. As the reaction time was the longest for Group A, the fact that the participants took the longest time to process pseudohomophones as compared to the other two conditions. These results show significant differences between each of the conditions, and also show that pseudohomophones take the longest to reject, thereby supporting the hypothesis of the study.
These findings have resulted in a similar outcome to previous studies. Bringing back the study done by Rubenstein (1971): the phenomena of taking longer to identify a pseudohomophone as not a real world occurs because of the pseudohomophones passing through the brain and later getting rejected during visual processing. Another study was done by McCann et al (1988). They conducted replication studies using the lexical decision task. The results of this study revealed that pseudohomophones yielded longer reaction time than nonpseudohomophones.
With the support of a number of studies to maintain the pseudohomophone effect, there are arguments against it as well. For example, the argument that Baron proposed suggesting that visual mediation played the important role in word processing. Another study worth mentioning was well against the pseudohomophone effect. According to Underwood et al (1998), “when readers encountered homophones during a training phase of the experiment, then a pseudohomophone effect was observed in a later block of trials which contained no homophones during either phase of the experiment and they did not show a pseudohomophone effect”. According to his study, people used phonological evidence when it was necessary and did not use it when it was not needed.
This study could have a few limitations, the first being the factor of language. The experiment was conducted in the English and although all the participants were well-versed in the language, it was not the first language for a majority of the participants. This could have been a mere advantage for those who had English as their first language. The second limitation could be the fact that the participants weren’t divided equally into the three groups (Group C contained 2 participants lesser than the other two groups) and this might have a significant difference on the average results, although they might be relatively small.
Keeping these limitations in mind, it can be suggested that for further research, homogeneity of the participants (especially considering language) could be kept in mind. If improved, this study can be used in various ways; trying to understand reading/learning in young children, understanding and maybe even helping with dyslexia etcetera.
1. Herdman, C. M., LeFevre, J.-A., & Greenham, S. L. (1994). Implicating the lexicon: Base word frequency effects in pseudohomophone naming. Journal of Experimental Psychology: Human Perception and Performance, 20, 575-590.
2. McCann, R. S., & Besner, D. (1987). Reading pseudohomophones: Implications for models of pronunciation assembly and the locus of word-frequency effects in naming. Journal of Experimental Psychology: Human Perception and Performance, 13, 14–24.
3. McCusker, L. X., Hillinger, M. L., & Bias, R. G. (1981). Phonological recoding and reading. Psychological Bulletin, 89, 217-245.
4. Monsell, S., Patterson, K. E., Graham, A., Hughes, C. H., & Milroy, R. (1992). Lexical and sublexical translation of spelling to sound: Strategic anticipation of lexical status. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18, 452-467.
5. Seidenberg, M.S., Petersen, A., MacDonald M.C., Plaut, D.C.. (1996). Pseudohomophone Effects and Models of Word Recognition. Journal of Experimental Psychology: Learning, Memory, and Cognition. 22 (1), 48-62.
6. Underwood, G; Roberts, M; Thomason, H. (1988`). Strategical invariance in lexical access: The reappearance of the pseudohomophone effect.. Canadian Journal of Psychology. 42 (1), 24-34.
7. Van Orden, G. C., Johnston, J. C., & Hale, B. L. (1988). Word identification in reading proceeds from spelling to sound to meaning. ‘Journal of Experimental Psychology: Learning, Memory, and Cognition, 14, 371-386.
8. Ziegler, J. C., Jacobs, A. M., & Kl¨uppel, D. (2001). Pseudohomophone effects in lexical decision: Still a challenge for