- NG LI JIA
For several decades, researchers tried to understand the dynamics of short-term memory in speech processing and word recall. However, data were contradictory due to different lines of inquiry using varying methodologies. Understanding the underlying mechanisms of word recall and short-term memory remains an elusive, but not an impossible task. It is important to pinpoint how short-term memory and long-term memory interact so that light can be shed on illnesses which can affect speech articulation (Caplan, Michaud, & Hufford, 2013).
Current research focused on neighborhood size effect on the serial recall of words. Neighborhood size refers to the number of phonologically-similar words on a list that is serially recalled. Many experiments have shown that word recall was worse using lists with many phonologically similar words, i.e. words with a large neighborhood size.
Roodenrys (2009) concluded the efficacy of word recall was more affected by large neighborhood size and frequency due to phonological similarities. However, when Roodenrys replicated Goh & Pisoni’s (2003) experiments using words that were not neighbors on the same list, no effects were found for either large or small populations.
The following experiment utilize lists where neighbors are present in set but not on same lists. It will also utilize lists where neighbors are present in list but not present in set. Therefore, hypothesized that serial recall will be more significantly than small neighborhood sizes as opposed to large neighborhood sizes where words are present in list but not in same set. In addition, it is hypothesized that significant effects will be noted by small neighborhood sizes on serial word recall when no neighbors appear in the set. Examination of past experimental methodologies is in order to determine where errors were made and how to navigate the best approach to prove or disprove the aforementioned hypotheses.
Jalbert, Neath & Suprenant (2011) showed that word length and concurrent articulation negated neighborhood size effects in mixed lists. Concurrent articulation is defined as when a word that is to be recalled is spoken — or “articulated” — almost simultaneously, adding a so-called “cognitive load” as well as “noise to the to-be-remembered items” (Jalbert et al., 2011). This study demonstrated that small neighborhood size was affected very little by concurrent articulation.
The study’s design was marred by the fact that such small sample was used. That is, only ten subjects were tested — small number from which to collect meaningful data. Furthermore, it stands to reason that studies should use subjects who have similar educational backgrounds and/or similar ages, thereby eliminating confound. Older people may show quicker decay time when shown a word, thus slowing down process of redintegration.
However, recalling longer words from larger neighborhood sizes, higher-educated people would be more familiar with lists and/or sets of words of longer lengths. In this study, English was native language of the subjects. However, they failed to mention whether subjects had proficiency in second language, a skill which could help recalled CVC (consonant-verb-consonant) words, or even with longer nonsense words. Example, if a subject was shown the nonsense word “geto”, subject may use secondary language to recall similar word “gato” which, in Spanish, means “cat”, or homonyms such as paro (a nonsense word) and “pero”, meaning “but” in Spanish, assisting in a different mechanism of recall other than a phonological loop.
Goh and Pisoni (2008) used 56 subjects of roughly the same age and educational level in the first experiment of their study, thus potentially eliminating age and educational level variability and gathered more data from the subjects. Goh and Pisoni (2008) also considered short-term memory span (STM) and measured this variable accordingly, using the nine-digit span. The researchers found that all subjects scored about the same.
However, this study’s aim to measure lexical competition based on frequency and density of neighborhood sizes. They found that lexical competition among item-specific information for “easy” versus “hard” words in non-repeated lists was performed mainly in long-term memory (LTM) (Goh and Pisoni, 2008). The researchers assumed LTM and STM are static entities and they are not in flux. Moreover, their definitions of experimental effects fit STM versus LTM are ambiguous. The abovementioned researchers did not use nonsense words in their experiments, which would have served as a significant variable for determining boundaries of LTM and STM, because nonsense words are less frequent in large populations. Nonsense words would help eliminate lexical density in neighborhood populations and further define process of redintegration, i.e. how words decay into STM “traces” are somehow retrieved and reconstructed.
Roodenrys et al. (2002) conducted similar study using 24 Australian subjects, all of them native English speakers. Roodenrys et al.’s (2002) contradicted other findings that found speech perception in lexical properties playing a role in redintegration of words. Rather, these experiments pointed squarely to phonological processing rather than speech perception. While the design and methodology were sound, researchers failed to account for cultural. The results of all four experiments revealed a counterintuitive result, i.e. effects of neighborhood frequency had little impact on word recall. Roodenrys (2002) stressed that most of the word recall effects in their experiments were mediated by speech-production mechanisms.
Researchers — especially Roodenrys — contended lexical memory equates with LTM. However, their line of reasoning is unsound. It seems the time elapsed between seeing a word and its recall should constitute one definition of how STM of word lists and intrusion errors across sets could delineate the boundary between STM and LTM. Another consideration, is the effect of pattern recognition on recall accuracy as well as the potential effect on LTM. Furthermore, there may not be a universal “wiring” schematic for individual.
Finally, many of these questions will be investigated in this experiment and variables will be held constant to accurately calculate the effect of independent variable(s), and conclusive data will be collected to prove or disprove hypotheses that predict significant effects of small neighborhood sizes when no neighborhood words are on list but in a set and predicting significant effects of small neighborhood sizes (as opposed to large neighborhood sizes) when no neighborhood words appear in the set.
Caplan, D., Michaud, J., & Hufford, R., (2013). Short Term Memory, Working Memory, and Syntactic Comprehension in Aphasia. Cognitive Neuropsychology 30(2). doi: 10.1080/02643294.2013.803958.
Goh, W. D., & Pisoni, D. B. (2003). Effects of lexical competition on immediate memory span for spoken words. The Quarterly Journal of Experimental Psychology, 56A, 929-954.
Jalbert, A., Neath, I. & Suprenant, A.M. (2011). Does length or neighbourhood size cause the word length effect. Memory & Cognition, 39, 1198-1210.
Roodenrys, S., Hulme, C., Lethbridge, A., Hinton, M., & Nimmo, L.M., (2002). Word-Frequency and Phonological-Neighborhood Effects on Verbal Short-Term Memory. Journal of Experimental Psychology: Learning, Memory, and Cognition 28(6): 1019- 1034.
Roodenrys, S., (2009). Explaining Phonological Neighbourhood Effects in Short-Term Memory.
Faculty of Health and Behavioural Sciences. Retrieved from http://ro.uow.edu.au/hbspapers/1693/
Cite This Work
To export a reference to this article please select a referencing style below: