The word length effect

The word length effect is crucial to hypotheses explaining instantaneous memory. This means that there is a short-term mechanism that stores restricted information for a limited time (Kotzé et al., 2013). As a result, it has been argued that humans recall lists or items like one-syllable words more readily and precisely than longer words. The Word Length Effect has resulted in the direct development of working memory and the phonological loop, which is the best evidence for time-based decay. The ability to remember phonological information is significant since it distinguishes how human memory works. . This initial idea brings out the essence of rehearsal. The Word Length Effect occurs as the rehearsal is faster in short words than in long words. It has also been observed that the Word Length Effect depicts free-recall experimentations whereby people can retrieve the words in a random order.

Experiments have shown that instantaneous memory is not constant but the variation comes is dependent on the length of the words being recalled (Goldstein, 2011). Results have concluded the following; first, memory span is directly related to the word length, secondly, when the amount of syllables are constant, words of temporal time span are easily remembered than words with long syllables. Thirdly, the prediction of the span can be based on words read by the person in approximately 2 seconds, and lastly, in cases where the articulation is suppressed by for instance subjecting irrelevant sound to the subject then the word length effect disappears with the optical exhibition however the effect is maintained when the presentation is auditory (Goldstein, 2011). The paper therefore is an analysis of the word length effect with keen reliance on accurate data. It provides an item-order description of word-length effects in recall and recognition.

Hypothesis

The article seeks to explore the hypothesis that an immediate memory span varies with how long the words to be recollected are. The findings expected includes: memory span is negatively correlated to lengths of words and when phonemes remain unaltered, words with short duration can be easily recalled as compared to ones with longer time span. More so, the duration is predicted on the basis that words under consideration. Lastly, the hypotheses seeks to test whether word length effect diminishes when visual presentation is eliminated.

Materials and Methods

The data reported was collected through sampling and the data was then analyzed using SPSS software. Participants under study were 71 in number. Participants were collected across a wide spectrum ranging from age, weight and ethnicity. Each session comprised of 16 words presented sequentially depending on the number of syllables. Thereafter, each study subject was passed through a recall test. Words are drawn from a huge pull ranging from 1600 to 2000. Each word has a delay interval of 1200 ms before the next one appears on the screen. Every listed word was followed by 700-1100 ms stimulus interval. Every partaker was then given 75 seconds to recall the items that were presented. After collection of data, the use of SPSS was employed. This methodology provided an analytical representations under the topics: descriptive statistics, pairwise comparison, t-test, linear models and graphical representation.

The approach used is the time based decay process. The assumption based on information on word length effect is that each neuron is dedicated to a certain memory network. Also, the retrieved items act ass internal clue for the following item based on similarity between the related items. The explanation for this is that each retrieval process is directly associated with the memory representations of the listed items (Weinberg & Abramowitz, 2002). Dynamics in retrieval is represented by a certain sequence of items that are meant to be recalled by the participants. The first word used for comparison purposes has highest similarity to the presently recalled word. Ultimately, the recall duration comes to an end of the reclamation process when it enters a cycle where no more can be retrieved. One experimental procedure used was W = 1700, n – 0; 1, with N representing the neurons in the network. For each of the trial, 16 words were chosen by use of two methodologies. There were two intervals used. In the beginning, items were collected autonomously while in the second trial the words were selected randomly. Subsequent transitions were determined by a similarity matrix. The recall process was terminated during retrieval especially when the items enter a cycle which would mean that no more items would be retrieved (Neath and Nairne, 2005)

Results and Findings

An analysis was conducted on the data set collected from the lab report. The number of individuals investigated for one, two, three and four syllables were fixed at 71. The minimum statistic for one syllable was 0.14 while the minimum statistic for two syllable was 0.11. The three and four syllables were estimate at 0.14 and 0.17 respectively. All the trials were used, excluding intrusions and repetitions. For all the syllables, whole recall possibility was measured and computed as a percentage of the trials. Results from the SPSS analysis, shows that the mean for the one-syllable was 0.8101 while the mean for two-syllable was 0.6953, the average for the third and fourth-syllable was estimated at 0.6064 and 0.5827, respectively. Also, standard deviation for one syllable was 0.15865 while the one for a two syllable was 0.18735. The third and fourth syllables were calculated to have a standard deviation of 0.17883 and 0.17342 respectively. These results are in line with classical word length effect which lists short words as better recalled than long words. The constant standard error for skewness was 0.285 while the statistics were -1.723, -0.614, -0.405, and -0.474; all in sequential order.

In analyzing the proportion of correct words as a function of listed items, one-syllable had an 80 percent chance of correctness while two-syllable words had a 70 percent chance of correctness. Three-syllable had a 60 percent chance while four-syllable word had less than 60 percent chance of correctness. The results show a classical word length effect because they represent a negative recall stimuli as the number of words increase (Kumar, 2015). The chart below is a pictorial representation of the facts mentioned.

Furthermore, to test whether the word length effect is exactly as explained above, a correlation analysis was carried out. The measure used was an analysis of the mean by use of 95 percent confidence interval. The results for one-syllable shows a lower bound interval level of 0.773 and an upper bound result of 0.848; with a mean of 0.81 and standards error estimate of 0.19. Secondly, words with two syllables shows a decline of the lower bound interval to 0.651 and an upper bound value of 0.740. The respective means for this value is 0.695 and the standard error is 0.022. For the three syllable words, the confidence interval for the lower bound was 0.651 while the upper interval was 0.649. Estimated standard error and mean for this syllable were 0.021 and 0.606 respectively. Words with four syllables had the least confidence interval values. The lower bound value was estimated at 0.542 while recorded upper bound value was estimated at 0.624. These values are based on the mean value of 0.583 and a standard error value of 0.021.

The pairwise comparison table explains the significant difference in mean. The purpose of the pairwise comparison is to identify the differences that occurred between the mean values. Using Bonferroni Hoc approach, the analysis allows us to figure out where disparities occurred. The reason for using this method is because the ANOVA analysis above was significant. The planned contrast (difference between pre and post-intervention time) show significant differences in the longer words (Goldstein, 2011). For instance, in one-word syllables, the mean difference between the 2, 3 and 4 interventions were rather insignificant. However, as the words lengthen, the significance increases gradually. The tabulated results are in line with the overall analysis: the bigger the words, the longer the recall time.

Discussion

It is evident that articulatory suppression occurs as one moves from one-syllable to the four-syllable words. The experiment manipulates syllable length in an attempt to test effects of duration on the four measures of the period it takes to pronounce certain types of words. Serial recall produced robust effects as illustrated in the findings. From the graphical representation above, it is evident that one-word syllables have a success rate of 80 percent compared to four-syllable words, which according to the lab output, have a success rate of less than 60 percent. The central executive of the above results is that the function that ensures this effect is the working memory system (Goldstein, 2011). The purpose of the memory section of the brain is not only to store information but also coordinate how such info is distributed by the phonological loop. Therefore, the central executive plays a huge role in the storage of one task or in this case, two, three and four tasks.

A classical explanation of this subject is described by consistent complexity of items and increased rehearsal of items. All the statistical representation shows that brief words are generically easier to remember. In addition, shorter items can be easily learnt due to the little time used when rehearsing. The graphical representation and ANOVA analysis show a adverse correlation between the length of words and number of words recalled. Also, the average probability to recall a pool of items in random increases with increase with a representation of such items, as opposed to items with minimal representations. This is due to higher variability in terms of representation size. It is clear that memory is worse for long words especially if it takes a lot effort to pronounce (Goldstein, 2011). If the above chart was to be represented in a line graph, there would be a linear relationship between the time taken and the pronunciations rate. Other factors under study in such a case includes serial positioning of memory items, articulatory suppression and effects of modality.

Discussion concludes that a central feature to word length is that recall for short words is much better compared to that of long words. The current research is in line with these conclusions. The employed technique is the time-based approach which is presumably consistent with the decay theory. Just like the decay approach, this time-based system uses articulatory suppression to stop rehearsal (Baddeley, 2003). When suppression occurs, the visual stimulus transforms into phonemic code. One obvious implication of this research is the conclusion that memory span is strained by huge chunks of information. It is therefore attributed to temporal rather than structural limitations of the brain. One general assumption is that it is easier to pronounce monosyllabic that multisyllabic ones. On estimated marginal means, the difference is significant at the point of adjustment for multiple comparisons. These estimates make use of Bonferron-Holm, which is applicable to any family of pairwise comparisons. However, critics such as Baddeley et al. (2003) argue that stating the words repeatedly slowly eliminates the word length effect. This argument can be supported by the facts that probabilities show huge gaps even for words of a given similar length.

Conclusion

The research work has analyzed a large data set of recall experiments involving a mix of short and long words. The probability of recalled words are consistent across the arbitrarily chosen groups. The paper seeks to explain how word length contributes to the ability to recall. The study makes use of four syllables, each with a certain lengths. A repeated measure of ANOVA analysis is used to determine the whether the scholarly definition of word length approach has been achieved. Therefore, length of a words is the most prominent fact in the process of recalling.

References

Baddeley, A. D. (2003). Working memory and language: an overview. J. Commun. Disord. 36, 189–208. doi: 10.1016/s0021-9924(03)00019-4.

Goldstein, E. B. (2011). Cognitive psychology: Connecting mind, research, and everyday experience. Australia: Wadsworth Cengage Learning.

Kotzé, P., Marsden, G., Lindgaard, G., Wesson, J., Winckler, M., IFIP Technical Committee 13 on Human Computer Interaction & IFIP Conference on Human-Computer Interaction, INTERACT. (2013). Human-computer interaction -- INTERACT 2013: 14th IFIP TC 13 International Conference, Cape Town, South Africa, September 2-6, 2013 : proceedings. Berlin: Springer.

Kumar, A. A. (2015). Digital signal processing. Delhi: PHI Learning.

Lovatt, P. J. (1998). Immediate serial recall and the word-length effect.

Neath, I. and Nairne S. (2005).Word-length effects in immediate memory: Overwriting trace decay theory. Purdue University, West Lafayette, Indiana.

Weinberg, S. L., & Abramowitz, S. K. (2002). Data analysis for the behavioral sciences using SPSS. Cambridge: Cambridge University Press.