Effect Of Serial Position Biasness On Human Memory - Study And Findings

For the summative assessment component of the module, worth 70% of the overall grade, you are required to complete ONE FULL and ONE PARTIAL (Abstract, Method and Results only) lab report based on the two practical briefings you are required to attend in Semester 2. 

Human memory is a processor which gives the ability to analyse, recognize and decide. Acquired information from the environment is analyzed and recognized by the help of working memory.  Retention of information over a long or short period from gathered knowledge or previous incidents enables human to perform social and personal activities. In the current experiment two sets of ninety well defined words were chosen. Two experimenters divided the set of questions into 6 word lists with 15 words in each of the sets.  For randomness and un-biasness the words were split up in random though decent order. The difficulty level of the questionnaires was kept in mind during selection of the words.

The information in human memory passes through sensory memory to short term memory and finally gets stored in the long term memory.  These three memories differ from each other through their consciousness and characteristic of operations. Working memory of human brain processes information through it sensory memory to the short and long term memories.  The level of integration handled by these three memories is of specialized characters and degrees and cannot be handled by another system.  In the scope of the current experiment this particular factor was taken into account.  Methods in the study were adopted to evaluate the effect of channelization of information in human memory (Forgas, J. P., 2011).

The multi sector human memory reflects its working pattern which is observed in serial position effect. The serial position effect is it typical psychological situation where the brain shows tendency to remember the initial and the final information provided to it. It creates biasness and in the current study the effect of this biasness has been analyzed.  The short term memory stores the latest set of information and the long term memory accumulate the old set of information.  The tendency of the human working memory is to forget the in between information of an entire topic. The experiment set up was designed to study the effect for 34 participants compared to that of the control group.  

Serial position effect consist of primacy effect and recency effect.  In recency effect the short-term memory of the brain remembers the latest set of information and in the primacy effect the long term memory remembers the data which which has been collected during the initial stages of information gathering.  The centrally located information on the data set if homogenous in nature with the entire information, easily gets forgotten by both the memories (Parkin, A.J., 2000). This biased nature of memory has been studied in the experment by minimising the effect of serial position. Participants were tested for their verbal memomryy skills and recall pattern was noted.

Experimental Setup

In earlier research works the effect of primacy and recency of human brain have been discussed and methods to reduce these effects have been analysed (Tan, L., & Ward, G., 2000).

Joseph P. Forgas  in 2010 discussed about the elimination of  primacy factor buy negative effect by impression formation.  Tan, Lydia and Ward, Geoff in 2000 experimented with scheduled rehearsals to reduce the business of the brain and concluded that primacy and recency effects were still present in their study.

The current study tries to connect the loopholes of the earlier works by eliminating the primacy and recency effect of the brain. Experimental measures were taken to reduce these effects by two approaches which were expected to reduce the baisness of the human brain. Based on the theory of

double dissociation Brown Peterson effect was chosen to reduce the recency effect and incremental rehearsal strategy was adopted to reduce the primacy effect on the brain (Squire, L. R., 2004).  The control group participants were examined along with other participants, who went through Brown Peterson (delay) effect and incremental rehearsal strategy (n-1) effect.  It was a matter of interest whether adopted strategies to reduce these effects can actually do so. The experiment was based on the dual memory hypothesis and independent handling of primacy and recency effect was the main aim. Under the scope of the current study, finding parallel ways of managing both effects without affecting the other effect was the primary goal. Participants were separately studied for primacy and recency effects for this purpose. While adopting the measure to reduce primacy effect, recency effect of the subjects were noted. Again during the course of reduction of recency effect, primacy effect was noted down. The objective was to increase the number of recalls in middle section by balancing the serial position effect components independently.

Experiment with control group and reduced serial position groupsgroups separately. One-way analysis of variance was performed for characterizing the groups separately. It was expected that the subjects will tend to forget the initial few words from the list whem treated for thier recency effect. It was also expected that the subjects will tend to forget the final few words when (n-1) technique will be adopted. It was hypothesized that there was no difference in variance score for

primacy and recency under these three conditions. The alternate hypothesis was one directional in nature where the effect of the treatments were expected to reduce the average scores of primacy and recency independently compared to that of the control group. 

Results and Analysis

Subjects for the experiment were chosen from the University campus. An advertisement was distributed in handbill format to the University students where primary aim of the study was explained. A workshop was conducted with interested participants. After an initial verbal ability quiz a total of 34 subjects were chosen based on their scores in quiz. The chosen participants were asked to sign the ethical consent form of the school of psychology of the University.

The entire experiment was based on two factors, first on the word list and secondly on the primacy and recency reducing factors. Word lists were designed with 15 words in each set. Total 180 words were divided between two experimenters with 90 words each. Both of them constructed six word lists of their own by dividing the given words equally keeping in mind the difficulty level factor of the words. Control, delay and n-1, all these factors were permutated to construct six different orders in which the experiment could be performed.

Participants were provided with full size paper and two pencils and a pen. They were asked to mention their names and university roll numbers on the paper. Two experimenters read the words aloud to the participants and they were asked to recall the words. They were asked to write the recalled words in order of their recalling. The participants were tested under control condition, delay situation and n-1 condition.

The experiment was divided into three parts.  Participants were initially called in the first section of the experiment and a word list with 15 words was read to them by an expert. After a gap of two minutes a paper was given to them. Participants were asked to recall the words and write them on the paper in order of recalling. This was a control situation and no measure was taken to reduce serial position effect. Ten minutes were provided to the participants for recalling and writing the words. A gap of 30 minutes was provided to them. In the second session a new word list was read aloud to them. This time a variable delay of 0-30 seconds was imposed between the words. Participants were given an interpolated task in between the time gaps to stop rehearsing the words told to them. Participants then recalled the words and wrote in order of recalling. The third session was conducted after a break of two hours. The session started with a new word list of 15 words and words were read aloud to the participants. The experimenter adopted n-1 technique to dictate the words. The participants were asked to repeat the immediate predecessor of the word told by the experimenter. The primacy and recency scores were found from all the three sessions for the participants. The entire setup was replicated twice for randomness in sample data. 

The raw data was collected in a scoring sheet for control, delay and n-1 situations. Total 34 sheets were prepared for the experiment purpose. Every sheet contained six scores of a participant. Three scores were primacy scores under three conditions. And other three scores were recency scores under three experimental conditions. The average recall scores were found for control, delay and    n-1 conditions for each participant. This data is given in appendix (table 1). Average recall scores for 34 participants were found for all three experimental conditions and for all serial positions.

The mean of average primacy scores were 0.35 (S.D 0.18), 0.33 (S.D 0.19) and 0.22 (0.14) for three techniques. Levene test for the purpose of homogeneity was done. The statistic value was 1.63 where p value was 0.20 and hence the homogeneity of variances was rejected. A one way ANOVA was performed for primacy effect for the three groups. F value of 5.37 was significant as p value was 0.006 which was less than 0.05. Hence a significant difference in variance was observed between the groups. Tukey’s post hoc analysis was also performed for primacy scores. It was found that mean difference of control group and n-1 technique group was significant with p value less than 0.05. Significant difference in mean scores for delay and n-1 groups was observed where p value was 0.03.

The mean recency scores were 0.48 (S.D=0.17), 0.2 (S.D=0.14) and 0.53 (S.D=0.17) for three experiment conditions. Homogeneity test of Levene showed that variances were not at all significantly homogeneous. One way ANOVA revealed that difference in variances were very significant with p value of zero. The value of the F statistic was 42.74 and it was in the critical region. The null hypothesis was rejected and it was concluded that there was a significant difference 

in average primacy scores for three experimental situations. Tukey’s post hoc analysis was used to rest the difference in mean primacy scores. Difference in mean primacy scores of control group and n-1 technique group was not significant with p value less than 0.43. Significant difference in mean scores for control with delay and n-1 with delay group was observed where p value was less than 0.05..

The average primacy scores for control group was less than average recency score. This result was in line with the previous results. From table 2 of the appendix and average scores, it was observed that the control group scores were almost in parity with classical serial position effect. Figure 2 was drawn based on the calculated figures and the line graph was similar to the classical graph in figure1. The delay treatment was able to reduce the average score for recency condition. The average primacy score for delay treatment did not changed significantly. Hence reduction in recency effect was noticed and the reflection was in the line graph of figure 2. Earlier research work by Tan, Lydia and Ward, Geoff in 2000 was supported by our test results. The average primacy score for the delay treatment was less than that of control group because of obvious reason of working memory functioning. The post hoc analysis by Tukey HSD test calculated the significance value for this comparison as 0.91 which was greater than 0.05. Hence the comparison result was considered in significant.

The n-1 treatment on the other hand reduced average primacy score of the participants. The significant difference between the average primacy scores of n-1 treatment group with control group was observed. The average recency scores increased compared to the control group. The explanation of increase in recency score was quite obvious in nature. The working memory of a human brain balanced out the reduction in primacy effect. But it was observed that the average

recency score for n-1 treatment was not significantly greater than the mean score of the control group. Post hoc analysis by Tukey HSD produced p value for this comparison as 0.43 which was greater than 0.05. The results resembled with the search direction of Joseph P. Forgas who worked in this field in 2010.

The experiment results also indicated the fact that serial position effect is a combination of two independent conditions of primacy and recency effects. The dual memeory hypothesis was well supported by the test results and reduction of one affect withouth affecting the other score was significantly possible (Eysenck, M., 2001). The analysis of variance for primacy and recency effects revealed that theaverage scores of delay and n-1 condition were significantly different.

The effect size of the sample was not calculated in this experiment. Due to laboratory experimental purpose only fifteen words were selected and the entire experiment was done with 34 willing participants. In future work it will be interesting to increase the number of words in a list for the participants. A bigger sample size will also make a real time experiment with more constraints.

In a recent research work by Samuel W. Logan in 2011 the relation between receny and serial position effects and serial recall was examined (Logan et al., 2015).  The experiment was done by the help of bimanual end-state comfort task. The current study will become more interesting if execution of bimanual end-state comfort task for the participants gets included. The time factor was used in delay technique but not used for the purpose of the study. Variation in delay time and probability related to it was not included in the study. This critical aspect can restructure the entire experimental set up.

Visual information was also not used. It remains a scope for the future study to observe the serial position effect on visual slides instead of verbal information (Sands et al., 2014). 

References 

Eysenck, M. (2001) Principles of Cognitive Psychology. Hove: Psychology Press.

Egli, S. C., Beck, I. R., Berres, M., Foldi, N. S., Monsch, A. U., & Sollberger, M. (2014). Serial position effects are sensitive predictors of conversion from MCI to Alzheimer's disease dementia. Alzheimer's & dementia: the journal of the Alzheimer's Association, 10(5), S420-S424.

Forgas, J. P. (2011). Can negative affect eliminate the power of first impressions? Affective influences on primacy and recency effects in impression formation. Journal of Experimental Social Psychology, 47(2), 425-429.

Hurlstone, M. J., Hitch, G. J., & Baddeley, A. D. (2014). Memory for serial order across domains: An overview of the literature and directions for future research. Psychological bulletin, 140(2), 339.

Logan, S. W., & Fischman, M. G. (2015). The death of recency: Relationship between end-state comfort and serial position effects in serial recall: Logan and Fischman (2011) revisited. Human movement science, 44, 11-21.

Morrison, A. B., Conway, A. R., & Chein, J. M. (2014). Primacy and recency effects as indices of the focus of attention. Frontiers in human neuroscience, 8, 6.

Parkin, A.J. (2000). Essential Cognitive Psychology. Hove: Psychology Press.

Squire, L. R. (2004). Memory systems of the brain: a brief history and current perspective. Neurobiology of learning and memory, 82(3), 171-177.., Urcuioli, P. J., Wright, A. A., & Santiago, H. C. (2014). 21 SERIAL POSITION EFFECTS AND REHEARSAL IN PRIMATE VISUAL MEMORY. Animal cognition, 375.

Tan, L., & Ward, G. (2000). A recency-based account of the primacy effect in free recall. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26(6), 1589.

Warrington, E. K. (2014). The double dissociation of short-and long-term memory. Human Memory and Amnesia (PLE: Memory), 4, 61.