Effects Of Emotional Valence And Age On False Recall Of Words

Findings from Classical Literature and Empirical Research

Human memory is susceptible to distortion. In real life, people have been found to remember events that never occurred or fail to clearly recall portions of events that actually took place. False memory occurrence has also been studied in the laboratory by using an array of words with phonetic or semantic association with certain words of interest. Such experiments follow the DeeseRoediger-McDermott (DRM) paradigms (Roediger & McDermott, 1995). In such experiments, participants are exposed to a list of words that are related to a target (lure) word, which they are not presented with. An assessment of words recalled gives rise to the false recognition of the lure word although it was not presented to the participants.

According to Kaplan (2015), emotional stimuli activate the production of false memory alarms. Van (2013) studied the false recognition of lure and neutral words using the DRM paradigm. The results of his experiment demonstrated that there existed significant differences in the false recall of words based on the distinctiveness of the lures. Neutral words were more correctly recalled as compared to lure words. Further, the study participants were more prone to identify emotional lures when the word list contained other emotional words than when the list contained neutral words. These results, however, provided contrary findings to those of Huff (2013) who found that emotional words were more likely to be associated with false alarms as compared to neutral words. The researcher argued that emotion induced an attention bias, which was associated with a higher likelihood of identifying emotional stimuli.

Aging is another factor associated with memory accuracy and may be associated with the increased production of false alarms. In list-learning or recall experiments, older people may not be as effective as younger ones in differentiating studied from unstudied items. This is the case due to decreased efficiency in cognitive control processes among older people as compared to among younger individuals. According to Abichou (2020) older adults are more prone to the production of alarms than younger people. However, more false recalls are associated with neutral words than emotional words among older individuals.  Some studies have found equivalent memory for positive and negative stimuli among young and older people (Dennis, 2017). Other studies have, however, found that there is a difference in memory for positive and negative stimuli between young and older adults (Dang, 2020).

Based on the findings of classical literature and empirical research discussed above, researchers have found varying results on how the memory of words differs based on age and emotional valence of words. It is for this reason that the present study seeks to replicate elements of some of these studies with an aim of investigating the effects of the emotional valence of words and age on rates of false recall/ recognition.

Aim and Research Question

This research project sought to compare younger and older participants’ memory performance in relation to false recall of words with emotional valence. This quantitative study has non-directional hypotheses and aims to answer the question: Does the emotional valence of words and age have effects on rates of false recall/recognition?

The study sought to evaluate the following two hypotheses;

Hypothesis 1 – words with emotional valence will have an effect on the rate of words falsely recalled.

Hypothesis 2 – age of participants will have an effect on the rate of words falsely recalled.

This study employed the DRM paradigm for testing the false recall of the words with emotional valence among all the participants. The word lists contained neutral and negative emotional valence, which were derived from the study of Zhang et al., (2016). The research group members shared the data for further individual analysis.

A within-subjects design was employed to investigate hypothesis 1. Hypothesis 1 aimed at testing if there was any difference in the number of words falsely recalled by participants between word lists that are neutral, and words that have a negative valence. The independent variables were the neutral words and negative valenced words in lists. The dependent variable was the number of lure words falsely recognized.

A between-subjects design was followed to test the second hypothesis. Hypothesis 2 sought to investigate if there were differences between two age groups (18-49 and over 60) in recalling the words with negative valence. The independent variables were age and negative valenced words. The “negative” valenced variable was manipulated by the “neutral” variable to see their effects on recall. Participants viewed three neutral word lists and three negative valenced word lists in random order. The errors reported informed the dependent variable.

A stratified sample consisting of 81 participants took part in the study. The sample was stratified as follows.

1) Younger group (18 – 49) – 51 participants.

2) Older group (60+) – 30 participants.

Opportunistic sampling was conducted. The participants were drawn from among individuals well known to the research members from their social circle (family, friends, colleagues, acquaintances). The potential participants were contacted via telephone or email or instant messaging. Participation in the experiment was voluntary.

Participants were presented with six lists each containing ten words. Three of the lists consisted of neutral words and three had a negative valence. The lists used are six (three neutral and three negative valence) of the twelve lists created by Zhang et al. (2016) and as reported by them, words in each list are related to the lure word, the emotional valence is consistent for all words in a list, and the words do not vary in levels of emotional arousal.

Hypotheses

The word lists were presented as a video created in PowerPoint, with each list lasting twenty seconds, interspersed with a slide Formatted: Font: (Default) Arial, 12 pt Formatted: Font: (Default) +Body (Calibri), 11 pt indicating the start of the next list, lasting five seconds. The video lasted 2 minutes 40 seconds. In order to counteract any effect the order list may have, six videos were made, each containing the neutral words and negative valence word lists in different sequences, and were randomly assigned to participants.

Testing started immediately after the presentation of all six-word lists, as is the method in more than one published study that uses DRM – type lists with lure words (Roediger & McDermott, 1995; 2000). The participants were presented with an A4 sheet containing thirty-six words which contained a random arrangement of words 1, 4 and 8 from each of the viewed lists which were derived from Zhang et al. (2016) and the six critical lures associated with each list. There were also twelve unrelated items, which were selected to be consistent in word length, significance, and valence to the words included in the video.

Participants were asked to write down their age on the A4 sheet of 36 words before proceeding to mark all the words that they remember seeing in the lists presented on the video. The researcher made the participant aware that the test had been completed and thanked them for taking part before offering them the information sheet and a copy of their consent form. How they can gain access to debriefing information was also be made available to them. The dependent measure was the number of critical lure words that are (falsely) recalled, as in a study with variables that also relate to age and the recall of false lures (Lövdén, 2003). Participants were scored for each critical lure falsely recognized for both the categories in the six lists (neutral x 3, negative x 3). Scores could range from 0, 1, 2, 3 for each category, and overall scores of 0, 1, 2, 3, 4, 5, 6 across both categories. The total number of scores from each age group was then be compared.

Hypothesis 1- Paired sample t-test was used to determine how the participants performed across two conditions and whether the mean change was significantly different from zero between false recall for negative valence words and false recall for neutral words for each participant.

Methodology

Hypothesis 2 – Independent sample t-test was employed for examining the difference of false recall for negative valence words between two age groups. It compared the mean scores of the two different groups (older age group and younger age group) and allowed for determining whether there was statistical evidence that false recall is significantly different between these two age populations.

A total of 81 participants took part in the study. 51 were from the younger group while 30 from the older group. 47 of the participants were female while 34 were male. Among the younger group, there were 31 females and 20 males. Among the older group, there were 16 females and 14 males.

Table 1 depicts the distribution of neutral lure words while table 2 depicts the distribution of negative lure words by gender.

Hypothesis 1

To test the first hypothesis, a paired samples t-test was conducted.

The below bar graph depicts the mean and error bars for the no of negative and neutral lure words. 

The paired samples t-test results are presented and discussed below. 

The descriptive statistics contained in table 3 reveal that the average number of negative lure words (M=1.728, SD=0.8805) was higher than the average number of neutral lure words (M=1.469, SD=1.050). 

Table 4: Paired Samples Test
	Paired Differences	t	df	Sig. (2-tailed)
Mean	Std. Deviation	Std. Error Mean	95% Confidence Interval of the Difference
Lower	Upper
Pair 1	No. Of Negative Lure Words – No. Of Neutral Lure Words	.2593	1.0929	.1214	.0176	.5009	2.135	80	.036

Upon conducting the paired samples t-test, the analysis found that there exists a significant mean difference between false recall for negative valence words and false recall for neutral valence words (t=2.135, p=0.036). The results can therefore be interpreted as to support hypothesis 1.

The second hypothesis was tested by conducting an independent samples t-test for examining the difference of false recall for negative valence words between the two age groups.

The bar graph depicting the mean and error bars of negative lure words is as shown below; 

Below are the results of the independent samples t-test analysis.

The results contained in table 5 suggest that the mean number negative of lures was higher for the younger group (M=1.824, SD=0.9101) than for the older group (M=1.567, SD=0.8172).

The difference in means can be represented in the below table; 

Table 6: Independent Samples Test
	Levene’s Test for Equality of Variances	t-test for Equality of Means
F	Sig.	t	df	Sig. (2-tailed)	Mean Difference	Std. Error Difference	95% Confidence Interval of the Difference
Lower	Upper
No. Of Negative Lure Words	Equal variances assumed	.119	.731	1.273	79	.207	.2569	.2018	-.1448	.6586
Equal variances not assumed			1.309	66.286	.195	.2569	.1962	-.1349	.6486

The results of the independent sample t-test as shown in table 4 above suggest that the difference of false recall for negative valence words between the two age groups is not significant (i=1.273, p=0.207). This implies that hypothesis 2 is not supported.

The study sought to investigate whether the emotional valence of words and age have effects on rates of false recall/ recognition. Hypothesis one sought to investigate whether or not there exists a significant mean difference between false recall for negative valence words and false recall for neutral valence words. Results obtained upon conducting a paired samples t-test suggested that the average number of negative lure words (M=1.728, SD=0.8805) was higher than the average number of neutral lure words (M=1.469, SD=1.050). These results were significant implying that the hypothesis was justified. The findings of hypothesis 1 are consistent with those of Kuperman (2014) who suggested that there is a higher likelihood of false recall for negative valence words than for neutral lure words.  However, the findings are inconsistent with those of Zhang et.al (2016) who reported that words in each list created by them are related to the lure word. They further suggested that the emotional valence is consistent for all words in a list, and the words do not vary in levels of emotional arousal.

Participants and Sampling

Hypothesis 2 sought to evaluate the difference of false recall for negative valence words between the two age groups. Upon conducting an independent sample’s t-test, it was found that the mean number negative of lures was higher for the younger group (M=1.824, SD=0.9101) than for the older group (M=1.567, SD=0.8172). However, the difference was not significant (p>0.05). This implies that hypothesis 2 is not supported.

The findings of hypothesis 2 are contrary to those of Ballhausen (2015) who found that there is a significantly higher rate of false recall for lure words among older adults than among younger adults.

Conclusion

Findings of the present study contribute to the existing evidence that the emotional valence of words has an impact on false recall/ recognition, with negative lure words being more prone to false recall than neutral lure words. At the same time, the findings obtained upon analyzing the effect of age on false recall deviate from the existing body of evidence which suggests that as humans age, memories are susceptible to distortion. Further research should therefore be conducted to identify the problem which led to the contradiction in results. Further, research should be conducted to investigate the effect of the interaction between valence of words and age on false recall.

References

Abichou, K. (2020). False memory in normal ageing: empirical data from the DRM paradigm and theoretical perspectives. Geriatrie et psychologie neuropsychiatrie du vieillissement, 65-75.

Ballhausen, N. (2015). Emotional valence differentially affects encoding and retrieval of prospective memory in older adults. Aging Neuropsychology and Cognition, 544-559.

Dang, X. (2020). Age differences in false memories for visual scenes and the effect of prior recall. Journal of Pacific Rim Psychology, 14.

Dennis, N. (2017). The influence of perceptual similarity and individual differences to false memories in aging. Neurobiology of Aging, 221-230.

Huff, M. (2013). When does memory monitoring succeed versus fail? Comparing item-specific and relational encoding in the DRM paradigm. Journal of Experimental Psychology: Learning, Memory, and Cognition, 1246.

Kaplan, R. (2015). Emotion and False Memory. Emotion Review, 8-13.

Kuperman, V. (2014). Emotion and language: Valence and arousal affect word recognition. Journal of Experimental Psychology: General, 1065.

Roediger, H. L. & McDermott, K. B. (2000). Tricks of Memory. Current Directions in Psychological Science, 9(4), 123-127. https://doi.org/10.1111/1467- 8721.00075    

Van, D. (2013). Mood and the DRM paradigm: An investigation of the effects of valence and arousal on false memory. The Quarterly Journal of Experimental Psychology, 1060-1081.

Zhang, X., Lee, L. H., Hao, S., Wang, J., He, Y., Hu, J., … & Yu, L. C. (2016, June). Building Chinese affective resources in valence-arousal dimensions. In Proceedings of the 2016 Conference of the North American Chapter of the Association for Computation  Linguistics: Human Language Technologies (pp. 540-545).