Rebecca Bavone

 Effects of Scenarios with Positive or Negative Valence on Emotional Memory Rebecca Bavone Western Connecticut State University Abstract Every day we process stimuli such as advertisements, music, movies, and social media that could have a role in subconsciously influencing our emotions, with some media being more impactful and memorable than others. This study examined the effects of priming in the form of reading scenarios of either positive or negative valence on the recognition or false recognition of words with the same valence as the scenario. Participants who read the negatively emotional scenario did not recall or falsely recall significantly more words with negative valence than those who read the positively emotional scenario. Numerous limitations such as the absence of a rating task for valence strength and a mood test could have been the reason for no effect being found. In conclusion, how negative and positive words are stored may not differ dramatically. Effects of Scenarios with Positive or Negative Valence on Emotional Memory If someone asked you to tell them the first memory that came to mind, you would probably remember something emotional, whether positive or negative. Maybe you recall the time you scraped your knee the last day of third grade and all the other kids stared at you, or smile at the memory of your tenth birthday party. As we grow, our memory becomes filled with impactful information, such as a comedic movie or an emotional song, and these memories can influence our behavior and way of thinking. Substantial research has supported the idea that emotional items are better remembered than neutral words (Schmidt, 2012; Brierly, Medford, Shaw, and David, 2007). Studies that examined emotional memory in further detail have focused on concepts such as context, categorization, and biological aspects (Schmidt, 2012; White, Bruno, Rotello, & Ratcliff, 2014; Richardson, Strange, & Dolan, 2015) . Schmidt (2012) also found that when emotional words were placed in a highly emotional context, they were more likely to be remembered than ones in a low emotional context. An example of high (e.g. stronger) emotional and low (e.g. weaker) emotional contexts are: My dog died last night and My car died last night, respectively. White and colleagues (2014) found that medium and high proportions of emotional words in categories resulted in an increase in memory of emotional items. In other words, the greater the number of emotional words shown in a category, compared to neutral words, the greater number of emotional words were remembered. So, increasing the proportion and categorization of emotional words increases emotional memory. Talmi and Moscovitch’s (2004) study showed that categorizing neutral words makes them just as memorable as non-categorized emotional words. Many experiments have narrowed down the areas of the brain that are involved in emotional memory (Richardson, Strange, & Dolan, 2015; Mickley, Steinmetz, & Kensinger, 2009). Richardson and colleagues’ (2015) study of the hippocampus-amygdala interaction resulted in multiple findings. Damage to the hippocampus caused a decreased ability to recall both neutral and emotional words, while a damaged amygdala only affected emotional memory. This can be explained by the fact that the hippocampus is involved in memory formation in general, and the amygdala governs fear and other emotions. Damage to the amygdala affected hippocampal activity regarding emotional items and left hippocampal damage affected emotional memory activity of the amygdala. These results showed that not only are both the hippocampus and amygdala involved in the memory formation of emotional stimuli, but they also interact with each other during the process. Full activity of the amygdala and the left hippocampus seem to be required for successful emotional memory formation. Mickley and colleagues (2009) found evidence that regions in the temporal and occipital lobes take part in forming memories of aversive items and items with positive valence are associated with the frontal and parietal lobes. This separates the components of emotional memory into valence (negative or positive). Priming can play a role in processing emotional memory. Schneider and Shriffin (1977) defined automatic processing as being a type of subconscious storing of information. The explanation suggests that being presented with a stimulus automatically brings up related stimuli, which is the process of priming. Perry (2003) found, for example, that presenting a word primed the spelling of a non-word, a combination of letters without meaning. Participants spelled the non-word with the same vowels as the prime word. Hoedemaker and Gordon (2014) employed an ocular version of the lexical decision task, or, in simpler terms, had participants classify items shown in triplets as words or non-words using eye movement. They found a semantic priming effect. More specifically, participants spent more time looking at the middle word when it was preceded by a semantically related prime. McRae and Boisvert (1998) also studied semantic priming, and found that word pairs high in semantic similarity showed more of a priming effect. I would like to clarify that semantic relatedness and semantic similarity have different meanings. Semantic relatedness refers to terms that are relevant to each other in a more general sense (e.g. dog and cat) whereas semantic similarity between words is more specific (e.g. dog and leash or cat and sleep). A specific type of priming, called the evaluative priming method, involves a prime with either a positive or negative valence followed by valence classification of target words. Freytag, Bluemke, and Fiedler (2011) used this method and found that words of the same valence as the prime were classified the fastest, also known as the congruity effect. The prime words are affective cues of the environment, and therefore the participant becomes more focused on words with equal valence to the prime. So, semantic relatedness and semantic similarity, as well as emotion, can be used as primes. The following experiment combined the aspects of the evaluative priming paradigm and emotional contexts to study emotional memory. The study featured scenarios of positive or negative valence that served as primes and, rather than measuring the reaction time of classification, participants were shown single words and then given a “recognition test”. I hypothesized that after reading the negatively emotional scenario, participants would recognize or falsely recognize more negatively emotional words than those who read the positively emotional scenario. The reasoning behind this is that contexts were shown to affect emotional memory, the formation of these memories seem to be somewhat separated by valence, and valence and semantic relatedness used as primes were shown to affect classification of target words (Schmidt, 2012; Mickley et al., 2009; Freytag et al., 2011; Hoedemaker et al., 2014). I focused on negative emotional memory because, according to the difference in brain activity for memory formation, it is possible one type of emotional memory is stronger than the other. More specifically, negative emotions could be more impactful and memorable than positive emotions. Methods Participants Participants (n = 10) were undergraduate students of Western Connecticut State University who volunteered and were offered class credit. After they gave informed consent, they were randomly assigned to the positive valence group (PV) (n = 5) or the negative valence group (NV) (n = 5). Materials Scenarios. The written scenario described a couple walking down a New York City street, and when it started raining, a man approached them. It was the same in both experimental conditions except for the last statement. PV read that the man started to breakdance in front of the couple and NV read that he violently mugged the couple (Appx. 1). The NV was designed to prime a negative emotion, regardless of whether the words in the narrative were related to the word list. The PV narrative was a primer for positive emotion. Memory Test. Participants were asked to remember eighteen words (Appx. 2), which were presented after reading the scenario. The list contained an equal number of negative, positive, and neutral terms. A memory test was given to measure the number of words retained. Subjects were required to recognize the learned words among another presentation of the original eighteen words intermixed with eight new words (3 negative, 2 positive, 3 neutral). Participants were meant to be primed with emotions that would influence their memory, so if the mean of one of the groups was more than six words, that could possibly indicate priming was significantly more effective in that group than the other. In this experiment, however, falsely recognized words were treated the same as ones from the original list. Procedure The experiment took place in a quiet room used for testing. Expectation bias was minimized by avoiding the use of vocabulary words related to the experiment until debriefing. I felt it was important to note the observations of the participants’ moods. There were not any apparent mood differences. Seated in front of a computer, participants read one of the two scenarios. Then the test wordlist was shown with each word sequentially for 1,500 milliseconds. Immediately after, they were given the recognition test. Participants were instructed to press a button on the keyboard if they recognized the word from the original list, and another key if they did not. The recognition test list was randomized so those in PV and NV saw the words in a different order, but those within the groups saw the words in the same order. Only the performance on the negative emotional words were recorded. Results This study compared the effect of negative and positive emotional priming on memory. I found no significant effect of the different scenarios on emotional memory, tind (8) = 0.00, p > 0.05. Participants who read the negatively emotional scenario (M = 6.00, SD = 1.00) did not recognize or falsely recognized significantly more words of negative arousal than those who read the positively emotional scenario (M = 6.00, SD = 1.00). Although the descriptive statistics are the same for both groups and there were six negative words in the original list, not every participant remembered those six words. Some participants recognized 5 words or 7 words, and not all these words were from the original list. Discussion This study found no effect of scenarios with different emotional arousal on emotional memory, a result inconsistent with the current literature. As mentioned in the introduction, it has been found that emotional memory is better than memory of neutral words (Schmidt, 2012; Brierly et al., 2007). According to the findings, negative and positive emotional memory may not be incredibly different. Therefore, as there was no effect of priming on the memory of words with negative or positive valence, the effect may mainly concern general emotional memory and the memory of neutral words. There is research supporting the idea of a difference between memory for the two types of words (Mickley et al., 2009). White and colleagues (2014) found that the higher proportion and categorization of emotional words increase emotional memory compared to neutral words. This is evidence that the retrieval of emotional items, regardless of valence, is more easily influenced than the retrieval of non-emotional items. If this study were to be replicated, neutral words should serve as a third group that would show if the priming of general emotional memory could occur under the conditions of the experiment. Although there are different brain regions that activate for positive and negative emotional items, they may not be significantly involved in memory formation, or they may be activated in similar ways (Mickley et al., 2009). In other words, the difference between the memorization of words depending on their valence may not be large enough to be influenced by priming, the priming effect may have to be stronger, or memory formation may be too similar for an effect to be found. Emotional memory formation in general is a process involving the interaction between the hippocampus and the amygdala (Richardson et al., 2015). This activation could simply dominate over any other smaller activations that may separate items by valence - the occipital and temporal lobes, and the frontal and parietal lobes (Mickley et al., 2009). Schneider and Shriffin’s (1977) definition of automatic processing connects to priming. Items need to be primed appropriately for associations to be brought up. This study depended on emotional context as the prime, whereas other studies have used a prime word within a pair of words or a category of words (Perry, 2013; Hoedemaker et al., 2014; McRae et al., 1998). Schmidt found that high emotional context increased emotional memory (2012). The scenarios used may not have been emotionally strong enough to produce an effect. In a replication of this study, they should either be enhanced somehow to make them arouse higher levels of emotion or visual scenarios should be used. The scenario used in this study may have had words besides the prime word that were considered negative or positive to some (e.g. New York City). To control for this, any specific places or words that could bring up emotion should be eliminated or changed. Studies have found priming effects when words were semantically related or similar, and other studies have used emotion as a prime (Hoedemaker et al., 2014; McRae et al., 1998; Freytag et al., 2011). Although I utilized these aspects in this study by adding in some words of the same valence and some of the opposite valence as the scenario into a word list, the strength of emotional arousal was not determined. Some words may have been weak in negativity or positivity compared to the prime word and other words in the list. If all words, including ones of the opposite valence, were of equal strength and possibly of increased strength, then priming may have an effect on recollection or false recollection. Strong emotional arousal may have to be present for the prime word to bring up associated items. To test for the similarity and strength of valence, a separate group of participants should rate negative and positive valence of the words (this includes the prime word, the original word list, the added words for the “recognition test”, and the ‘filler’ words in the scenario). For future replications of this study, the experiment could be divided into conditions of valence strength to test for any effect of level of emotional arousal on priming of emotional memory. There are some limitations to this study. I chose to test the difference between positive and negative valence rather than general valence and neutrality because I wanted to focus on a possible difference in the biology of emotional memory formation. Adding in neutral words as a third group could show a difference between negative and positive words as this new group would act as a control group. I did not use a standardized test to control for mood. Some participants could have had a more dramatic positive or negative mood than observed and this could have affected the results of the study. If someone was in a negative mood, they could have focused more on the negative words, or could have been less affected by the positive scenario if they were assigned to that level. No rating task was given to make sure all the words were of the same valence strength (Schmidt, 2012). This would have been helpful because the weakest emotional words would have been discovered and removed. Even if valence strength was not an issue, the scenarios in general may not have been emotionally strong enough to prime the recollection or false recollection of positive or negative words. Order effects were only controlled for within groups. To control for order effects across all subjects, the “recognition test” should be randomized for every participant. Another issue could have been that the recognition test was not difficult enough to draw an effect. Adding more words to either the original list or distractors could address this problem. Visual scenarios may be more emotionally arousing than merely reading words. A significant effect might be found if pictures or videos are used rather than written scenarios. To avoid confounds, the experimenter should take the pictures themselves and use the same person or people. Although the findings were not significant, prime valence may still influence emotional memory. Using knowledge of possible brain regions that are activated during memory formation and by focusing on emotional memory, I have evidence that the biology of storing positive and negative words may not be very different. On the other hand, this study could be improved in numerous ways that might provide insight. Emotional memories are stronger than neutral ones, but the question that still has yet to be answered is whether a knee scrape during one’s childhood or a delightful tenth birthday party is more memorable. References Brierley, B., Medford, N., Shaw, P., & David, A. S. (2007). Emotional memory for words: Separating content and context. Cognition & Emotion, 23 (3), 495-521. Retrieved from http://web.b.ebscohost.com/ehost/detail/detail?vid=0&sid=371b9c60-5fb9-4be0-a157-f1e0e7de2f8a%40sessionmgr104&bdata=JkF1dGhUeXBlPWNvb2tpZSxpcCx1cmwsY3BpZCZjdXN0aWQ9d2VzdGNvbm4mbGFuZz1lcyZzaXRlPWVob3N0LWxpdmU%3d#db=aph&AN=-. Freytag, P., Bluemke, M., & Fiedler, K. (2011). An adaptive-learning approach to affect regulation: Strategic influences on evaluative priming. Cognition and Emotion, 25 (3), 426-439. Retrieved from https://www.researchgate.net/publication/-_An_adaptive-learning_approach_to_affect_regulation_Strategic_influences_on_evaluative_priming. Hoedemaker, R. S. & Gordon, P. C. (2014). It Takes Time to Prime: Semantic Priming in the Ocular Lexical Decision Task. Journal of Experimental Psychology: Human Perception and Performance, 40 (6),-. Retrieved from http://dx.doi.org/10.1037/a-. McRae, K. & Boisvert, S. (1998). Automatic Semantic Similarity Priming. Journal of Experimental Psychology: Learning, Memory, and Cognition, 24 (3), 558-572. Retrieved from http://eds.b.ebscohost.com/ehost/detail/detail?vid=2&sid=3407fee1-ec86-41a2-aaf3-8eefdcda9384%40sessionmgr104&hid=101&bdata=JnNpdGU9ZWhvc3QtbGl2ZSZzY29wZT1zaXRl#AN=643306&db=aph. Mickley Steinmetz, K. R. & Kensinger, E. A. (2009). The effects of valence and arousal on the neural activity leading to subsequent memory. Psychophysiology, 46,-. Retrieved from https://www2.bc.edu/elizabeth-kensinger/MickleySteinmetz_Kensinger_Psychophys09.pdf. Perry, C. (2003). Priming the rules of spelling. The Quarterly Journal of Experimental Psychology, 56A (3), 515-530. Retrieved from http://eds.a.ebscohost.com/ehost/detail/detail?vid=2&sid=07aa529d-0119-4ddb-84ff-921d0f41dcbc%40sessionmgr4008&hid=4205&bdata=JnNpdGU9ZWhvc3QtbGl2ZSZzY29wZT1zaXRl#AN=-&db=aph. Richardson, M. P., Strange, B. A., & Dolan, R. J. (2004). Encoding of emotional memories depends on amygdala and hippocampus and their interactions. Nature Neuroscience, 7 (3), 278-285. Retrieved from http://dx.doi.org/10.1038/nn1190. Schmidt, S. R. (2012). Memory for emotional words in sentences: The importance of emotional contrast. Cognition & Emotion, 26 (6),-. Retrieved from http://web.b.ebscohost.com/ehost/detail/detail?vid=0&sid=480ffc61-5eea-457f-996a-32c7d-%40sessionmgr102&bdata=JkF1dGhUeXBlPWNvb2tpZSxpcCx1cmwsY3BpZCZjdXN0aWQ9d2VzdGNvbm4mbGFuZz1lcyZzaXRlPWVob3N0LWxpdmU%3d#AN=-&db=aph. Schneider, W, & Shriffin, R. M. (1977). Controlled and Automatic Human Information Processing: I. Detection, Search, and Attention. Psychological Review, 84 (1), 1-66. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=-&rep=rep1&type=pdf Talmi, D., Moscovitch, M. (2004). Can semantic relatedness explain the enhancement of memory for emotional words? Memory & Cognition, 32 (5), 742-751. Retrieved from http://eds.b.ebscohost.com/ehost/detail/detail?sid=c--c040-4cc9-a413-7acb0eeec62a%40sessionmgr102&vid=0&hid=103&bdata=JnNpdGU9ZWhvc3QtbGl2ZSZzY29wZT1zaXRl#AN=-&db=aph. White, C., Kapucu, A., Bruno, D., Rotello, C. M., & Ratcliff, R. (2014). Memory bias for negative emotional words in recognition memory is driven by effects of category membership. Cognition & Emotion, 28 (5), 867-880. Retrieved from http://web.b.ebscohost.com/ehost/detail/detail?vid=0&sid=8b8b-a-4a74-80f9-e02bc86b4d5e%40sessionmgr120&bdata=JkF1dGhUeXBlPWNvb2tpZSxpcCx1cmwsY3BpZCZjdXN0aWQ9d2VzdGNvbm4mbGFuZz1lcyZzaXRlPWVob3N0LWxpdmU%3d#AN=-&db=aph. Appendix 1 Negative emotional scenario: Charlie and Claire were visiting New York City for a day. As they were walking down the street, hand in hand, it began to rain and everyone ran into the nearest building. Charlie and Claire, however, kept moving down the street. A few minutes later, a man neared quickly and violently mugged them. Positive emotional scenario: Charlie and Claire were visiting New York City for a day. As they were walking down the street, hand in hand, it began to rain and everyone ran into the nearest building. Charlie and Claire, however, kept moving down the street. A few minutes later, a man neared quickly and started to breakdance. Appendix 2 Word List: dog sad joy anger laugh table happy apple fear pain lamp funny lizard smile anxiety exciting despair grape Recognition Test: mouse, sad, hope, anger, laugh, table, happy, apple, afraid, pain, light, funny, lizard, smile, anxiety, exciting, despair, grape, depressed, cherry, dog, joy, lamp, fear, rage, giggle