Download as pdf
Download as pdf
You are on page 1of 10
2221 Spee Pohl Aon The Effects of Emotional Valence and Perceived Life Stress on Recalling Personal Experiences and Envisioning Future Events Azara Lalla and Signy Sheldon Department of Psychology, MeGill University Although i s underod that engaging in mental time tive 1 remember the past and imine the {uur relies on similar cognitive processes, tere are impetnt differences. Notably, thee is evidence that emotional valence difeenly affects how post and Tulue evens are accessed. Here, we explored the differential effect f emotional valence on pas and fue event generation in the context of personal stress. This is bused en findings that an individual's eurent fe sess ~a metric of mental heal ~ alters motional mental time tase. In an online experiment condhsted ding the novel consis 2019 (COVID-19) pandemic, 421 participants generated specific past and fate events toa series of positive snd negative eves and then rated the likeibood thatthe event would occur in the Fatre, the emotion conveyed in the event, and the dificully of generating the even. Participants also completes question naire estimating current le stress. We founda general bias toward gencrting specific postive future events that ws not prosent when sensratins pst events, Adtonlly, we found & smal bat since effect of stress levels on ratings of difficulty and likelihood for evens generate in espons to positive ues, These reults provide new insight into how an individual's curent sess selectively targets the ‘way positive ifs evens are persive. “Keywords: mental ime tavel,aulobiograhical memory, pode flare thinking, emuton, sess Mental time travel is the ability {0 aecess autobiographical ‘events from different time periods and involves both recollecting ‘experiences and envisioning future events (Addis, 2020; Buckner & Caroll, 2007; Schacter & Addis, 2007; Tulving, 1985). ‘Although these Facets of mental time tavel both depend on epi sodic memory to simulate and experience mental events (Mosco- vitch etal, 2016; Schacter & Addis, 2007b; Schacter eta, 2012; Sheldon etal, 2019: Sheldon & Levine, 2016), there are notable distinctions in the way they are formed, Foremost, fulure events fare generated with less episodic specificity than past events (Anderson & Dewhurst, 2009; Williams et al, 1996), which has been attributed to the greater cognitive effort needed to construct ‘imagined events rather than reconstruct experienced ones (Addis tal, 2007; Schacter ct al, 2012). Another notable difference is that future events include more positive emotional content than. pas events (Rasmussen & Bemtsen, 2013), These findings, how= ‘ver, rellect objective differences in how past and future events are generated, Teaving open questions about whether emotional ‘Thisatcle was published Online First November 15,2021 Ava Lalla © hipsorid or 000-043 2608-7430 Materials and da fro this stady are available on request. This research was funded by a Naural Sciences and Engineering Research Council of Canada Discovery Grint RGPIN-04241 awarded to Signy Sheldon, We ‘would like to acknowledge List Zh, Kayla Willams, and Catrina AAgostina for thei assistance ith data cleaning and scoring. The authors Faive no conflict of inert odsckne ‘Conespondence concerning this article should be adiesed to Signy ‘Sheldon, Department of Psychology. McGill University, 2001 McGill Avene, Montel, QCHA 1G1, Cana, Ena: siny seldom @mepilck 1392 events from the past and future are also evaluate differently. This is 4a important research question to address because the way pers perecives positive or negative evens in ther life can impact behav fons inthe real world. Additionally, a person's current level of ie ress might influence how emotions events from the pas a future are accessed and evaluated, based on research showing impairments to mental ime trvel in stress related disorders (Williams & Bros bent, 1986; Williams eta, 1996). Here, we aim to uncover how emotional valence affects mental Gime tnivel while considering the ellects of individual differences in caren life stress. ‘The Effect of Emotional Cues on Mental Time Travel ‘There isa reliable effect of positive emotional valence on men- tally time travelling to future events, such that positive events are rmore likely to he accessed than negative events (Baursics et l., 2016; Gallo eta, 2011; Newby-Clark & Ross, 2003; Rasmussen & Bemisen, 2013; Ross & Newby-Clark, 1998). One explanation {or this effect is thatthe ability to imagine postive Future events allows one to focus on future goals, which then guide adaptive behaviors Ballance et al, 2021; Taylor & Brown, 1988; Taylor & Brown, 1994). IL is no surprise, then, that the ability to imagine ‘uuure positive events has been positively linked to well-being (MacLeod & Conway, 2007) and negatively linked to depression (MacLeod & Byme, 1996; MacLeod et al., 2005; Quoidbach et 1 2008) Interestingly, the positivity bias seen when imagining, future events Is not consistently reported when people remember past events. For example, one study found that postive future events are accessed with more specificity than are negative future events, EMOTION, STRESS, AND MENTAL TIME TRAVEL wher 1 positive and negative past events are remembered with ‘qual Tevels of specificity (Gallo et al., 2011), Another study found that postive future events are accessed more quickly in response to cues than negative future events (D’Argembeau & Van der Lin- den, 2004; Newby-Clark & Ross, 2003). In contrast, whereas some work has found that positive memories are recalled swith ‘more detail relative to negative memories (Barry et al, 2021; Berntsen & Rubin, 2002; D’Argembeau & Van der Linden, 2004 ‘Talarico et al., 2000), this effect has been allributed to emotional arousal raher than valence (Sheldon et al, 2020; Simpson & Shel- don, 2020; Talarico et a, 2004), suggesting a distinction between the elfects of positive and negative emotion on past and future ‘mental time tavel To dale, the positivity bias in future imagination and the lack therco in remembering the past—has been evaluated by measuring the objective ability to access specific events, which are those ooeur- fing in one time and place, in response to emotional cic (Williams ‘& Broadbent, 1986). It is equally imporant to establish if this bias Persisis for the way events are evaluated, because event evaluation ‘can determine how an accessed mental event willbe used to guide behavior. For example, if person perceives @ negative past event (ea, being stood up fora date) as likely to happen again, this evalu ation can impact person's future behaviors (DeSteno et, 2000; ‘Wright & Bower, 1992), Altemately, if person has dificuly imag ining a postive fuure event (ex, attending a binthday purty, that individual might be les likely to engage in the steps for tht event {0 occur (eg, planning the party; Danziger et al, 2006; DeSteno etal, 2000; Hadklock, 2001), ‘The Influence of Individual Difference Mental Time Travel Stress on (One factor known to influence both the acess to and evaluation ‘of emotional events isan individuals" mental health status, Deficits in mental time tavel are a hallmark symptom of many stess- related mental illnesses, including major depressive disorder (MDD; Dalgleish & Wemer-Seidler, 2014), Several studies have reported a reduction in the ability to recollect or imagine specific personal events, a phenomenon known as overgeneral memory, in MDD (Buss ot a, 2004; Kleim & Ehlers, 2008; Moore & Zool ner, 2007; Williams & Broadbent, 1986; Williams et a., 1996). Additionally, a common finding is that MDD reduces the ability © ‘imagine spectic, positive events, whereas enfuancing the abilty 1 recall negative autobiographical knowledge (D’Argembeau et al, 2011; Gamble et al, 2019; Thimm et al, 2013; Williams et al, 1996). These effects might be present in individuals at risk for MDD, such as those with high levels of current life stress (Richter-Levin & Xu, 2018 Yang etal, 2015). Thus, itis impor- tant to consider how individual variability in current stress might affect emotional mental time travel To summarize, the overall aim ofthe current study was to refine knowledge on how emotion affects the access and evaluation of past and future events while also considering individual differen cet in current life stress. This consideration is particularly relevant ‘as many people have experienced notable increases in life stress during the COVID-19 pandemic. We hypothesized that if stress affects emotional mental time travel in a similar fashion as ‘observed in MDD (Flory & Yehuda, 2015; Kleim etal, 20145, Moore & ZoelIner, 2007; Ono etal, 2016; Sumner eta, 2010), 1393 then poople reporting higher stress will have selective alterations {tohow positive events are accessed and imagined Method Participants Participants (W = 645) were recruited from the online crowd: sourcing platform Amazon Mechanical Turk (MTurk; Crowston, 2012), This sample size was based on previous work assessing uobiographical events in an online sample (Verschuere & Klci bers, 2017). Data were collected from April 20 through June 3, 2020, Prior to scoring, 34.73% of the participants were excluded due to submiting nonsensical responses (eg, numbers, strings of letters, copy and pasted text) or completing less than 50% of the task. A final sample of 421 participants were included in the study: 46.8% of the participants completed Version 1, and 53.2% com- pleted Version 2. All participants lived in Canada or the United States and sere free of any neurological or psychiatric disorders Tinformed consent was obtained prior testing, and compensation was $4.77 USD. All procedures were in accordance with the guidelines outlined by the Research Ethics Board of McGill University, Procedure ‘A link was generated wsing the online survey hosting platform SurveyMonkey (Momentive Inc., 2021), which was provided to participants through MTurk. Participants first filled out a demo- graphics questionnaire that involved answering questions about basic sociodemographic information. They then completed an autobiographical memory and future cueing task (Williams & Broadbent, 1986), which we refer to as the past event task and the uur event task, respectively. This was followed by the Perceived, ‘Stress Scale-10 (PSS-10; Cohen, 1988). Stimuli ‘Twelve words were used as cues with which to prompt subjects for responses, six postive and six negative words, matched for ‘word frequency and chosen to be high on emotionality (Williams & Broadbent, 1986; Willisms et al, 1996). The words were di- vided into (wo matched sels of six words, with three positive and. yee negative words in each set. One set was used for the past ‘event task and one set was used For the Future event task. Past Event and Future Event Tasks ‘Version | of the task consisted frst of the past event task fol- Towed by the future event task; Version 2 consisted of the future event tek followed by the past event task. Instructions for each task were provided prior 10 the begining of each. Each task eon sisted of six trials. In toa, there were 12 trials and all participants access six past events and six future events Participants were insiucted to remember oF imagine & specific personal event that happened in the last few years oF that could happen in dhe next few yeats in response 10 a word cue. Specific events were described as personally experienced events cecurting none location within one day. They were (old that their accessed 1304 ‘events could be important or trivial and related or unrelated 10 the word cue, a long as they were spacfic events. Examples of spe cific past events (eg, “Fating dinner for my birthday lst June at nice laian restaurant”) and specific future events (c.g. “Attending ‘my nieces wedkling in August 2022") were provided. If they could ‘ot access a specitic event within I min, they were instructed 10 type SKIP. Participants were tested om their comprehension of the instructions by answering two questions about whether example events were specific or nonspecific During the task, participants were presented with a positive (ea, “huppy") or negative (eg, "sadness" cue and then instructed totypea brie, one sentence description of a specific pastor future ‘event, They wore also instructed to rate the emotion of the event (C-What isthe emotion of ths event?") on a S-point seale, ranging from 0 (very negative) to 4 (very positive); the likelihood of the ‘event (“How likely is an event lke this to occur in your life in the future?) on a S-point scale, ranging from 0 (very unlikely) to 4 ikely); and the difficulty of accessing the event (How difficult, was ito think of this event”) on a S-point scale, ranging from 0 (not at al 104 (a great deal), PSS-10 After the event tisk, participants completed the PSS-10. The PSS-I0 is a widely used self-report questionnaire that measures the degree to which participants perceive their lives during the last. month to be unpredictable, uncontrollable, and overloading. The {questionnaire has ten items and participants rate the frequency of siressiul events on a 4:point scale ranging from 0 (never) to 4 (very often). The PSS-10 has adequate reliability (2 = .78; Cohen, 1988) and convergent validity (Mitchell et al, 2008). Scoring The typed event descriptions were scored as “I” if they const tuted a specific event or as “0” if they didnot. Events that aceumred in one location and lasted for less than one day (eg, “Going to a BBQ last June with my friend Alice”) were scored as specific. A ther responses were scored as nonspecific, Two Filers were Irtined ‘with a randomly selected st of 54 past events from an unrelated ‘study. Interratr reliability was assessed with the intraclass correla tion coefficient (CC) of 80. For the future event trained with a randomly selected set of 48 future events from an ‘unrelated study and had an ICC of 86, We caleulated the propor tion of specific events (event specilicity) generated foreach partic pant foreach time period and cue emotion condition, to rlers were Analysis, All data was analyzed using R statistical sofware (R Core ‘Team, 2017). A series of linear mixed effects models (Baayen et 1, 2008) were conducted with the tme# library (Version 1-1-17; Bates et al., 2014). Post hoe tests using the Tukey correction wer conducted with the emmeans library (Version 1.6.1; Searle etal, 1980), Independent variables of cue emotion (positive, negative), time period (pis, future), seore on the PSS-10, and their interac tions, were run on the dependent variables of imerest (Le. event specificity, likelihood rating, dificulty rating, emotion rating) sith participant and experimental yersion entered as random effects. LALLA AND SHELDON Results, Event Specificity “The linear mined-ffets move preictne propoton of specie eens from cue emton, ime perio, and PSS-10 (see Table 1), revealed a vend toward sizmeat main effect of PSS-10 (Bh “029, p =.058). A sinitisant main effect of time period in the ‘negative direction (i = ~.12, p < .001) suggested that event speci- fcty was lower fr the future time period than the past. izi= cand intraction Intwoon tw prior nal cw emo, wes ab ‘dented ()= 18, p = 0 se Figure 1). Pest oe eompaions revealed tha this was due toa significant diference between pos tive ad negative future events, 1,257) = 85, p< OM No si nificant dfereace between poniive and negative past evens Was ‘ented, (1,257) = 27, p = 79. Patipants were severely ‘mpaied onthe Fate even! tsk particule when the cue emo tiom was negative a compared to ther performance othe past event xk Likelihood Rating ‘The linear mixed-effects model predicting likelihood rating from cue emotion, time period, and PSS-10 (see Table 2) revealed a significant main effect of PSS-10 score in the negative direction (f= -.11, p < 01), a significant main effect of time Period in the negative direction (B=~27, p< 001), and a sig nificant main effeet of cue emotion in the negative direction (B ~ 53, p <.001). These main effects suggest that lower ratings of likelihood were associated with higher scores on the PSS-10, that future events were rated as less likely to occur, and that neg- ative events were rated as less likely to occur, respectively. We ‘Table 1 Results of Linear Mixed-Esfects Model Predicting the Proportion of Accessing Specific Events From the Score on the PSS-10, Time Period (PastFuture), Cue Emotion (Positive/Negative), and the Interaction Among Factors Propartion of specific evens Pretoria effect Fatima ? Predictor There ost <0 Pss-10 03. 058 ioe pesiod on <0 Cue emotion 0.00 8S SS-10 % Time Period ‘oot ma SS-10 % Cue Emotion 0.00 su ‘Time Period % Cue Emotion ors <0 SS-10 X Time Period % Cue oan, #91 ration Random eect * oor aa 03, {hee 000 ec O31 Naja 24 Nyce Observations 18 Marginal Conditional ‘04200395 ‘Nowe PSS-10= Perceived Stress Seale-10 be disseminated bron. ely for the personal use ofthe EMOTION, STRESS, AND MENTAL TIME TRAVEL 1395 Figure 1 The Average Proportion of Specific Event Responses to the Positive and Negative Cues for the Past (Blue Bars) and Future (Grey Bars) Time Periods 1.00: ‘ } s i E ‘Note. Standard eror bars ae shown. *** p < 001. See the online ani- cle forthe color version ofthis iar, ‘also identified a small but significant interaction between PSS-10 score and cue emotion (B = 12, p = 025; see Figure 2A) and a Significant interaction between time period and cue emotion (B=.17, p = O18; see Figure 2B). Post hoc comparisons indi- cued the latter interaction was driven by a significant difference in likelihood rating between positive past and future events, 1(1,229) = 5.20, p <.001. No significant difference in ikelinood bberween negative past and future events was identified, 1,231) = 1.85, p=.065. Difficulty Rating ‘The linear mixed effects model predicting difficulty rating of accessing an event from cue emotion, task, and PSS-10 (see Table 3) revealed a significant main effect of PSS-10 scote in the posi- tive direction (fb = 26, p< O01; see Figure 3A), a significant iain effect of time period in the positive direction (P= 32, p< £001), und a significant main effet of cue emotion in the positive direction (B = .19, p < O01). These main effects suggest that higher seores on the PSS-10 were associated with greater ratings (of difficulty, that future events were rated as more dificult to access than past events, and that negative events were also rated a8 ‘more difficult to access than positive events. We also identified ‘small, significant interaction between PSS-I0 score and cue emo- 0085; see Figure 3B). Emotion Rating ‘The linear mixed-effects model predicting the emotionality of accessed events from cue emotion, time period, and PSS-10 (see Table 4) revealed a small main effect of PSS-10 score in the ne: ative direction (B =~06, p = 026; see Figure 4A), a significant main cffcct of time period in the negative direction (B= 31, p< 001), and a significant main eflect of cue emotion in the negative direction (B = ~3.15, p <.001). These main effects sug- ges that higher scores on the PSS-10 were associated with more negative emotion ratings, that future events were rated more neg- atively han past events, and thal events generated in response 10 ‘a negative cue were rated as more negative than events generated in response to a positive cue, We also identified significant interaction between time period and cue emotion (B = .74, p< (001; see Figure 4B), Post hoc comparisons revealed that this ‘was due to a significant difference in emotionality rating between positive past and future events, (1,232) = 7.86, p< .0O1, such that past events were rated as more positive, as well a a signif cant difference in emotionality rating between negative past and future events, (1,239) = -10.56, p <.001, such that past events ‘were rated as more negative Discussion ‘The present study aimed to assess how individual differences in current life stress experienced during the COVID-19 pandemic affects mental time travel, focusing on how emotional events remembered from the past and imagined in the future are accessed and evaluated. In response 1o emotional cues, participants were asked to acoess specific past and Future personal events and then to rate these events on how likely they were to occur in the future, Table 2 Results of a Lincar Mixed-Effects Model Predicting the Likelihood of Accessing From the Score on the PSS-10, Time Period (Past/Future), Cue Emotion (Postive/Negative), and the Interaction Among Factors Preditorfand effect Exinate ? Predicor There <0 Pss-10 “008 ‘Tew peiod Time Period Cue Emotion ‘cognitively demanding than remembering past experiences (Addis Random effect tal, 2007), Ascot the constructive episaiesimutation y= on othesis, whereas both forms of mental time travel require epi it Pothesis, whereas both forms of’ mental # require P= cenn oor sodic memory 10 construct mental representations of events, 1¢C) 038 Imagining specific future events places greater demands on these Nsajauan a ‘memory processes. Unlike past events that ean be constructed — _Nezwn 2 from existing memory traces, imagining future events requires Observations Mexibly recombining details Grom disparate aulobiographical —_Mainal Conditional RE events in novel ways (Addis et al, 2007; Schacter & Addis, Nowe PSS-10= Peroived Stress Seale10 ely forthe personal vse of 3 eis 5 : i z i This EMOTION, STRESS, AND MENTAL TIME TRAVEL Figure 1397 A: Average Difficulty Rating for Past (Blue Bar) and Future (Grey Bar) Time Periods): B: Average Difficulty Rating in Response to Positive (Green Line) and Negative (Blue Line) by PSS-10 Score A) Dithcuty Rating Past Future ‘Time Period B) Ditheuty Rating ° oi Pss.t0 Note, PSS-10 = Psroived Stress Scale-10, Stand cor hars re shown,“ p = 001, Seo the anne ate forthe color version ofthis ire, accessed an equ number of specific positive an negative past events. This positivity bias for future events is consistent sith prior reports (Newby-Clark & Ross, 2003; Rasmussen & Berisen, 2013) and likely serves an adaptive function. Imagining positive Tuture evens nets as a motivator of behavior and eneouraging one to plan for specific, anticipated events (Taylor & Brown, 1988), aligning with reports that imagining postive Future events is linked to increased well-being (MacLeod & Conway, 2007), and decreased hopelessness and depression (MacLeod et al, 2005; MacLeod & Byme, 1996; Quoidbach et al. 2008) ‘A novel finding from our study was that emotional cues affected ‘how participants evaluated the lkelthood that a recalled or imag jned event would occur in dhe future und the difficulty of accessing these evens. We found positive events from both past and future time periods were judged as more likely to oeeur inthe Future and ‘were more easily accessed than negative events. There ate a few possible explanations for this finding. According to prior work, the presence of positive emotion has a broadening elfet on cog tive processing, enhancing the ability to access related content to presented information (Ashby et al, 1999; Clore & Muntsinger, 2007; Madan et al., 2019). Thus, positive eue words may have activated more content than the negative cue words, decreasing the difficulty of idemttying a specific episode and increasing per- ceptions of the event as likely. Another possibility is that this effect is related to fading afect bias, whereby negative cognitions fand perhaps also negative autobiographical information become inaccessible more rapidly than positive cognitions; a bias that is important for an individual's wellbeing (Adler & Pansky, 2020; Szpunaret al, 2012), The lading of details from these neg: tively valunced cognitions may thus lower likelihood estimates and raise difficulty ratings for negative events. Finally, partic pants may have consciously suppressed access t0 negative events inthe interest of emotional regulation (Benoit eta, 2016). Taken together, we suggest that our findings ofa positivity bias for future ‘event specificity, alongside the finding that positive events ar eas Jer to access than negative events and are rated as more likely 10 happen reganiless of event time period, illustrate the directive role ‘of postive events in imagining outcomes ‘Tabled Results of a Linear Mixed-Effects Model Predicting How Emotional Participants Rated Accessed Events From the Score ‘om the PSS-10, Time Period (Pasw Future), Cue Emotion (Positive/Negative) and the Interaction Among Factors mation rating Prodistorandom effect Estimate > Ted Tnereept 388 <0 Pss.10 ‘06 025 ‘Time pesiod 031

You might also like