Trends Suggest a Potential Effect of Episodic Recall Primes on Ratings of Task Confidence

Daniel Baldini, advised by Tyler Livingston


The present research used a between-participants experimental design to examine the effect of episodic priming on self-perceived task confidence. An analysis of the interaction between participant self-esteem and task type revealed a significant effect of self-esteem on a job task, but not on an athletic task. Self-esteem was positively associated with confidence in job-related tasks. Differences in self-perceived task confidence in job and athletic contexts trended toward significance for participants primed with thoughts of success versus failure. Because these relationships were statistically nonsignificant, more research is needed to demonstrate a relationship. Other nonsignificant trends emerged related to gender: Men reported more confidence on athletic tasks versus job tasks, whereas women reported more confidence on job tasks versus athletic tasks. Based on these findings and theoretical analysis, we suggest that corporations and athletic teams can use episodic priming tasks to achieve maximum success in relevant aspects of performance. If future projects show that women are more confident than men in job tasks, corporations could implement programs that aim to increase the confidence of male employees. Similarly, extensions of the present findings could reveal coaching strategies to enhance confidence among women athletes.

1 Introduction

Athletics is shifting gradually to a mentally focused field. Coaches and athletes alike are putting more emphasis on the mental side of athletics, with various research being conducted on strategies for success on the mental side of the job. Some coaches have researched connective mechanisms that build a sort of pyramid of successful team qualities [5]. Social psychologists have investigated other psychological variables that can affect athletic performance, including coaches’ biases [8] and perhaps players’ self-affirmations [3] (see McQueen & Klein, 2006 for meta-analysis). These studies illustrate the importance of empirical psychological research to test the effects of social variables on team members’ performance. With the sports market having grown to about 68 billion dollars annually, an increase in pressure for athletes to perform is inevitably taking place. Therefore, athletes find it ever more critical to succeed in the mental aspect of their performance, not just the physical [5]. This increased interest in the psychological aspect of career performance makes research regarding the psychology of sports and success on job performance a field that is becoming increasingly relevant. Studies have shown that affirmations are a helpful motivation tactic [7]. The research indicates that self-affirmations covering a wide spectrum of styles are necessary for success. Sufficient research has been done to predict outcomes but not enough to draw any steadfast relationships. Through meta-analysis, we have found that there are both studies involving affirmations on task confidence in sports and task confidence in jobrelated environments. However, the variables exclusively investigate the effects between self-affirmation practices and either sports or job-related tasks [3]; [7]. Studies have failed to investigate interacting aspects of self-affirmation and its effects between job and sports-related tasks and the interaction between these tasks and self-esteem. There have also been no studies that look at gender differences when participants are presented with confidence measures. The current study aims to address this gap in research by looking at the comprehensive interactions between selfesteem, athletic confidence, workplace confidence, and gender. Finally, previous research has shown that episodic memory recall primes are effective in experimental designs measuring task performance. These primes involve prompting participants to recall and write about a specific type of memory from their past, such as a time you felt powerful (to prime feelings of power; [2]) or your most arousing sexual experience (to prime sexual arousal; [6]. After writing about these memories, participants respond to relevant dependent variables that, theoretically, should vary as a function of the episodic recall prime. In certain tasks, math problems were done at a greater level when primed to perform free-recall of positive affirmation-based events (Schimel et al., 2004). In the current study, we used this method of episodic recall priming to prime concepts related to success or failure. The purpose of the present study is to expand upon each aspect of the preliminary literature review. By combining methods used in previous studies and incorporating them in a unified project designed to compare the variables discussed, we plan on investigating the effects of affirmations between participants. Based on the reviewed literature, we present two hypotheses. The first hypothesis was that participants in the “success” condition would have a higher average confidence level in both the job and athletic tasks than participants in the “fail” condition. The second hypothesis was that participants indicating higher self-esteem would present with higher confidence overall, thus demonstrating a positive relationship between indicated self-esteem and task confidence.

2 Method

The research used a two-group between-participants experimental design. The Qualtrics online survey distribution platform randomly assigned participants to one of two priming conditions. Participants subsequently reported their self-perceived level of confidence in their ability to complete each of two separate tasks. We analyzed these dependent measures separately using one-way ANOVA models. Thirty-nine Amazon Mechanical Turk (MTurk) workers completed the 10-minute online survey. Each worker received 1 dollar in exchange for participating. One participant was excluded from analyses due to an incomplete survey. Twenty-two women and 17 men completed the survey and were included in the reported analyses. The mean age of the participants was 42.9 years. The survey included an informed consent agreement, 10-item self-esteem scale (Rosenburg, 1989; see Table 1), demographics, an episodic recall task priming either success or failure (depending on experimental condition; see the Appendix), and finally, confidence ratings on both a job and athletic performance task (see Figure 1). MTurk workers participated in an online study designed to produce primed recall of past success or failure on either a job or athletic task. Each participant indicated their informed consent to participate. Participants then responded to the 10-item Rosenberg Scale, a self-esteem assessment that is regarded for its effectiveness and replicability (Trzesniewski, Robins, & Hendin, 2001). Once completed, participants responded to a demographic’s questionnaire and proceeded to the experimental portion of the survey. In the experimental portion of the survey, participants wrote a short essay response to a prompt asking them “to describe a time in which you succeeded in an athletic or job environment. By succeeded, we mean achieving something you desired or attained something that gave you satisfaction” (success condition) or “to describe a time in which you failed in an athletic or job environment. By failed, we mean falling short of achieving a desired outcome” (failure condition; see the Appendix for full manipulation text). In prior research, prompts involving affirmations and free recall have shown to be an effective method of producing behavioral results [7]. Eleven participants in the “success” condition wrote about a job-related success. Eight participants in the “success” condition wrote about an athletic-related success. Nine participants in the “failure” condition wrote about a job-related failure. Eleven participants in the “failure” condition wrote about an athletic-related failure. Thus, participants were relatively evenly distributed among experimental conditions. After the participants finished the experimental portion of the survey, they reported their self-perceived confidence on a scale from 0 percent to 100 percent on each of two different hypothetical tasks in random order (see Figure 1 for descriptions of the tasks). In one task, participants reported self-perceived confidence in successfully making a shot in table pong into one of the ten cups during their first attempt. We believed that this task provided an analogue to athletic performance. In the other task, the participant reported their self-perceived confidence in completing a task assigned to them by a supervisor at their workplace. We believed that this task is primarily based on the worker’s self-confidence in their ability to exceed demands in the workplace, therefore lining up well with the goal of creating an imaginary task based on confidence level on the job. After completing this final portion of the survey, each participant received a personalized code to receive monetary compensation.

3 Results

We conducted statistical analyses using R data analysis software and retrieved several tables of descriptive statistics using Excel software. Analyses based on our hypotheses and an added demographic analysis based on reported gender revealed three categories of relevant relationships, reviewed below. Category 1: The Rosenburg self-esteem scale was highly reliable (α = .94, ω = .96). Regression analyses revealed that self-esteem positively predicted confidence on the job task (b = 19.70, t(37) = 3.10, p = .004, R2 = .19). However, self-esteem did not predict confidence on the athletic task (b = 2.70, t(37) = .41, p = .004, R2 = -.02). (See Table 2) Category 2: Participants in the “success” condition reported nonsignificant higher confidence on the imaginary tasks than participants in the “failure” condition. Participants in the “success” condition had a mean confidence of 53.32 percent on the job task, whereas participants in the “failure” condition had a mean value of 41.50 percent. This pattern was also seen in the athletic task, on which participants in the “success” condition had a mean confidence of 57.74 percent, while the participants in the “failure” condition had a mean confidence of 50.9 percent. Although these trends were consistent with our hypotheses, they were nonsignificant (ps ¿ .23; see Figure 2A). Thus, more research is needed to demonstrate a relationship between these variables (see Table 3). Category 3: Participants were grouped in a secondary analysis by self-reported gender categories (man vs. woman) to examine gender differences on task confidence. On the athletic task, men tended to report greater confidence than did women: The mean confidence among men was 59.41 percent, whereas the mean among women was 50.23 percent (p = .33). This difference was nonsignificant, so we cannot conclude a relationship existed between gender and confidence among our participants (see Figure 2B). An analysis of the job task values also demonstrated an inverse relationship between gender and task confidence. On the job task, women tended to report greater confidence than did men: The mean confidence among women was 48.77 percent, whereas the mean confidence among men was 45.29 percent (p = .73). Again, we cannot conclude a relationship existed between these variables among our participants because the difference was nonsignificant (see Figure 2B). (see Table 4)

4 Discussion

Through a thorough analysis of the collected data, three different relationships can be interpreted to guide future research. An important conclusion to discuss is how selfesteem affects the different task types. The results show a significant positive relationship between reported selfesteem and confidence in the job task. However, there is an absence of this same relationship in the athletic task. These results could lead to a mechanism of improvement in motivation for employers to use when attempting to get more success out of employees. The current project supports previous findings showing that certain psychological variables can help improve athletic and workplace success [5]. If this were to be achieved, a useful step in that process would be for employers to develop strategies for improving their employees’ self-esteem. If this relationship is continuously replicated through future investigation, it will imply that improving self-esteem would raise employees’ confidence in their ability to complete tasks assigned to them within their job. Thus, leading to more productive employees. A different step that future research could take would be to look at the effects the other way around. If researchers want to grasp the interactions between self-esteem and success/failure entirely, they could design a study that looks at the effects of priming on self-esteem. If confirmation of the hypothesis that priming does affect self-esteem is achieved, having a mental health professional on staff at interested companies who specializes in resilience training after a significant failure during job tasks could help improve overall company performance and morale. Another important implication of the study to discuss is the absence of a relationship within athletics. While it was clear the relationship between job confidence and self-esteem existed, athletics did not see this clear correlation. This lack of correlation implies that individual confidence in athletics is separate from self-confidence and is a takeaway from this project that is worth future investigation. Researchers should attempt to replicate these results with a larger sample to help identify a mechanism that can increase self-confidence in athletic performance. A possible relationship could be that athletics affects self-esteem instead of self-esteem affecting athletics. Researchers in the future could prime participants to discuss failure or success in only athletic environments and then investigate that sort of priming and its effects on the self-esteem of the participants. This hypothesis would be that athletics affects the results of the Rosenburg scale. A future study based on environmental priming should also be performed. Due to the lack of specificity, there was no control for whether or not participants were primed with a job or athletic task; they were only controlled to be primed for “success” or “failure” conditions. It could be helpful to divide groups further. The proposed groups would be primed based on only athletic episodic memory of failure or success, or only job-related episodic memory of failure or success. The researchers could then analyze results between environmental groups. This level of specificity could be critical in increasing the effect of the primes. Consistent with The Theory of Planned Behavior, it is noted that specificity of intention allows for more perceived control over behavior, thus allowing for better prediction of behavior [1]. In specifying the behavior recalled, it would be hypothesized that the effect of the primes might be stronger, resulting in statistically significant data that otherwise would not have been produced. The next result to discuss is a suggested trend that the project found. The original aim of the research was to prime participants to recall either an episodic memory of success in athletics-related or job-related environments or a memory of failure in the respective environments. We hypothesized that the nature of the prime would affect the resulting confidence in imagined job and athletic tasks. The results showed a trend that suggested an effect. However, that trend was not significant. This information is still promising and can be used in future research. In both the athletic and jobrelated tasks, when participants were primed to retrieve memories of success, they did show higher confidence than the participants that retrieved memories of failure. The trending results gave promising information that should be further researched. We believe that if the experiment can be appropriately replicated with a larger sample size, the trending differences observed in the current study could reach statistical significance. If a future study were to reach significance through a larger sample size, this would add to the research on the effectiveness of using episodic recall priming when attempting to evoke certain attitudes and states of affirmation [3]; [7]). The same situation happened when investigating the confidence between gender and task. Gender as a variable produced a trend in the difference of confidence in tasks. Although the differences were nonsignificant, trends suggested that men felt more confident than women in the athletic task while women felt more confident in the job task than men. If researchers were to choose to stem off of these results, the best way to replicate it would be to do the same project with similar variables, but control for the equal participants in each gender group and increase the sample size so that a more appropriate power is reached. The hypothesis is that a higher power will produce significant data providing more clear evidence than the trending data in the present project. Practical use of this information could see employers’ focus on raising the confidence in their male employees and could work well in increasing the level of their companies’ performance. Researchers should primarily utilize this study as a guide for future investigations. A practical method in raising the confidence of both workers and athletes is a necessary venture, as mental health is progressing as an essential field for both success of companies and the well-being of their employees/ athletes. Using the information from the current project, behavioral researchers should attempt replication in a much larger sample size. Ideally, a sample size of 120, approximately 30 per group, as this demonstrated sufficient power in similar studies using essay-based recall primes [2], (groups proposed would be two gender groups and the primed recall groups, making it four total groups). It is hypothesized that the larger sample size would allow for more clear representations of data.

5 Appendix

Success condition

We would like you to describe a time in which you succeeded in an athletic or work environment. By succeeded, we mean achieving something you desired or attained something that gave you satisfaction. Please describe this situation in detail including what you did, how you felt, how others responded to your success, etc. Your response should be approximately 10 sentences.

The above prompt was used with the purpose of priming participants for episodic-memory recall of a success condition, in hypothesis that this priming would affect outcomes in a confidence-based imaginary task.

Failure condition

We would like you to describe a time in which you failed in an athletic or work environment. By failed, we mean falling short of achieving a desired outcome. Please describe this situation in detail including what you did, how you felt, how others responded to your failure, etc. Your response should be approximately 10 sentences.

The above prompt was used with the purpose of priming participants for episodic-memory recall of a failure condition, in hypothesis that this priming would affect outcomes in a confidence-based imaginary task.

Below is a list of statements dealing with your general feelings about yourself. Please indicate how strongly you agree or disagree with each statement.

Fig. 1: This table shows The Rosenburg self-esteem scale.

Note: The above Rosenburg Self-Esteem scale was analyzed as a moderator of the relationship between the confidence manipulation and performance on the imaginary task (Figure 3). Imagine you are playing the game depicted above. The object of this game is to throw a small, hollow ball, such as a ping pong ball, into one of the red cups. Each red cup contains a few ounces of water to prevent the cup from easily tipping over.

Fig. 2: Simple regression analyses testing associations between self-esteem (predictor variable) and self-rated confidence on athletic and job tasks.
Fig. 3: Two-way ANOVA testing effects of episodic recall priming (success vs. failure) on self-rated confidence on athletic and job tasks.

From 0 percent to 100 percent, what do you think are the chances that you could successfully shoot the ball into one of the cups on the first try?

Imagine you are in the meeting depicted above. Your boss, on the right, called the meeting to assign you a new job task. This kind of task typically takes three weeks to complete, but your boss forgot to assign the task to you earlier and the deadline is in 4 days.

From 0 percent to 100 percent, what do you think are the chances that you could successfully complete the task in 4 days? The above imaginary tasks were used to investigate participant’s confidence in athletic and work-related environments. The confidence measured was then compared between subjects.

Fig. 4: Two-way ANOVA testing effects of participant gender on self-rated confidence on athletic and job tasks.
Fig. 5: This image shows a depiction of throwing a ball into a cluster of cups.
Fig. 6: This image depicts two cartoon individuals having a meeting at a table.
Fig. 7: The above figure illustrates participants’ selfreported confidence on each task as a function of experimental condition (succeeded vs. failed).
Fig. 8: The above figure illustrates men’s and women’s self-reported confidence on each task.

[1] I. Ajzen. “The theory of planned behavior. Organizational Behavior and Human Decision Processes”. In: 50.2 (1999). Retrieved from 90020-T (1991), pp. 179–211.

[2] A. D. Galinsky, D. H. Gruenfeld, and J. C. Magee. “From power to action.” In: Journal of Personality and Social Psychology 85.3 (2003). Retrieved from (2003), pp. 453–366.

[3] A. McQueen and W. M. P. Klein. “Experimental manipulations of self-affirmation: A systematic review. Self and Identity”. In: 5.4 (2006). Retrieved from (2006), pp. 289–354.

[4] S. O’Leary and S. Brown. “Self-efficacy and the physiological stress response. In Self- efficacy, adaptation, and adjustment”. In: Springer (1995). Retrieved from 1- 4419- 6868- 5_8” (1995), pp. 227–246.

[5] D. Perez, S. Van Horn, and M. P. Otten. “Coach John Wooden’s pyramid of success: A comparison to the sport psychology literature.” In: International Journal of Sports Science and Coaching 9.1 (2014). Retrieved from 9541.9.1.85 (2014), pp. 85–101.

[6] P. R. Rerick, T. N. Livingston, and D. Davis. “Does the horny man think women want him too? Effects of sexual arousal on men’s perceptions of women’s sexual willingness.” In: The Journal of Social Psychology 160.4 (2019). Retrieved from 1080/00224545.2019.1692330 (2020).

[7] J. Schimel et al. “Not all self-affirmations were created equal: The cognitive and social benefits of affirming the intrinsic (vs. extrinsic) self. Social Cognition”. In: 22.1 (2004). Retrieved from 1521/soco. (2004).

[8] J. Weaver, J. F. Moses, and M. Snyder. “Self-fulfilling prophecies in ability settings.” In: The Journal of Social Psychology 156.2 (2016). Retrieved from https: / / doi . org / 10 . 1080 / 00224545 . 2015 . 1076761 (2015).