ListenLarger documents may require additional load time.
Exploring "rosy" bias and goal achievement in marathon runners
Journal of Sport Behavior. 36.1 (Mar. 2013): p66.

The present study investigated runners' predictions before participating in a marathon, their mid-race experiences, and their post-race recollections. Specifically, we examined whether runners exhibited "rosy" biases, in which they expected to feel better--and remembered feeling better--than they reported during the race itself. While pre-race expectations were not significantly different from mid-race reports, runners remembered feeling better than they actually reported feeling while the race was underway, illustrating a rosy retrospective bias. Participants 'intentions to run in a future marathon depended on whether they had achieved their pre-race goal, but also their temporal perspective. Immediately after the race, participants who had failed to meet their pre-race time goal reported less interest in running another marathon compared to people who had met their goal. However, four weeks later, when runners were fully recovered, people who had failed to meet their time goal reported greater desire to run another marathon than people who had met their goal. This crossover pattern may be explained by differences in psychological construal, with runners focusing on concrete experiences (e.g., pain or pleasure) immediately after the marathon, but focusing on abstract representations (e.g., goal achievement) when the past race and future races are temporally distant.

Full Text: 

"You have to forget your last marathon before you try another. Your mind can't know what's coming." Frank Shorter, U.S. Olympic Marathon Gold Medalist (Will-Weber, 2008)

Marathon running has become increasingly popular over the past 40 years. In the United States alone, there were approximately 507,000 marathon finishers across 625 different races in 2010 (Running USA, 2011). The vast majority of these runners are recreational athletes, including many first-time participants. Despite its growth in popularity, long-distance running is frequently accompanied by discomfort, fatigue, and injury during the training stage as well as during the race (van Gent et al., 2007). The frequency and severity of negative experiences reported by many runners makes one wonder why people choose to run marathons in the first place, and importantly, why they choose to run additional marathons.

A substantial body of literature demonstrates that people make systematic errors in predicting how they will feel during and after future experiences (e.g., Dunn & Latham, 2006; Gilbert, 2006), suggesting that marathon runners may miscalibrate their expectations about running a marathon prior to the actual event. For example, one study observed that participants preparing for a three-week bicycle trip predicted that the trip would be quite enjoyable. However, the bicyclists failed to anticipate the effect that long days of cycling, often in the rain with whiny companions, would have upon their enjoyment of the trip, and they reported substantially less positive evaluations of the trip while it was taking place than they had expected beforehand--an effect known as rosy prospective bias (Mitchell, Thompson, Peterson, & Cronk, 1997).

Similarly, to the extent that marathoners recall past running experiences when deciding whether to participate in future ones, they may recall the experience in a way that differs from their mid-race conception. The bicyclists noted previously, for example, recalled the experience as substantially more positive, once it was over, compared to their reports during the trip--an effect termed rosy retrospective bias (Mitchell et al., 1997). In another study, students on spring break vacations showed similar bias; moreover, students' desires to take similar vacations in the future were mediated by recollections that were substantially more positive than their actual experiences (Wirtz, Kruger, Scollon, & Diener, 2003). In the present research, we examined whether marathon runners exhibit rosy prospective and retrospective biases, as such errors in predicting how they will feel and remembering how they felt may illuminate decision-making about participation in future marathons.

While marathoners may exhibit rosy biases, marathon running also is distinct from bicycle holidays (Mitchell et al., 1997) or spring break vacations (Wirtz et al., 2003) because of its goal-directed focus. Marathoners frequently cite goal achievement, such as trying to run faster, trying to beat a certain time, or improving upon their previous performance as a primary reason for their participation (Havenar & Lochbaum, 2007; Masters & Ogles, 1995; Ogles & Masters, 2003). Attainment of one's goal also is cited as a reason for not wanting to run in a future marathon, as indicated in a survey of runners in the 1980 Melbourne marathon. Among runners who did not intend to run another marathon, the most frequent reason given was "will have achieved my goal" (Summers, Machin, & Sargent, 1983). Thus, examining runners' attainment or non-attainment of their goals is likely one key to understanding why, upon completing one marathon event, runners choose to undertake additional ones. In the present study, we assessed runners' goals prior to the marathon, determined whether they achieved their goal, and recorded their post-race intention to run another marathon.

Assessing Prospective and Retrospective Biases in Marathon Runners

No prior research on marathon running has included mid-race assessments, presumably due to the impracticality of administering survey measures to people engaging in an intense physical activity outside of a laboratory environment. Yet, ecological assessments of experience during the episode itself are necessary to evaluate the presence of biases in prospection or recollection. In the present research, we devised a method to question participants orally as they were participating in the marathon. We questioned runners about their experience at two time points during the marathon, in addition to surveying them before and after the race. By comparing predictions and recollections with actual experiences, we were able to examine our hypotheses about rosy prospective and retrospective biases (Mitchell et al., 1997).

While the pattern of rosy prospective bias predicts that runners will expect to feel better than they actually end up feeling during the race, this type of overestimation may be unlikely in a marathon, given that there are important reasons why marathon runners might temper their expectations. A survey of 363 recreational marathoners found that the vast majority of runners expected the race to be difficult, with 52% expecting the race to be "moderately hard" and 44% expecting it to be "very hard" (Summers, Sargent, Levey, & Murray, 1982). Summers et al. (1983) similarly found that first-time marathoners anticipated that the race would be difficult, with a mean expectation of 5.64 on a 1-7 scale where 7 = "very hard." The pattern of rosy retrospective bias, on the other hand, predicts that runners will remember feeling better during the race than they actually did. Consistent with research by Mitchell et al. (1997) and Wirtz et al. (2003), we anticipated that participants' reports of their experiences after the race would be more positive than their mid-race reports.

Goal Achievement, Temporal Perspective, and Intention to Run Another Marathon

In addition to asking participants about their expectations prior to the race, we also asked them to report their goals. Based on prior surveys (Masters & Ogles, 1995, 2003; Summers et al. 1983), we anticipated that participants who did not meet their goals for the race (versus those who did) would express greater interest in running another marathon. To evaluate this hypothesis, we asked runners to report their time goals (if they had them); we then compared goal times to actual completion times to determine whether the participants did in fact run faster than their stated goal time. Participants also were questioned about their intention to run another marathon immediately following the race and again four weeks later.

While goal attainment (or non-attainment) is frequently cited by runners as a motivation, we posit that the relation between goal achievement and the desire to run in another marathon, after one's last, may vary as a function of time. Construal-level theory (CLT; Trope & Liberman, 2003, 2010) specifies that events that have greater psychological distance (such as separation in time) are represented more abstractly than events that are perceived as closer. For example, Liberman and Trope (1998) showed that participants thinking about an activity such as taking a test in the distant future were more likely to construe it in abstract terms ("showing one's knowledge"), whereas they were more likely to construe the event in concrete terms ("answering questions") when it was described as temporally close. Differences in construal level have implications for future planning: in another study, students indicated greater willingness to engage in an interesting but difficult academic activity when it was temporally distant than when it was temporally close.

If running a marathon follows this pattern of temporal differences in construal, then immediately following the race, runners are likely to construe the experience in concrete terms, such as focusing on their immediate physical discomfort. However, four weeks later runners are likely to generate a more abstract construal of the experience, such as focusing on their achievement goals. Consistent with this theory, we predicted that runners who failed to achieve their goal in the race (vs. those who achieved their goal) would report a greater desire to attempt a future marathon; however, we expected this effect to be stronger four weeks after the marathon--when the previous race and any future race would both be construed in abstract terms--than immediately after the marathon.



Runners and walkers were recruited from a marathon training group in Bellingham, WA. Ten women and seven men were registered for the Bellingham Bay Marathon in 2008 and 14 women and five men were registered for the race in 2009. Participants ranged in age from 22 to 62 (M = 41.72, SD = 10.76). Most of the participants (77%) were first-time marathoners (1). Among those who had previous marathon experience, the average number of races completed was M = 1.56 (SD = 1.33). To ensure that participation in the study would not interfere with marathon performance, we only recruited walkers or runners with an expected pace no faster than 8:30/mile. Participants in the 2008 sample were not compensated; participants in the 2009 sample were offered $10 for their participation (three declined payment).


Participants completed web-based surveys to assess their predictions and memories about the marathon. The pre-race survey asked two primary questions about how participants expected to feel: "During the marathon, how do you think you will feel physically" and "during the marathon, how do you think you will feel mentally/emotionally?" Responses were given on a Likert-type scale from 1 (very bad) to 10 (very good). These items were chosen to obtain a brief global assessment of runners' feelings that also could be easily administered during the race itself. A similar single-item scale with the same endpoints, the "feeling scale," has been shown to be a valid measure of affective responses during exercise (Hardy & Rejeski, 1989; Rejeski, Best, Griffith, & Kenney, 1987). In addition, participants were asked, "Are you planning to run another marathon after the Bellingham Bay marathon?" Participants responded to this question on a Likert-type scale from 1 (definitely no) to 10 (definitely yes). Participants also were asked to indicate whether they had a goal for the marathon and, if so, to describe it. For the 2008 sample, this question was asked in open-ended format. Because some 2008 participants did not indicate a time goal, for the 2009 sample, participants were asked specifically to indicate whether they had a time goal, and, if so, to specify what it was.

The immediate post-race survey included the same key questions as the pre-race survey, worded to ask about physical and mental experiences in the just-completed marathon, (e.g., "During the marathon, how did you feel physically overall?") as well as intention to run another marathon. Responses were provided using the same 1-10 Likert-type scales as in the pre-race survey. In addition, participants indicated whether they finished the race and, if so, to rate how they felt at the finish line on a Liken-type scale from 1 (very bad) to 10 (very good) (2). The four-week post-race survey was identical to the immediate post-race survey, with an additional question asking participants to rate how well they had recovered from the marathon on a scale from 1 (still have a long way to recover) to 10 (completely recovered).


Participants were recruited in person at group training meetings two to four weeks prior to the marathon. Potential participants were informed that the study was about "perceptions of running a marathon." Participants signed consent forms at the time they were recruited.

Two weeks prior to the marathon, participants were emailed a link to the pre-race survey and were asked to complete it within one week. In the pre-race email, participants were told that they would be approached at water stops near mile 16 and mile 23 in the race and that a researcher would run or walk alongside at the participants' pace to ask them questions. Water stops were chosen because people tend to slow down as they approach these locations, facilitating our ability to question the participants. These specific locations were chosen because many marathoners report "hitting the wall" at approximately mile 20 in the race (Buman, Omli, Giacobbi, & Brewer, 2008; Schuler & Langens, 2007; Summers et al., 1983); by questioning participants at mile 16 and mile 23, we were able collect one data point from just before and one from just after this point. Prior to the race, participants' race bibs were marked with a bright sticker so they could easily be identified on the course, and participants provided an estimate of their pace so the researchers would know approximately when they would arrive at each question point. This procedure allowed us to successfully identify every participant at both locations.

Because we were concerned about whether participants would be able to answer questions while running a marathon, we informed them ahead of time about the exact questions they would be asked and the corresponding response options (i.e., rate their current physical and mental feeling and indicate the extent to which they want to run another marathon on 1-10 scales). As participants passed by the designated locations, a research assistant jogged alongside and collected their responses.

Immediately following the race, participants were contacted and asked to complete the post-race survey within 24 hours. Seven participants returned their surveys via mail, and the remaining participants completed the survey online. All but one did so within 24 hours of the race; the final participant submitted the survey three days after the race. Twenty-eight days after the race, participants were emailed a link to the four-week follow-up survey and asked to complete it as soon as possible. All participants completed the final survey within seven days, except for one woman who was unable to be contacted. Participants' finishing times were obtained from the results posted on the marathon website. Finishing times were recorded as "chip" times, a measure of the time elapsed between when the racers crossed the start line and when they crossed the finish line, as noted by an electronic device attached to their shoes. Upon completion of the study, participants were debriefed about the hypotheses of the study.


All participants completed the marathon, with an average finish time of 5 hours and 13 minutes (SD = 0:58). Patterns of reported physical and mental feeling and interest in running another marathon were similar between the 2008 and 2009 samples. Year of participation was initially included as a variable in all of the reported analyses, but it produced no significant main effects or interactions with any of the relevant variables. Thus, all subsequent results are reported on the entire sample of 36 participants.

Rosy Bias in Physical and Mental Feeling

Ratings of physical and mental feeling were very strongly correlated at all time points (rs ranging from .71 to .82, all ps < .001), thus these responses were averaged into a single score for analysis. Participant ratings of physical and mental feeling before, during, and after the race varied considerably, as shown in Figure 1. A pattern indicating both rosy prospective and retrospective biases would be supported by the presence of a significant quadratic pattern, which tests for higher mean ratings in the pre-race and post-race responses relative to ratings during the race. A one-way repeated-measures ANOVA comparing ratings of physical and mental feeling at all time points showed significant differences among the ratings (3), F(2.66, 87.61) = 5.11, p = .004, partial [[eta].sub.2] =. 13. The predicted quadratic contrast also was significant, F(1, 33) = 5.65, p = .023, partial [[eta].sub.2] =. 15, consistent with an overall pattern of prospective and retrospective biases.

Rosy prospective bias would be indicated if participants' mid-race ratings were lower compared to their pre-race predictions. Pairwise comparisons using a Bonferroni adjustment showed that participants' reported feelings at mile 16 were slightly higher than their predictions, but this difference was not statistically significant, [t.sub.(33)] = -1.06, ns, Cohen's d = .18. Consistent with a pattern of rosy prospective bias, participants' ratings at mile 23 were lower than their pre-race predictions, but this difference also was not statistically significant, [t.sub.(33)] = 1.87, ns, d = .33. Rosy retrospective bias, in contrast, was supported by participants' significantly higher post-race recollections compared to their actual feeling at mile 23--both immediately after the race, [t.sub.(33)] = 3.02, p < .05, d = .53, and four weeks after the race, [t.sub.(33)] = 4.61, p < .05, d = .80. However, although post-race recollections were significantly higher than reports at mile 23, comparisons of recalled feeling with mile 16 yielded no significant differences, either immediately after the race, [t.sub.(33)] = 1.44, ns, d = .25, or four-weeks later, [t.sub.(33)] =. 13, ns, d =.02.

Pre-race and post-race questions assessed participants' expectations and recollections about the race overall, whereas reports at mile 16 and mile 23 assessed participants' feelings at specific points during the race. Thus, to provide an additional test of rosy prospective and retrospective biases, pre-race predictions and post-race recollections were compared to the average of the mid-race ratings. Predictions were not significantly different from mid-race ratings, [t.sub.(35)] = -.64, ns, d =. 11, nor were immediate post-race recollections, [t.sub.(35)] = 1.08, ns, d =. 18. However, four weeks after the race, participants showed evidence of rosy recollection bias, recalling that they felt better during the race than they actually did, according to their own reports, [t.sub.(33)] = 3.53, p <. 001, d = .61.


In summary, the strongest evidence for rosy prospective and retrospective bias results from comparisons that focus on mid-race reports at mile 23 of the race, when runners were feeling significantly worse than at mile 16, [t.sub.(33)] = 3.17, p < .05. This difference is consistent with the idea that participants may have "hit the wall" at some point between mile 16 and mile 23, an experience of extreme fatigue commonly reported among marathon runners late in the race (Buman et al., 2008; Schuler & Langens, 2007; Summers et al., 1983).

Goal Achievement, Temporal Perspective, and Intention to Run Another Marathon We suspected that participants' experience of achieving or not achieving their goal would be an important predictor of their intention to run another marathon in the future4. Twelve of 17 participants (70.5%) in the 2008 sample spontaneously provided a time goal when asked about their race goals in general, reporting an average goal time of 5:06. Participants in the 2009 sample were specifically asked to provide a time goal if they had one; 14 of 19 participants (74%) did so, with an average time goal of 4:56. Reported goals were compared to actual finish times, which varied across the full sample from 2 hours and 3 minutes slower than goal to 52 minutes faster than goal, with an average discrepancy of 12.6 minutes slower than goal.

Participants were divided into groups according to whether they achieved their time goal (equal to or faster than goal, n = 9) or did not achieve their goal (finished slower than goal, n = 17). Intentions to run another marathon immediately following the race and four weeks following the race were compared as a function of goal achievement using a 2 x 2 mixed-model ANOVA. While the analysis showed no main effects of either goal achievement, F(1,24) = 0.05, ns, or time of rating, F(1,24) = 0.80, ns, the predicted time x goal achievement interaction was observed, F(1,24) = 4.06, p = .055, partial [[eta].sup.2] =. 15. As shown in Figure 2, participants who had achieved their goal reported greater intentions to run another marathon immediately following the race, compared to participants who had failed to achieve their goal. However, four weeks later, this pattern reversed, with participants who had failed to achieve their goal reporting greater intentions to run another marathon than participants who had achieved their goal. Thus, the observed crossover pattern fits well with predictions derived from Construal Level Theory (Trope & Liberman, 2003, 2010).



Why do people run marathons? Further, once having completed a marathon, with its often grueling physical and emotional effects, what influences the intention to run another? We compared marathon runners' expectations prior to the event, and recollections following the race, with their reported physical and mental experience during the race. As predicted, runners in the Bellingham Bay Marathon showed evidence of a rosy retrospective bias for physical and mental feeling, recalling that they had felt significantly more favorable than their reported experience late in the race. In other words, following the race, runners' recollections of how they had felt during the marathon, physically and mentally, became inflated relative to the reports they had themselves provided less than a day earlier, during the event, and this effect was strengthened when they recalled the experience again four weeks later.

In contrast, we found less convincing evidence for rosy prospective bias. Participants felt worse than expected at mile 23, consistent with our predictions, but this effect was not significant. While our small sample size may have contributed to the lack of statistical significance, the lack of a stronger rosy prospective bias is not entirely surprising, given the unique nature of marathon running. Previous research that has shown evidence of rosy prospective bias focused on upcoming events that most people would expect to be pleasant, such as vacations (Mitchell et al, 1997; Wirtz et al., 2003). The marathon, in contrast, is well-known for its challenge and difficulty. Further, the participants for this study were recruited from an organized marathon training group; presentations by team leaders may have provided even first time marathoners with a fairly accurate expectation about what the race would feel like. Runners who have only recently decided to begin training for a race, a process that typically begins months in advance, may be more likely to show a rosy prospective bias compared to people who have largely completed their training, particularly if they have been training through an organized group with experienced leaders. It also may be the case that expectations for the race become less favorable as runners' training increases and the race gets closer.

Goal Achievement and Intention to Run Another Marathon

Knowing that marathon runners are often motivated by goal achievement, particularly regarding time goals (e.g., Masters & Ogles, 1995; Ogles & Masters, 2003), we anticipated that participants who had failed to achieve their time goal would report greater interest in running another marathon compared to people who had achieved their goal, and that this effect would be amplified when questioned four weeks after the marathon, compared to immediately after the race. Consistent with this prediction, we found a cross-over interaction which indicated that, immediately following the race, people who had achieved their time goal reported greater interest in running another marathon compared to people who had not, but four weeks later people who had not achieved their time goal reported greater interest in running another marathon compared to people who had.

This shift in intentions to run another marathon across time is consistent with construal-level theory (Trope & Liberman, 2010), which specifies that psychologically proximal events will be construed in more concrete, low-level terms whereas distant events will be construed in more abstract, high-level terms. Immediately following a marathon, for instance, participants may construe the experience in terms of their current feelings. Four weeks later, the physical and mental feelings experienced during the race are a distant memory and the race can be more easily construed in high-level, abstract terms such as success or failure of goal achievement. Supporting this notion, participants who failed to achieve their goal reported feeling significantly worse at the finish line (M= 5.88, SD = 2.08) compared to people who achieved their goal (M = 8.00, SD = 2.34; [t.sub.(24)] = 2.35, p < .05). The group who did not achieve their goal may have had concrete thoughts such as, "I feel horrible right now," leading to a relatively lower desire to run another marathon, compared with those who achieved their goal. In contrast, the group who did achieve their goals may have had concrete thoughts such as "I feel pretty good right now!"

Four weeks after the race, participants were asked how well they thought they had recovered, on a scale from 1 to 10, with higher scores indicating greater recovery. All participants reported high levels of recovery, with similar responses for people who had not achieved their race goal (M = 9.29, SD = 1.57) as for people who had achieved their goal (M = 9.67, SD = .71; [t.sub.(24)] =.67, ns). At this point, both groups would be able to construe the previous race in abstract, high-level terms such as, "I had been hoping to finish the race in under five hours, but I didn't quite make it," or "A sub-five-hour marathon is one item I can scratch off my bucket list!" Similarly, any future race is chronologically and psychologically distant and is likely to be construed in high-level terms that focus, for example, on the satisfaction that might be obtained from improving upon one's time. Unfortunately, we did not directly assess participants' construal of the race experience, so these examples are speculative and will need to be explored more thoroughly in future research. However, one of the study participants, who did not meet her time goal, spontaneously expressed high-level construal in an open-ended comment on the four-week post-race survey, noting, "I learned a lot from my first marathon, therefore I'm determined to do better at my next marathon!"


While the present study demonstrates that marathon runners exhibit a rosy retrospective bias, and that the effect of goal achievement on the intention to run another marathon depends on the amount of time elapsed since completing the race, we also must describe several limitations of the present research. First, our solution to the challenge of assessing runners' experiences during a marathon--questioning them at water stops--could potentially affect participant responses. To ensure that participants would be able to answer our questions while running, we informed them ahead of time where they would be questioned and what they would be asked. This allowed us to successfully locate and question every runner, but it also introduced the potential for participants to consider how they would respond prior to being asked. This method also potentially confounds temporal perspective with other factors (e.g., mid-race questions were posed verbally by research assistants while pre- and post-race questions were completed via internet surveys). The mid-race method of questioning participants may introduce demand characteristics, such as participants trying to impress us with their responses or trying to anticipate what we wanted them to say (e.g., Orne, 1962). However, it is worth noting that participants were not informed about any of our hypotheses prior to the race. Similarly, questioning participants at water stops introduces the possibility that their responses could be affected by the (presumably welcome) availability of water and an external excuse to slow down. Due to the difficulties of questioning runners during a marathon, we were only able to assess runners' mid-race experience based on two time points, at miles 16 and 23 of the race. Thus, our data are limited with regard to other points during the race. It is possible, for instance, that prospective and retrospective biases may be more or less pronounced depending on which portions of the race are assessed. Indeed, we found the strongest evidence for rosy bias when we compared predictions and recollections to participants' feelings at mile 23, a particularly difficult period of the marathon that may be hard to anticipate even with adequate training. Similarly, we asked runners to recall and summarize their global experience during the marathon; yet, we do not necessarily know which moments of the marathon were integrated into participants' recollections, or how different moments were weighted (cf., Fredrickson & Kahneman, 1993; Redelmeier & Kahneman, 1996).

Second, participants' reports of physical and mental experience could be separated in future research into positive and negative components (e.g., Wirtz et al., 2003). This would allow researchers to examine how positive and negative affect, for example, might be differentially weighted when participants globally summarize the experience of running a marathon and decide whether to run another. Both of these limitations reflect the challenges of assessing runners' experiences during an actual marathon: to avoid interfering with runners' performance, it was desirable to question them infrequently, at only two points during the race (which coincided with water stops) and to ask a limited set of questions that could be answered quickly and easily. Yet, the assessment of experience during an actual marathon, even at only two time points, is also one of the strengths of the present investigation relative to past ones (e.g., Schuler & Langens, 2007; Summers et al., 1982, 1983; Wirtz & Kruger, 2002).

Third, though our results indicate that the intention to run another marathon depends on both whether participants achieved their goal and--consistent with construal-level theory (Trope & Lieberman, 2010)--the time elapsed since completing the marathon, we did not directly measure how concretely versus abstractly participants construed their experience. Thus, our evidence for this conclusion is indirect, and we were only able to infer the thoughts that runners may have had in the aftermath of the event.


We began this article with the words of Olympian Frank Shorter, who suggested that the desire to run a marathon, particularly another marathon after one's first, depends on the forgetting of its harsh physical and mental toll. The present study offers several insights that modify this assertion. Rather than being forgotten, memories of marathon running are instead subject to a rosy bias in recollection that makes the experience seem more positive, compared with self-reports during the race. Further, goal achievement was one key to understanding runners' intentions to participate in a marathon again. As the time from their last marathon increased, runners who failed to achieve their goal became more motivated to run again, while those who achieved their goal became less motivated to run again.

Sport psychologists and trainers can benefit from taking into account the effect of perspective (prospection versus retrospection) and temporal construal when working with individuals preparing for, running in, and considering repeating a marathon. For example, while concrete cognitions--such as those related to one's physical or mental state--are likely to be prevalent immediately before or after a marathon, abstract cognitions--such as those related to an individual's goals--may predominate when the event is more distant. Similarly, we expect athletes at all levels of experience, from novice to advanced, to be subject to the bias we observed (cf., van Dijk, Finkenauer, & Pollmann, 2008). Thus, we believe integrating these considerations into the typically lengthy training regimens associated with marathon running could better inform both recreational and advanced athletes.

Author Note

This research was supported by a grant from the Office of Research and Sponsored Programs at Western Washington University. Portions of the research described here were presented at the annual meeting of the Society for Personality and Social Psychology, January 2010.


Buman, M. P., Omli, J. W., Giacobbi, P. R., & Brewer, B. W. (2008). Experienced and coping responses of "hitting the wall" for recreational marathon runners. Journal of Applied Sport Psychology, 20, 282-300.

Dunn, E. W., & Latham, S. M. (2006). Affective forecasting: A user's guide to emotional time travel. In J. P. Forgas (Ed.), Affect in social thinking and behavior (pp. 177-193). New York: Psychology Press.

Fredrickson, B. L., & Kahneman, D. (1993). Duration neglect in retrospective evaluations of affective episodes. Journal of Personality and Social Psychology, 65, 45-55.

Gilbert, D. (2006). Stumbling on happiness. New York: Knopf.

Hardy, C. J., & Rejeski, W. J. (1989). Not what, but how one feels: The measurement of affect during exercise. Journal of Sport and Exercise Psychology, 11, 304-317.

Havenar, J., & Lochbaum, M. (2007). Differences in participation motives of first-time marathon finishers and pre-race dropouts. Journal of Sport Behavior, 30, 270-279.

Liberman, N., & Trope, Y. (1998). The effect of temporal distance on level of construal. Journal of Personality and Social Psychology, 75, 5-18.

Masters, K. S., & Ogles, B. M. (1995). An investigation of the different motivations of marathon runners with varying degrees of experience. Journal of Sport Behavior, 18, 69-79.

Mitchell, T. R., Thompson, L., Peterson, E., & Cronk, R. (1997). Temporal adjustments in the evaluation of events: The "rosy" view. Journal of Experimental Social Psychology, 33, 421-448.

Ogles, B. M., & Masters, K. S. (2003). A typology of marathon runners based on cluster analysis of motivations. Journal of Sport Behavior, 26, 69-84.

Orne, M. T. (1962). On the social psychology of the psychological experiment: With particular reference to demand characteristics and their implications. American Psychologist, 17, 776-783.

Redelmeier, D. A., & Kahneman, D. (1996). Patients' memories of painful medical treatments: Real-time and retrospective evaluations of two minimally invasive procedures. Pain, 66, 3-8.

Rejeski, W. J., Best, D. L., (Griffith, P., & Kenney, E. (1987). Sex-role orientation and the responses of men to exercise stress. Research Quarterly for Exercise and Sport, 58, 260-264.

Running USA (2010). 2010 Marathon, half-marathon and state of the sport reports. Downloaded from on June 17, 2011.

Schuler, J., & Langens, T. A. (2007). Psychological crisis in a marathon and the buffering effects of self-verbalizations. Journal of Applied Social Psychology, 37, 2319-2344.

Summers, J. J., Machin, V. J., & Sargent, G. I. (1983). Psychosocial factors related to marathon running. Journal of Sport Psychology, 5, 314-331.

Summers, J. J., Sargent, G. I., Levey, M. J., & Murray, K. D. (1982). Middle-aged, non-elite marathon runners: A profile. Perceptual and Motor Skills, 54, 963-969.

Trope, Y, & Liberman, N. (2003). Temporal construal. Psychological Review, 110, 403-421.

Trope, Y, & Liberman, N. (2010). Construal-level theory of psychological distance. Psychological Review, 117, 440-463.

van Dijk, W. W., Finkenauer, C., & Pollmann, M. (2008). The misprediction of emotions in track athletics: is experience the teacher of all things? Basic and Applied Social Psychology, 30, 369-376.

van Gent, R. N., Siem, D., van Middelkoop, M., van Os, A. G., Bierma-Zeinstra, S. M. A., & Koes, B. W. (2007). Incidence and determinants of lower extremity running injuries in long distance runners: a systematic review. British Journal of Sports Medicine, 41, 469-480.

Will-Weber, M. (Ed.). (2008). The quotable runner: Great moments of wisdom, inspiration, wrongheadedness, and humor. Halcottsville, NY: Breakaway Books.

Wirtz, D., & Kruger, J. (2002). [Predicted, on-line, and remembered experience of marathoners]. Unpublished raw data.

Wirtz, D., Kruger, J., Scollon, C. N., & Diener, E. (2003). What to do on spring break? The role of predicted, on-line, and remembered experience in future choice. Psychological Science, 520-524.

Kristi M. Lemm

Western Washington University

Derrick Wirtz

East Carolina University


(1) The number of participants with previous marathon experience (n = 8) was too small to test for differences between experienced and first-time marathons.

(2) The question regarding how participants felt at the finish line did not specify physical or mental feeling.

(3) Degrees of freedom and corresponding p-values for this analysis and all other one-way repeated measures ANOVAs reflect a Greenhouse-Geisser adjustment, necessary because the covariance matrix did not meet the assumption of sphericity.

(4) Participants' reported intention to run another marathon did not differ significantly across the five time points, F(2.92,96.46) = 1.99, ns. Participants reported moderately high interest in running another marathon prior to the race (M = 6.56, SD = 2.95), at mile 16 (M = 7.60, SD = 2.46), at mile 23 (M = 6.96, SD = 2.90), immediately after the race (M = 7.44, SD = 2.78), and four weeks later (M = 7.20, SD = 3.08). Thus, participants maintained a relatively high level of interest in running another marathon, even at points during the race when they were feeling their lowest physically and mentally.

Address correspondence to: Kristi Lemm, Department of Psychology, Western Washington University, 516 High Street, Bellingham, WA 98225-9172. E-mail:

Source Citation   (MLA 8th Edition)
Lemm, Kristi M., and Derrick Wirtz. "Exploring 'rosy' bias and goal achievement in marathon runners." Journal of Sport Behavior, vol. 36, no. 1, 2013, p. 66+. Expanded Academic ASAP, Accessed 25 Sept. 2017.

Gale Document Number: GALE|A321334820