Through your partner’s eyes: Perspective taking tempers optimism in behavior predictions

Abstract

People tend to be overly optimistic when predicting their future behaviors. This research examines how taking someone else’s perspective affects predictions of relationship behaviors. Study 1 (N = 82) showed that taking the partner’s perspective when predicting how many relationship-enhancing behaviors one might perform over the next week reduced the number of predicted behaviors and consequently reduced optimistic bias. Study 2 (N = 244) replicated the reduction in predicted behaviors when taking the partner’s or a friend’s perspective. Study 2 also showed that predictions from another person’s view are similar to predictions for another person’s behavior. Study 3 (N = 149) replicated the reduction in predicted behaviors and the similarity to predictions for other people’s behavior. Furthermore, Study 3 suggests that one reason why adopting another’s perspective affects predictions is an attenuation of the link between forecasts and relationship quality and increase of the link of forecasts with conscientiousness, which tends to be a better predictor of behavior.

Keywords

Interpersonal relationships optimism perspective taking prediction pro-relational behavior

People are often called upon to estimate what they will do: People have to make predictions about their future behaviors when committing to a plan, setting up a meeting, when preparing for a vacation, or when promising to do something for a loved one. Predictions about one’s own behavior are often optimistically biased—across life domains, people predict they will accomplish more desirable behaviors than they actually do (Armor & Taylor, 1998; Dunning, 2007). Such optimism can be damaging, when biased predictions create expectations that cannot be met. For example, when a person’s behavior falls short of what he or she predicted and promised, his or her partner might trust him or her less and attribute the shortfall to a lack of care and love. Given the potential for negative consequences of inaccurate predictions—and the frequency of optimistic bias in behavior predictions—how can predictions of relationship behaviors be rendered more accurate? In three studies, we examine whether taking the perspective of another person while making a prediction can reduce the optimism in their prediction.

Optimism in predictions

Errors in prediction are common; people’s expectations are often overly optimistic (Armor & Taylor, 1998; Buehler, Griffin, & Ross, 1994; Dunning, 2007). Not only do people expect to finish tasks sooner than they really do, save more money than they really do, perform better on tests than they really do, they also predict to perform more positive acts for their partner than they really do. For example, when predicting specific behaviors they wanted to do for a romantic relationship partner over 2 weeks, participants only completed a fraction of the promised actions (Peetz & Kammrath, 2011). Such overestimation can have negative consequences. In the realm of relationship predictions, for example, the partner may perceive the broken promise as a relational transgression (Cameron, Ross, & Holmes, 2002; Metts, 1994; Siegert & Stamp, 1994); and seeing a partner fall short of their promises might be taken as an indicator of lack of motivation to make the partner happy and/or lack of effort (Fincham & Bradbury, 1993; Karney & Bradbury, 2000; McCullough et al., 1998; McCullough, Worthington Jr., & Rachal, 1997; Miller & Rempel, 2004).

Blind spots for self-predictions

Research suggests that the motivation to achieve a goal strongly influences predictions of goal-directed behaviors (Buehler, Griffin, & MacDonald, 1997; Koehler & Poon, 2006; Krizan & Windschitl, 2007; Peetz & Buehler, 2009), despite the fact that motivation alone is not a sufficient driver of many behaviors. For example, those who were more motivated to finish a task early predicted to finish earlier than other participants, but actually ended up finishing at a similar time (Buehler et al., 1997). Especially behaviors that require high self-control (such as temporally delayed actions) depend on people’s ability to self-control or their level of conscientiousness rather than their motivation to do the behaviors (Koehler & Poon, 2006; Peetz & Kammrath, 2011). Predictions about relationship behaviors may be particularly prone to be optimistically biased, if the motivation to make a partner happy blinds people as to what they will realistically be able to do. For example, promises in romantic relationships tend to be linked more with the motivation to make the partner happy and with relationship quality rather than actual predictors of follow-through behaviors, such as personal self-regulatory ability (Peetz & Kammrath, 2011). Thus, people tend to have blind spots when predicting what behaviors they will do for their loved ones, exhibiting optimistic bias in predictions about relationship behaviors (Peetz & Kammrath, 2011, 2013).

Predicting other’s behavior

The same blind spots are less likely to occur when generating expectations for other people’s behavior. People tend to have a more accurate understanding of what is possible (by relying on more diagnostic information such as ability to perform the behavior) when making predictions for other people rather than for the self (e.g., Epley & Dunning, 2006). Indeed, people tend to evaluate their own behavior according to different standards than others’ behaviors (Epley & Dunning, 2000; Kruger & Gilovich, 2004), tend to be less optimistically biased when making judgments about other people than about themselves (Buehler et al., 1994; Helzer & Dunning, 2012; MacDonald & Ross, 1999), and tend to recognize biases in other people’s judgments and behavior but not in their own (Miller & Ratner, 1998; Van Boven, White, Kamada, & Gilovich, 2003).

Predictions relevant to relationships have shown this reduction in bias in observer’s judgments as well. When forecasting how long a relationship would last, people tended to be overly optimistic about their own relationships—whereas their parents and roommates made relatively more accurate predictions (MacDonald & Ross, 1999). Similarly, a couple’s social network (e.g., friends) held more negative—and arguably, less optimistically inflated—perceptions about the relationship than the couple themselves (Agnew, Loving, & Drigotas, 2001).

One of the reasons why predictions for other’s behavior might be more accurate than self-predictions is that people are more likely to take self-regulatory ability and personality into account than when making predictions about others’ (vs. their own) behavior (Buehler et al., 1994; Newby-Clark, Ross, Buehler, Koehler, & Griffin, 2000; Peetz & Kammrath, 2013). For example, when participants predicted the likelihood of 14 pro-relational behaviors, their perceived relationship quality was strongly linked to the extent of predicted behaviors when predicting their own behavior but when predicting their partner’s behavior, predictions were linked more with their self-reported conscientiousness (Peetz & Kammrath, 2013).

It is not always possible to obtain an outside perspective on one’s own actions. Even if someone else might make a less biased prediction, people might not be willing or able to engage such a proxy in their forecasts. In the present studies, we examine whether the same benefits of prediction accuracy can be achieved when simply taking the perspective of someone else.

Perspective taking

Perspective taking might refer to imagining both what someone else feels (e.g., Batson, Early, & Salvarani, 1997) and what someone else thinks (e.g., Galinsky & Moskowitz, 2000). Forecasting from another person’s point of view refers to the latter, cognitive, form of perspective taking. The ability to take someone’s perspective has been linked to a number of prosocial benefits including increased social bonding (Galinsky, Ku, & Wang, 2005), more mimicry to improve social interactions (Chartrand & Bargh, 1999), and reduced stereotyping (Galinsky & Moskowitz, 2000).

In the context of romantic relationships, the ability to take a partner’s perspective has been linked to better relationship quality overall (Arriaga & Rusbult, 1998; Franzoi, Davis, & Young, 1985; Kimmes, Edwards, Wetchler, & Bercik, 2014; Long & Andrews, 1990) as well as specific benefits such as constructive responses during conflict (Rusbult, Verette, Whitney, Slovik, & Lipkus, 1991), forgiveness (Kimmes & Durtschi, 2016), and more positive and fewer negative relationship behaviors generally (Davis & Oathout, 1987). Individuals in romantic relationships reported performing behaviors from a list of 17 positive behaviors more frequently the greater their dispositional perspective taking and empathic concern, and their partners reported seeing these behaviors performed more often (Davis & Oathut, 1987). Participants were more likely to respond with relationship-maintaining responses when their partner received negative feedback if they were experiencing more empathy (Pinkus, Lockwood, Marshall, & Yoon, 2012). When married couples were instructed to take the perspective of their partner, they solved a hypothetical relationship conflict more constructively than when they took their own perspective or when they did not receive any instructions (Arriaga & Rusbult, 1998). We expect that perspective taking also benefits the accuracy of self-predictions. Making predictions about one’s own behavior from someone else’s perspective might reduce the optimistic bias, just as making predictions about others’ behaviors tends to be less optimistically biased.

Overview of current studies

We hypothesized that taking the perspective of someone else might reduce the optimistic bias that is often evident in people’s predictions about their own behaviors. In three studies, we instructed participants to take someone else’s perspective while forecasting their own relationship behaviors. In a first study, we examined whether predictions are less optimistic and more in line with actual behavior reports when predictions are made from a romantic partner’s perspective rather than from one’s own. In a second study, we extended the test of a perspective taking effect to examine how predictions differ when taking the perspective of a partner versus the perspective of a friend. The second and third studies also examined whether predictions from another’s point of view are more similar to predictions made for other people than predictions made for the self. The second and third studies also examined one possible underlying reason for the proposed reduction in predicted behaviors: We expected that participants’ predictions would be linked less with their perceived relationship quality and be linked more with their level of conscientiousness when adopting another person’s point of view, rather than their own. When taking someone else’s perspective, forecasters might look toward more objective predictors of behavior (conscientiousness) and be less biased by motivating influences (their positive feelings for the partner or relationship quality) than control forecasters.

Study 1

This study examined the predicted and actual frequency of 21 positive relationship-enhancing behaviors (Maxwell, Joel, Peetz, MacDonald, & Baucom, 2017) to assess optimism in pro-relational behavior predictions. We expected that predictions about future behaviors one might do for a partner over the course of a week would be overly optimistic; that is, participants would fall short of the predicted behaviors. However, predictions about relationship behaviors made from a partner’s perspective should be less optimistic and closer to actual behavior.

We also explored whether perceived relationship quality and conscientiousness were linked to predictions. Because our main objective in this first study was to examine the effect of perspective taking on behavior predictions, we measured conscientiousness, relationship quality, and other exploratory scales after the prediction task to avoid carryover effects.

Method

Participants

We recruited 196 participants from CrowdFlower (U.S. and Canadian participants only), an online crowdsourcing platform. We aimed for 50 complete data sets per condition, which would allow us to detect medium effects between two groups with 70% power (GPower). We overrecruited at Time 1 to account for possible attrition at Time 2 (i.e., incomplete data sets). We advertised for participants in committed relationships and excluded four participants who indicated that they were single at the time of the study. The final sample included 103 male and 89 female participants (M _age = 35.68, SD = 12.61) in romantic relationships (M _{relationship length} = 113.97 months [9 years and 6 months], SD = 116.47 [9 years and 8 months]). One hundred and thirty-eight participants consented to be contacted again a week later and 82 of these actually completed the follow-up part when they were contacted (M _age = 39.02, SD = 13.61; M _{relationship length} = 133.12 months [11 years, 1 month], SD = 124.22).

Procedure

After a brief demographic questionnaire assessing age, gender, relationship status, and length, participants completed a behavior prediction task, followed by a number of exploratory personality and attitude scales. A full copy of the survey is included in the Supplementary Materials.

Behavior prediction task

Participants were randomly assigned to a control condition or a perspective taking condition. In the control condition, instructions simply read “Think about next week. In the spaces provided, indicate whether you will complete the following behaviors for your partner.” In the perspective taking condition, participants first read:
Perspective taking is the imaginative tendency to put yourself in the place of another person in order to understand their point of view. For the purpose of the following survey put aside your own frame of mind and attempt to adopt the point of view of your partner
and were then instructed to “indicate whether your partner would think you will complete the following behaviors for your partner.” For each of 21 relationship behaviors, participants marked for each of the 7 days of the next week whether they would do the behavior that day or not. We added all behaviors, resulting in a predicted behavior score between 0 and 147 for each participant (21 behaviors × 7 days). Behaviors were simple, concrete, quantifiable (e.g., “I will tell my partner that I appreciate them”), and have been established to be widely perceived as positive, relationship-enhancing behaviors (Maxwell et al., 2017).

Exploratory scales

Participants also completed a number of exploratory personality and relationship attitude scales. Most pertinently, participants reported their perceived relationship quality (e.g., “How much do you love your partner?”) using a 6-item scale (a short form of the Perceived Relationship Quality Scale from Fletcher, Simpson, & Thomas, 2000; Cronbach’s α = .89) with response scales ranging from not at all* (1) to very much (7). They also reported their conscientiousness (e.g., “I am someone who…does a thorough job”), using a 9-item scale (Big Five Inventory [BFI] subscale for conscientiousness; John & Srivastava, 1999; Cronbach’s α = .86) with response scales ranging from disagree strongly (1) to agree strongly (7). They also completed the perspective taking subscale of the Interpersonal Reactivity Index (Davis, 1980). Participants responded to 7 items (e.g., “I try to look at everybody’s side of a disagreement before I make a decision,” α = .72) on a scale does not describe me well (1) to describes me very well (5). Additional information on these scales including validation information and scoring procedures can be found in the Supplemental Materials.

Behavior report

Participants who indicated they would be willing to be contacted received an e-mail 1 week later (and one reminder e-mail 24 hours after that) inviting them to sign on to a follow-up survey. Most of those 82 participants who completed the follow-up completed it either 7 or 8 days after the initial survey (n = 74, 89%), eight participants completed it 9 or 10 days later. Notably, those participants who actually completed the follow-up survey predicted to do more behaviors overall, t(179) = 4.46, p < .001, d = .67, were older, t(189) = 3.27, p = .001, d = .49, in longer relationships, t(189) = 1.97, p = .050, d = .29, and more likely to be female, χ²(1, N = 192) = 11.34, p = .001, d = .50, than participants who did not do the follow-up survey.

In the follow-up survey, participants were asked to indicate for each of the 21 relationship behaviors whether they had done the behavior on each day of the week after the initial survey. We added all 21 behaviors, resulting in an actual behavior score between 0 and 147 for each participant (21 × 7 days).

Results

Does perspective taking affect behavior predictions?

First, we examined the effect of perspective condition on predicted behaviors among everyone who completed both surveys. Participants who took the perspective of their partner predicted to perform significantly fewer behaviors in an upcoming week (M = 83.47, SD = 46.75) than participants who predicted their behaviors in the control condition (M = 101.87, SD = 35.19), t(81) = 2.05, p = .044, d = .45. This effect also held true when considering all participants regardless of whether they completed the follow-up survey (Ms = 68.07 vs. 87.33), t(180) = 2.91, p = .004, d = .43).

Does perspective taking affect behavior?

Next we examined whether perspective taking also affected the number of behaviors participants actually reported doing for their partner over the course of a week. Participants performed about as many behaviors for their partner in the perspective taking condition (M = 80.69, SD = 40.96) as in the control condition (M = 87.32, SD = 37.67), t(80) = .76, p = .450, d = .17. Consequently, while participants in the control condition significantly overestimated the number of behaviors they would do for a partner, t(46) = 3.54, p = .001, d = .39, participants in the perspective taking condition did not, t(34) = .75, p = .461, d = .11 (see Figure 1), suggesting that taking the partner’s perspective de-biased their prediction. Another way to examine the effect of condition on the bias in predicted behaviors is to control for actual behaviors in a multiple regression. When regressing predicted behaviors on condition while holding the number of actual behaviors constant, condition continued to have a marginal effect on predicted behaviors, β = −.15, t(79) = −1.80, p = .076, d = .41.

Figure 1.
Number of predicted and actual behaviors for those predicting their pro-relational behaviors from their own or from their partner’s perspective in Study 1.

Exploratory scales

We next examined participants’ trait perspective taking (M = 3.55, SD = .63), trait conscientiousness (M = 3.99, SD = .73), and perceived relationship quality (M = 5.82, SD = 1.12) as reported at the end of Session 1.¹ If participants took this information about themselves into account when making a prediction, these ratings should correlate with their predicted relationship behaviors. Before testing the link between relationship quality and conscientiousness with predicted behaviors, we examined whether they differed by condition (see Table 1 for mean values). The perspective taking manipulation condition did not affect participants’ reported ability to take someone else’s perspective, t(81) = .03, p = .980, d = .01, and their reported conscientiousness, t(81) = 1.14, p = .257, d = .25. However, participants in the perspective taking condition reported lower relationship quality than participants in the control condition, t(81) = 2.35, p = .021, d = .52.

Table 1.
A summary of means by conditions in Study 1.

Measured Variables Self-prediction about own behavior Self-prediction about own behavior, taking partner’s perspective

n 47 36

Positive feelings 6.07_a (.90) 5.50_b (1.29)

Conscientiousness 4.07_a (.68) 3.88_a (.79)

Perspective taking ability 3.55_a (.90) 3.54_a (1.29)

Mean of predicted relationship behaviors 101.87_a (35.19) 83.47_b (46.75)

Mean of actual relationship behaviors 87.32_a (37.67) 80.69_a (40.96)

Note. Means in one row with different subscripts are significantly different from each other at p < .05. Standard deviations are presented in parentheses.

Next, we examined the link between predictions and the exploratory scales. Participants rating themselves higher on conscientiousness also predicted more relationship behaviors, r(81) = .40. Participants reporting higher relationship quality also predicted they would do more behaviors for their partner, r(81) = .41. Participants rating themselves better able to take someone’s perspective also predicted they would do more behaviors for their partner, r(81) = .42. The same links were present for correlations with actual behaviors completed over the week of the experiment—however, these links were largely driven by the high correlation between predicted and actual behaviors, r(81) = .68 and were reduced to nonsignificance when predicted behaviors were controlled for in a partial correlation.

Finally, we examined whether the links between conscientiousness and perceived relationship quality with behavior predictions might differ by condition. We analyzed the data in multiple regressions. All variables were centered prior to analysis. Perspective condition (0 = control; 1 = perspective taking), conscientiousness, and relationship quality were entered on the first step of a linear regression, and their interaction terms were entered in the second step. The condition effect remained significant, β = −.14, t(173) = −2.10, p = .037, B = −12.74. Conscientiousness remained a significant predictor of behavior predictions, β = .28, t(173) = 3.96, p < .001, B = 18.08, and so did relationship quality, β = .32, t(173) = −4.19, p < .001, B = 12.99. None of the interaction terms were significant, ts < .03, ps > .739. In other words, participants used information about their positive feelings and about their personality to a similar degree in both conditions.

Discussion

This study suggests that instructing people to take their partner’s view when predicting their own relationship behaviors can reduce the overoptimism in judgments about relationship behaviors—participants’ predictions were more in line with their behavior when they were instructed to take the partner’s perspective.

One possible explanation for this effect may be that taking another’s perspective may lead to a different process of arriving at the prediction, similar to how predictions for other people are generated (e.g., Peetz & Kammrath, 2013). Forecasters may be less biased by their motivation to make their partner happy when taking an outside point of view. If this was the case, we would expect weaker links between perceived relationship quality (a proxy for motivation to make the partner happy) and predicted behavior in the perspective taking condition than in the control condition and stronger links between conscientiousness and predicted behavior. We tested this idea in the current study but found no support for it: Predictions were linked to both greater conscientiousness and more positive feelings for the partner, regardless of perspective taking instructions. However, this result may not be conclusive due to several limitations of the present study. First, the study was underpowered to detect small effects, given the high attrition rate. Second, conscientiousness and relationship quality were rated after the behavior prediction, because our primary concern in this first study was to examine effects on the behavior predictions. Thus, the exploratory scales might not have been salient when making a prediction and may have been influenced by the prediction rather than vice versa. Indeed, participants may have inferred attitudes (e.g., perceived relationship quality) from their forecasted behavior. Such inference of attitudes may explain why positive feelings were lower in the perspective taking condition, despite the finding that empathy and perspective taking are generally beneficial for relationships (Arriaga & Rusbult, 1998; Franzoi et al., 1985).

Another limitation of the initial study was that participants took the perspective of their partner. Thus, it is unclear whether the perspective taking effect is limited to adopting the perspective of the person that the behavior is directed toward. Could taking the perspective of another person outside the relationship (e.g., a friend) also reduce bias in relationship predictions? On the one hand, as an insider, the partner would have greater knowledge about the relationship and about the forecaster’s abilities in this domain than someone outside the relationship. On the other hand, as the recipient of relationship behaviors, the partner may still be motivated to overestimate the frequency of these behaviors, whereas someone outside the relationship would not have such a motivation.

Study 2

The next study examined again whether perspective taking affects forecasts about relationship behaviors. We aimed to replicate the effect of instructing forecasters to take the perspective of their partner on their predicted relationship behaviors. We extended on this design by also including a condition in which people were instructed to take the perspective of someone not involved in the relationship when making their forecast: a close friend. We expected that similar to taking the perspective of the partner, taking a friend’s perspective would lower the number of predicted relationship behaviors.

Another aim of this study was to test whether taking someone’s perspective is similar to making predictions for another person. To test this assumption, we included a condition in which participants predicted their partner’s behavior. We expected that the mean level of behaviors in this partner condition would be similar to the mean level of predicted behaviors when taking someone else’s perspective (the partner’s or the friend’s) but lower than the control condition.

Finally, this study aimed to address some of the limitations of Study 1 in testing the links between relationship quality and conscientiousness with predicted relationship behaviors. Unlike the previous study, this study measured perceived relationship quality and conscientiousness before the behavior predictions. We matched these scales to the actor in each condition: When predicting a partner’s behavior, participants also rated the partner’s perceived relationship quality and the partner’s conscientiousness, whereas those who predicted their own behavior rated their own perceived relationship quality and their own conscientiousness. We expected that when taking someone else’s perspective or when predicting someone else’s behavior, forecasts might be linked more with objective predictors of behavior (conscientiousness) and linked less with biasing influences (relationship quality) than when predicting one’s own behavior without additional instructions in the control group.

Method

Participants

We recruited 261 U.S. participants from Mechanical Turk, an online crowdsourcing platform. We aimed for 60 data sets per condition, which would allow us to detect small effects between groups with 70% power (GPower). We overrecruited by 10% to account for possible exclusions. We advertised for participants in committed relationships and excluded four participants who indicated that they were single at the time of the study and six participants who did not list initials for their partner (writing in “mystery lover” or “no idea”). We also excluded six participants who indicated that some of the questions did not work on their device. The final sample included 141 male and 103 female participants and 1 participant who identified as “other” gender (M _age = 31.79, SD = 8.40) in romantic relationships (M _{relationship length} = 78.33 months [6 years and 6 months], SD = 81.44 [6 years and 9 months]). About half of the participants (46.7%) were engaged or married and the rest were dating and living together (27.5%) or dating and not living together (25.8%).

Procedure

After a brief demographic questionnaire assessing age, gender, relationship status and length, and partner initials, participants completed two scales: a brief conscientiousness and relationship quality scale followed by a behavior prediction task. A full copy of the survey is included in the Supplementary Materials.

Scales

Participants reported their perceived relationship quality using the same 6-item scale as in Study 1 (adapted from Fletcher et al., 2000; Cronbach’s α = .92). They also reported their conscientiousness, using the same 9-item scale (BFI subscale for conscientiousness; John & Srivastava, 1999; Cronbach’s α = .89) as in Study 1. Participants in the partner condition predicted the perceived relationship quality and conscientiousness of their partner, rather than themselves.

Behavior prediction task

Participants were randomly assigned to a control condition, a partner perspective taking condition, a friend perspective taking condition, or a partner condition in which they predicted the partner’s pro-relational behavior. The control condition and partner perspective taking condition were identical to Study 1’s behavior prediction tasks. In the friend perspective taking condition, participants first identified a friend of the same age and gender also in a relationship and the perspective taking instructions read “…put aside your own frame of mind and attempt to adopt the point of view of your friend” and were then instructed to “indicate whether your friend would think you will complete the following behaviors for your partner.” In the partner condition, participants predicted their partner’s behavior toward themselves: “indicate whether you think your partner will complete the following behaviors for you.” In each of the four conditions, we added all 21 predicted behaviors for each day of the following week, resulting in a predicted behavior score between 0 and 147 for each participant (21 × 7 days).

Results

Does perspective taking affect behavior predictions?

First, we examined the effect of condition on predicted behaviors. In an omnibus test of differences between all four conditions, condition did not significantly affect likelihood predictions overall, F(3, 241) = 2.07, p = .104, d = .32 (see Table 2 for mean values). However, recall that we expected three of the conditions to be very similar and to differ from the control condition. More pertinently, therefore, we ran preplanned contrasts between the control condition and each of the other three conditions. Replicating Study 1, participants who took the perspective of their partner predicted to do significantly fewer behaviors in an upcoming week than participants who predicted their behaviors in the control condition, t(240) = 1.99, p = .048, d = .36. Note that the mean levels of predicted behaviors (71.92 vs. 87.59 behaviors per week) were similar to those of the entire sample in Study 1 (but lower than the subset of participants in Study 1 which completed the follow-up). Participants who took the perspective of a friend also predicted to do significantly fewer behaviors in an upcoming week than participants who predicted their behaviors in the control condition, t(240) = 2.29, p = .023, d = .42. A contrast comparing the control condition to all other three conditions was also significant, t(240) = 2.31, p = .022, d = .34.

Table 2.
A summary of means by conditions in Study 2.

Measured Variables Self-prediction about own behavior Self-prediction about own behavior, taking partner’s perspective Self-prediction about own behavior, taking friend’s perspective Self-prediction about partner’s behavior

n 61 59 59 65

Relationship quality 5.93_a (1.07) 5.63_a (1.20) 5.90_a (1.02) 5.87_a (1.22)

Conscientiousness 4.00_a (.75) 3.77_a,b (.81) 3.88_a,b (.91) 3.60_b (.86)

Mean of predicted relationship behaviors 87.59_a (45.63) 71.92_b (43.41) 69.58_b (47.09) 77.05_a,b (36.26)

Note. Means in one row with different subscripts are significantly different from each other at p < .05. Standard deviations are presented in parentheses.

Does perspective taking lead to similar mean levels of predicted behaviors as predictions for others?

Next, we examined whether the prediction of one’s own pro-relational behaviors from another’s perspective resulted in similar forecasts as predictions for someone else’s behavior (i.e., the partner). Although participants predicting their partner’s behavior predicted fewer behaviors than participants in the control condition, this difference was not significant, t(240) = 1.37, p = .172, d = .24. Predictions in the partner condition did not differ from the partner perspective condition, t(240) = −.66, p = .509, d = .12, or from the friend perspective condition, t(240) = −.96, p = .337, d = .17.

Does perspective taking increase the link between conscientiousness and predictions?

We next examined participants’ trait conscientiousness (M = 3.81, SD = .84) and relationship quality (M = 5.83, SD = 1.13). If participants took this information about themselves into account when making a prediction, these ratings should be linked with their predicted relationship behaviors. Before testing the link between relationship quality and personality with predicted behaviors, we examined whether they differed by condition (see Table 1 for mean values). Participants’ relationship quality did not differ between conditions, F(3, 241) = .91, p = .435. Conscientiousness differed by condition, F(3, 241) = 2.60, p = .053, as participants saw their partner as less conscientious than themselves, t(241) = 2.36, p = .019, d = .35.

Next, we examined the link between pro-relational behavior predictions and conscientiousness and relationship quality. Across conditions, participants predicted they or their partner would do more relationship behaviors if they rated themselves/their partner higher on conscientiousness, r(242) = .24, and if they reported higher relationship quality, r(242) = .27. We examined whether the links between conscientiousness and relationship quality with behavior predictions differed between control participants and those who were instructed to take another’s perspective. We regressed the number of predicted behaviors on perspective condition (0 = control condition; 1 = both perspective taking conditions), conscientiousness, relationship quality, and all two-way interaction terms. All variables were centered prior to analysis. The perspective condition effect remained significant, B = −15.20, SE = 7.02, β = −.16, t(178) = −2.17, p = .032. Relationship quality remained a significant predictor of behavior predictions, B = 6.08, SE = 3.01, β = .16, t(178) = 2.03, p = .044. The link between predicted behaviors and conscientiousness dropped to marginal significance, B = 7.78, SE = 4.34, β = .14, t(178) = 1.79, p = .075. Condition did not interact with conscientiousness, B = −16.56, SE = 9.91, β = −.14, t(178) = −1.67, p = .096, or relationship quality, B = 5.24, SE = 7.10, β = .06, t(178) = .74, p = .461. In other words, participants’ behavior predictions were linked to their conscientiousness and their perceived relationship quality regardless of whether they took someone else’s perspective or not.

Discussion

This study replicated Study 1 showing again that instructing people to take their partner’s view when predicting their own relationship behaviors can reduce the number of predicted relationship behaviors. This study also showed that taking a friend’s perspective had a similar effect as taking the perspective of the partner, suggesting that the perspective taking effect is not limited to persons who are involved with the predicted behaviors (i.e., the partner would be the recipient of the forecasted pro-relational behaviors).

We had argued that one reason for the perspective taking effect may be that taking an outside view leads to a different process of arriving at a prediction, similar to how predictions for other people are generated (Buehler et al., 1994; Newby-Clark et al., 2000; Peetz & Kammrath, 2013). This study provided some evidence that predictions from someone else’s view are similar to predictions for other people—the mean levels of predicted behaviors when predicting the partner’s behavior were comparable to the perspective conditions. However, as in Study 1, there was no evidence of predicted behaviors being linked more strongly to conscientiousness and less strongly with perceived relationship quality. While this study was adequately powered to detect small effects and measured conscientiousness and relationship quality before the behavior prediction task, it may be that the behavior prediction task did not link well with these two underlying factors. In the final study, we changed the prediction task to one that has been previously shown to be linked to conscientiousness and relationship quality differentially for self and other predictions (Peetz & Kammrath, 2013).

Study 3

The next study aimed to replicate both the perspective taking effect and the comparison with predictions for someone else’s behavior using a different behavior prediction task. Rather than indicating the instances of 21 relationship behaviors over 1 week, participants were asked to rate the likelihood of performing each of 14 behaviors at least once, which was aggregated into an average likelihood judgment. We assigned participants to predict how likely it would be that they themselves would perform these behaviors or to predict how likely their partner would think it would be that they will perform these behaviors. We also included two conditions in which participants predicted someone else’s behavior: They predicted either how likely their partner would be to perform these behaviors or how likely a friend would be to perform the behaviors for the friend’s partner. We expected that the findings of the previous studies would replicate with this different measure of forecasting, namely, that participants taking the partner’s perspective would predict lower likelihood than control participants and similar likelihood levels as participants predicting their partner’s or their friend’s behavior. We also again examined whether different types of information tend to be considered when predictions are made from one’s own or from someone else’s perspective.

Method

Participants

We recruited 149 participants from CrowdFlower (U.S. and Canadian participants only). We aimed for 39 data sets per condition, which would allow us to detect medium effects (as those found in Studies 1 and 2) in contrasts for four groups with 70% power. We advertised for participants in committed relationships and excluded six participants who indicated that they were single at the time of the study. The final sample included 71 male and 78 female participants (M _age = 36.58, SD = 12.06) in romantic relationships (M _{relationship length} = 107.20 months [8 years and 11 months], SD = 118.53 [9 years and 10 months]). About half of the participants (55.7%) were engaged or married and the rest were dating and living together (16.1%) or dating and not living together (28.2%).

Procedure

A brief demographic questionnaire assessed age, gender, relationship status, relationship length, and partner’s initials. Participants were randomly assigned to the control condition, the perspective condition, the partner condition, or the friend condition. Participants in the friend condition were asked to think of a friend of similar age and gender as themselves who also is in a relationship. Next participants completed background scales on relationship quality and conscientiousness. Participants in the control and perspective condition reported their own perceived relationship quality (e.g., “How satisfied are you with your relationship?”), participants in the partner condition reported their partner’s relationship quality (e.g., “How satisfied is your partner with your relationship?”), and participants in the friend condition reported their friend’s relationship quality (e.g., “How satisfied is your friend with his or her relationship?”) using the full, unabbreviated, 18-item perceived relationship quality scale (Fletcher et al., 2000; Cronbach’s α = .97) with response scales ranging from not at all* (1) to very much (7). Participants completed the BFI subscale for conscientiousness (John & Srivastava, 1999; Cronbach’s α = .86), matched to condition as in Studies 1 and 2. That is, participants in the control and perspective condition reported their own conscientiousness, participants in the partner condition reported their partner’s conscientiousness, and participants in the friend condition reported their friend’s conscientiousness.

Next, participants predicted behavior likelihood for 14 pro-relational behaviors (adapted from Peetz & Kammrath, 2013; Cronbach’s α = .88) on response scales ranging from not at all likely (1) to extremely likely (7). Participants in the control and perspective condition predicted the likelihood that they themselves would do the behaviors (e.g., “Surprise your partner with their favorite beverage or candy.”). Participants either made these predictions from their own perspective (“Please predict the likelihood of each behavior based on how likely you are to perform each behavior for your partner”) or from their partner’s perspective (“Please predict the likelihood of each behavior based on how likely your partner would think you are to perform each behavior for you.”). Participants in the partner condition predicted the likelihood that their partner would do the behaviors (e.g., “Please predict the likelihood of each behavior based on how likely you think your partner is to perform each behavior for you.”). Participants in the friend condition predicted the likelihood that their friend would do the behaviors for the friend’s partner (e.g., “Please predict the likelihood of each behavior based on how likely you think your friend is to perform each behavior for his or her partner.”).

Results

Does perspective taking affect behavior predictions?

First, we examined behavior likelihood predictions. In an omnibus test of differences between all four conditions, condition did not significantly affect likelihood predictions overall, F(3, 141) = 2.14, p = .123, d = .38 (see Table 3 for mean values). However, recall that we expected three of the conditions to be very similar and all to differ from the control condition. More pertinently, therefore, we ran preplanned contrasts between the control condition and each of the other three conditions. Participants predicting their own behavior rated the likelihood of performing the 14 behaviors higher than those participants instructed to take the perspective of the partner, t(141) = 1.96, p = .052, d = 45. A contrast comparing the control group to all three other groups was significant, t(141) = 2.42, p = .017, d = .45.

Table 3.
A summary of means by conditions in Study 3.

Measured Variables Prediction about own behavior Prediction about own behavior, taking partner’s perspective Prediction about partner’s behavior Prediction about friend’s behavior

n 39 37 36 37

Relationship quality 5.52_a (1.07) 5.48_a (1.16) 5.74_a (.86) 5.38_a (1.41)

Conscientiousness 4.61_a (.74) 4.39_a (.73) 4.48_a (.78) 4.42_a (.80)

Mean likelihood of pro-relational behaviors 5.29_a (1.06) 4.82_b (1.09) 4.80_b (1.06) 4.82_b (.98)

Note. Means in one row with different subscripts are significantly different from each other at p ≤ .05. Standard deviations are presented in parentheses.

Does perspective taking lead to similar mean levels of predicted behaviors as predictions for others?

Participants in the control group also rated the likelihood of performing the behaviors higher than participants predicting their partner’s likelihood of performing the behaviors, t(141) = 1.98, p = .050, d = .46, or participants predicting their friend’s likelihood of performing the behaviors, t(141) = 1.96, p = .053, d = 45. Participants in the perspective condition did not differ in their likelihood judgments from those predicting their partner’s behavior, t(141) = .05, p = .960, d = .01, or their friend’s behavior, t(141) = −.01, p = .989, d = .001. Finally, an exploratory contrast comparing the control group to all three other groups was significant, t(141) = 2.42, p = .017, d = .45.

Does perspective taking increase the link between conscientiousness and predictions?

We next examined participants’ trait conscientiousness (M = 3.81, SD = .84) and relationship quality (M = 5.83, SD = 1.13). If participants took this information about themselves into account when making a prediction, these ratings should be linked with their predicted relationship behaviors. Before testing the link between relationship quality and conscientiousness with predicted behaviors, we examined whether they differed by condition (see Table 3 for mean values). Participants’ ratings of the quality of their relationship and their conscientiousness did not differ by condition, F(3, 145) = .63, p = .600, d = .23, and F(3, 145) = .59, p = .625, d = .22, respectively. None of the contrasts comparing the control condition to the other conditions was significant, ts < 1.20, ps > .231.

We examined whether perspective condition interacted with the links between conscientiousness and relationship quality and behavior likelihood judgments. We regressed likelihood judgments on perspective condition (0 = control condition; 1 = perspective taking condition), conscientiousness, relationship quality, and all two-way interaction terms. All variables were centered before analysis. Results showed that the Condition × Relationship Quality interaction term was significant, B = −.44, SE = .20, β = −.24, t(66) = −2.25, p = .028, and the Condition × Conscientiousness interaction term was significant, B = .62, SE = .30, β = .22, t(66) = 2.05, p = .044. Main effects of perspective condition, B = −.51, SE = .21, β = −.23, t(66) = −2.49, p = .015, relationship quality, B = .42, SE = .10, β = .44, t(66) = 4.15, p < .001, and conscientiousness, B = .36, SE = .16, β = .24, t(66) = 2.67, p = .027, remained significant. We next broke down the interactions by examining the simple slopes of conscientiousness and relationship quality in each condition (Aiken, West, & Reno, 1991). The interactions are portrayed in Figure 2. For participants predicting their own behavior, relationship quality was strongly linked to likelihood judgments, B = .54, SE = .10, β = .55, t(66) = 5.20, p < .001, but conscientiousness was not significantly linked to likelihood judgments, B = .21, SE = .15, β = .14, t(66) = 1.33, p = .188. For participants who predicted behaviors from the perspective of their partner, however, both relationship quality, B = .31, SE = .12, β = .32, t(66) = 2.52, p = .014, and conscientiousness, B = .52, SE = .20, β = .35, t(66) = 2.62, p = .011, were linked to likelihood judgments. The figure illustrates that estimated likelihood judgments are widely different for participants low (−1 SD) in relationship quality versus high (+1 SD) in relationship quality in the control condition, a gap that is reduced to more moderate proportions in the perspective taking condition. In contrast, estimated likelihood judgments were not at all different for participants low (−1 SD) in conscientiousness versus high (+1 SD) in conscientiousness in the control condition, whereas a moderate gap exists in the perspective taking condition.

Figure 2.
Average likelihood judgments for those considering their pro-relational behaviors from their own or from their partner’s perspective in Study 3 by those low (−1 SD) and high (+1 SD) in relationship quality and conscientiousness, respectively.

Discussion

This study replicated the perspective taking effect using a different behavior prediction task: Rather than predicting the frequency of relationship behaviors over 1 week as in Studies 1 and 2, participants predicted the likelihood of doing 14 behaviors at least once. Participants’ likelihood judgments were more similar to judgments for another person’s behavior (either the partner or a friend in a similar situation) than for their own behavior when not taking someone else’s perspective. Stepping outside one’s own head appears to lead people to treat predictions for their own behavior similar to predictions for other people’s behavior. This study yielded evidence that participants appeared to take the more objective type of information—conscientiousness—more into account when taking another’s perspective in predicting one’s own behavior. However, given that the two previous studies did not find the link between conscientiousness and behavior predictions to be different by perspective taking condition, this result may be interpreted with caution.

General discussion

People tend to be optimistically biased when predicting positive behaviors that they would do for their partner. In three studies, we examined how perspective taking may shift predictions of relationship behaviors. Consistent with the hypotheses, we found that taking the perspective of another person when making predictions about how much one would do for a romantic partner reduced the number of predicted behaviors (Studies 1 and 2) and reduced the judged likelihood of performing relationship behaviors once (Study 3). Consequently, the lowered predictions were less optimistically biased and more accurate compared to self-reported behavior a week later (Study 1). Predictions about one’s own behavior from another person’s perspective resembled predictions about other people’s behavior (Studies 2 and 3).

Why does perspective taking improve prediction accuracy?

We expected that perspective taking shifts people’s focus away from their feelings (which are not good predictors of behavior) toward their more enduring personality traits (which are better predictors of behavior). There was some evidence that predictions were linked more with perceived relationship quality and less with conscientiousness when forming predictions from a partner’s perspective (Study 3). However, this pattern of results was not present in the other two studies when participants were predicting instances of behaviors instead of judging the likelihood of behaviors (Studies 1 and 2). Given this inconsistency, more research is needed to pin point the particular mechanisms by which perspective taking can improve prediction accuracy.

It is possible that the benefits of perspective taking for prediction accuracy may be due to other cognitive shifts in how predictions are generated. For example, people might give more attention to past behavior when looking at their behavior through the lens of another person’s view—just as people tend to use knowledge about past behavior more when making judgments about others rather than the self (Helzer & Dunning, 2012).

Limitations

All three of the present studies use samples from crowdsourcing platforms (Mechanical Turk, CrowdFlower), where participants receive minimal compensation (50 cents for each study) for participating in research. On the one hand, samples recruited in such a way may lower the validity of the research, as participants completed the research online, rather than in the presence of the researcher, because they might be motivated only by the monetary gain, because they might be experienced participants who have completed several surveys or because they do not answer the study questions thoughtfully or carefully. On the other hand, crowdsourcing platform samples reflect a broader cross section of society and longer term relationships than a convenience student sample would. As crowdsourcing has become a more common recruitment method, several studies that have emerged give testimony to data obtained in this way being of acceptable quality (see Chandler & Shapiro, 2016, for a review). For example, the rate of compensation does not affect data quality, and data have been shown to be as reliable as those from more traditional samples (Buhrmester, Kwang, & Gosling, 2011).

Future research might examine several perspectives (e.g., take the perspective of parents, strangers) and several types of behaviors (task completion times, saving money, exercising, prosocial behavior) to determine boundary conditions of perspective taking effects on prediction. There may be some people outside the relationship who are superior forecasters and whose perspective would be most helpful in improving predictions. For example, there is evidence that female friends of the female in a couple are particularly good at making other relationship predictions, predicting whether a relationship will last more accurately (Agnew et al., 2001) or as accurately as the couple themselves (Loving, 2006).

Future research might also examine how perspective taking affects participants’ predictions for other people. Perspective taking has been shown to produce a greater overlap between persons’ self-view and their judgments about other people (Davis, Conklin, Smith, & Luce, 1996). It is possible therefore, that taking the perspective of someone else while predicting that person’s behavior will result in the same optimistic bias people show for their self-predictions. A preliminary examination of this possibility is included in the Supplemental Materials.

Benefits and side effects of perspective taking

Perspective taking in relationships is generally linked with generally positive outcomes such as relationship satisfaction (Franzoi et al., 1985) and marital adjustment (Long & Andrews, 1990), and the present work also shows positive outcomes in a specific instance (forecasting relationship-relevant behaviors). However, there is also evidence that perspective taking in relationships can increase biases such as the transparency illusion, leading to negative outcomes for couples instructed to take each other’s perspective during an interaction (Vorauer & Sucharyna, 2013). Additionally, when a partner’s thoughts and feelings reflect negatively on the relationship, greater empathic accuracy in perceiving these potentially damaging thoughts can spell trouble for the relationship (Simpson, Orina, & Ickes, 2003). For example, when seeing their partner discuss romantic alternatives, anxious–ambivalent people reported more empathic accuracy, but this accuracy led to more distress and reduced closeness with the partner (Simpson, Ickes, & Grich, 1999). It is important to note that our studies were limited to examining one instance of perspective taking rather than repeated (Vorauer & Suchuryna, 2013) or chronic (Franzoi et al., 1985; Long & Andrews, 1990) perspective taking. We also only examined cognitive (what would your partner think) not emotional (what would your partner feel) perspective taking.

Furthermore, while perspective taking may increase accuracy, it may be shortsighted to assume that accuracy is always advantageous, as optimistic illusions may also be beneficial (Taylor & Brown, 1988). In relationships, accuracy can be more important at some stages of the relationship than others (Gagné & Lydon, 2004). We argue that because falling short on expectations or breaking promises can cause a number of problems in relationships (Cameron et al., 2002; Hui, Bond, & Molden, 2012; Kammrath & Peetz, 2012; Metts, 1994), improving the accuracy of specific predictions about behavior relevant to the relationship is beneficial.

Implications

The present research contributes to the literature examining promises and predictions about behavior in interpersonal relationships (Cameron et al., 2002; Peetz & Kammrath, 2011). People tend to make more promises to their partners than what they can keep up with (Peetz & Kammrath, 2011), which can have damaging consequences for close relationships. “Broken Promises” was one of the most-cited problems in one study of conflict in relationships (Metts, 1994). This research provides some insights that can improve a promiser’s ability to predict the pro-relational behaviors that they can feasibly do—ones that they will not fall short of. This research may be applied to make more reasonable promises in a relationship context. When committing themselves to a course of action, promisers may do well to ask themselves “What would my partner think I can do?”

More generally, this research contributes to the study of relationship judgments (for reviews, see Fletcher & Kerr, 2010; Gagné & Lydon, 2004). Our research fits with the general pattern of bias and accuracy in relationship judgments, as participants were both accurate (their predicted behavior correlated highly with their actual behavior, Study 1) and biased (mean-level difference between predicted and actual behavior)—although the present studies targeted judgments about individuals’ own (relationship-relevant) behavior rather than judgments about the relationship or the partner.

Even more broadly, the present research contributes to the literature on predictions (e.g., Buehler et al., 1994; Newby-Clark et al., 2000). Whereas past research shows that predicting the behavior of another person can enhance the accuracy of predictions, the present research extends these findings by showing that simply adopting another person’s perspective can improve predictions concerning one’s own behavior. Past research has documented systematic optimistic biases present in a number of prediction domains including performance and achievement (Helzer & Dunning, 2012), task completion times (Buehler et al., 1994), and personal finances (Peetz & Buehler, 2013). Taking another person’s perspective could potentially temper optimism in those domains, as well.

Conclusions

Making forecasts about one’s own behavior can be biased in a number of ways. Especially in relationships, falling short on behaviors that were promised to the partner can spell trouble for the relationship. This research suggests an easy and convenient technique that people can practice to improve their prediction accuracy: Taking another’s perspective before predicting what they will be able to do. As our studies suggest, such perspective taking will temper their optimism as much as obtaining an outside perspective on one’s actions. Thus, this mental “trick” may contribute to promises and predictions that are better aligned with actual behavior and have positive downstream consequences for relationships.

Supplemental material

Supplementary_materials_for_Perspective_Taking_proofedJPeetz_(4) - Through your partner’s eyes: Perspective taking tempers optimism in behavior predictions

Supplementary_materials_for_Perspective_Taking_proofedJPeetz_(4) for Through your partner’s eyes: Perspective taking tempers optimism in behavior predictions by Johanna Peetz, Aaron Maccosham, and Kali May in Journal of Social and Personal Relationships

Measured Variables	Self-prediction about own behavior	Self-prediction about own behavior, taking partner’s perspective
n	47	36
Positive feelings	6.07_a (.90)	5.50_b (1.29)
Conscientiousness	4.07_a (.68)	3.88_a (.79)
Perspective taking ability	3.55_a (.90)	3.54_a (1.29)
Mean of predicted relationship behaviors	101.87_a (35.19)	83.47_b (46.75)
Mean of actual relationship behaviors	87.32_a (37.67)	80.69_a (40.96)

Measured Variables	Self-prediction about own behavior	Self-prediction about own behavior, taking partner’s perspective	Self-prediction about own behavior, taking friend’s perspective	Self-prediction about partner’s behavior
n	61	59	59	65
Relationship quality	5.93_a (1.07)	5.63_a (1.20)	5.90_a (1.02)	5.87_a (1.22)
Conscientiousness	4.00_a (.75)	3.77_a,b (.81)	3.88_a,b (.91)	3.60_b (.86)
Mean of predicted relationship behaviors	87.59_a (45.63)	71.92_b (43.41)	69.58_b (47.09)	77.05_a,b (36.26)

Measured Variables	Prediction about own behavior	Prediction about own behavior, taking partner’s perspective	Prediction about partner’s behavior	Prediction about friend’s behavior
n	39	37	36	37
Relationship quality	5.52_a (1.07)	5.48_a (1.16)	5.74_a (.86)	5.38_a (1.41)
Conscientiousness	4.61_a (.74)	4.39_a (.73)	4.48_a (.78)	4.42_a (.80)
Mean likelihood of pro-relational behaviors	5.29_a (1.06)	4.82_b (1.09)	4.80_b (1.06)	4.82_b (.98)

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a grant from the Social Science and Humanities Research Council of Canada (#435-2012-1211) to the first author.

Supplemental material

Supplemental material for this article is available online.

Note

References

Agnew

C. R.

Loving

T. J.

Drigotas

S. M.

(2001). Substituting the forest for the trees: Social networks and the prediction of romantic relationship state and fate. Journal of Personality and Social Psychology, 81, 1042–1057.

Aiken

L. S.

West

S. G.

Reno

R. R.

(1991). Multiple regression: Testing and interpreting interactions. New York, NY: Sage.

Armor

D. A.

Taylor

S. E.

(1998). Situated optimism: Specific outcome expectancies and self-regulation. Advances in Experimental Social Psychology, 30, 309–379.

Arriaga

X. B.

Rusbult

C. E.

(1998). Standing in my partner’s shoes: Partner perspective taking and reactions to accommodative dilemmas. Personality and Social Psychology Bulletin, 24, 927–948.

Batson

C. D.

Early

Salvarani

(1997). Perspective taking: Imagining how another feels versus imagining how you would feel. Personality and Social Psychology Bulletin, 23, 751–758.

Buehler

Griffin

MacDonald

(1997). The role of motivated reasoning in optimistic time predictions. Personality and Social Psychology Bulletin, 23(3), 238–247.

Buehler

Griffin

Ross

(1994). Exploring the “planning fallacy”: Why people underestimate their task completion times. Journal of Personality and Social Psychology, 67, 366–381.

Buhrmester

Kwang

Gosling

S. D.

(2011). Amazon’s Mechanical Turk a new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6, 3–5.

Cameron

J. J.

Ross

Holmes

J. G.

(2002). Loving the one you hurt: Positive effects of recounting a transgression against an intimate partner. Journal of Experimental Social Psychology, 38, 307–314.

10.

Chandler

Shapiro

(2016). Conducting clinical research using crowdsourced convenience samples. Annual Review of Clinical Psychology, 12, 53–81.

11.

Chartrand

T. L.

Bargh

J. A.

(1999). The chameleon effect: The perception–behavior link and social interaction. Journal of Personality and Social Psychology, 76, 893–910.

12.

Davis

M. H.

(1980). A multidimensional approach to individual differences in empathy. JSAS Catalog of Selected Documents in Psychology, 10, 85.

13.

Davis

M. H.

Conklin

Smith

Luce

(1996). Effect of perspective taking on the cognitive representation of persons: A merging of self and other. Journal of Personality and Social Psychology, 70, 713–726.

14.

Davis

M. H.

Oathout

H. A.

(1987). Maintenance of satisfaction in romantic relationships: Empathy and relational competence. Journal of Personality and Social Psychology, 53, 397–410.

15.

Dunning

(2007). Prediction: The inside view. In Kruglanski

A. W.

Higgins

E. T.

(Eds.), Social psychology (pp. 69–90). New York, NY: The Guilford Press.

16.

Epley

Dunning

(2000). Feeling “holier than thou”: Are self-serving assessments produced by errors in self-or social prediction? Journal of Personality and Social Psychology, 79, 861–875.

17.

Epley

Dunning

(2006). The mixed blessings of self-knowledge in behavioral prediction: Enhanced discrimination but exacerbated bias. Personality and Social Psychology Bulletin, 32, 641–655.

18.

Fincham

F. D.

Bradbury

T. N.

(1993). Marital satisfaction, depression, and attributions: A longitudinal analysis. Journal of Personality and Social Psychology, 64, 442–452.

19.

Fletcher

G. J.

Kerr

P. S.

(2010). Through the eyes of love: Reality and illusion in intimate relationships. Psychological Bulletin, 136, 627–658.

20.

Fletcher

G. J.

Simpson

J. A.

Thomas

(2000). The measurement of perceived relationship quality components: A confirmatory factor analytic approach. Personality and Social Psychology Bulletin, 26, 340–354. doi:10.1177/0146167200265007

21.

Franzoi

S. L.

Davis

M. H.

Young

R. D.

(1985). The effects of private self-consciousness and perspective taking on satisfaction in close relationships. Journal of Personality and Social Psychology, 48, 1584–1594.

22.

Gagné

F. M.

Lydon

J. E.

(2004). Bias and accuracy in close relationships: An integrative review. Personality and Social Psychology Review, 8, 322–338.

23.

Galinsky

A. D.

Wang

C. S.

(2005). Perspective-taking and self-other overlap: Fostering social bonds and facilitating social coordination. Group Processes and Intergroup Relations, 8, 109–124.

24.

Galinsky

A. D.

Moskowitz

G. B.

(2000). Perspective-taking: Decreasing stereotype expression, stereotype accessibility, and in-group favoritism. Journal of Personality and Social Psychology, 78, 708–724.

25.

Helzer

E. G.

Dunning

(2012). Why and when peer prediction is superior to self-prediction: The weight given to future aspiration versus past achievement. Journal of Personality and Social Psychology, 103, 38–55.

26.

Hui

C. M.

Bond

M. H.

Molden

D. C.

(2012). Why do (n’t) your partner’s efforts at self-improvement make you happy? An implicit theories perspective. Personality and Social Psychology Bulletin, 38, 101–113.

27.

John

O. P.

Srivastava

(1999). The Big-Five trait taxonomy: History, measurement, and theoretical perspectives. In Pervin

L. A.

John

O. P.

(Eds.), Handbook of personality: Theory and research (Vol. 2, pp. 102–138). New York: Guilford Press.

28.

Karney

B. R.

Bradbury

T. N.

(2000). Attributions in marriage: State or trait? A growth curve analysis. Journal of Personality and Social Psychology, 78, 295–309.

29.

Kammrath

Peetz

(2012). You promised you’d change: How incremental and entity theorists react to a romantic partner’s change attempts. Journal of Experimental Social Psychology, 48, 570–574.

30.

Kimmes

J. G.

Durtschi

J. A.

(2016). Forgiveness in romantic relationships: The roles of attachment, empathy, and attributions. Journal of Marital and Family Therapy, 42, 645–658.

31.

Kimmes

J. G.

Edwards

A. B.

Wetchler

J. L.

Bercik

(2014). Self and other ratings of dyadic empathy as predictors of relationship satisfaction. The American Journal of Family Therapy, 42, 426–437.

32.

Koehler

Poon

C. S. K.

(2006). Self-predictions overweight strength of current intentions. Journal of Experimental Social Psychology, 42, 517–524.

33.

Krizan

Windschitl

P. D.

(2007). The influence of outcome desirability on optimism. Psychological Bulletin, 133, 95–121.

34.

Kruger

Gilovich

(2004). Actions, intentions, and self-assessment: The road to self-enhancement is paved with good intentions. Personality and Social Psychology Bulletin, 30, 328–339.

35.

Long

E. C.

Andrews

D. W.

(1990). Perspective taking as a predictor of marital adjustment. Journal of Personality and Social Psychology, 59, 126–131.

36.

Loving

T. J.

(2006). Predicting dating relationship fate with insiders’ and outsiders’ perspectives: Who and what is asked matters. Personal Relationships, 13, 349–362.

37.

MacDonald

T. K.

Ross

(1999). Assessing the accuracy of predictions about dating relationships: How and why do lovers’ predictions differ from those made by observers? Personality and Social Psychology Bulletin, 25, 1417–1429.

38.

Maxwell

J. A.

Joel

Peetz

MacDonald

Baucom

(2017). Expectations drive perceptions of how much effort a romantic partner puts into the relationship from week to week. Manuscript in preparation.

39.

McCullough

M. E.

Rachal

K. C.

Sandage

S. J.

Worthington

E. L.

Jr Brown

S. W.

Hight

T. L.

(1998). Interpersonal forgiving in close relationships: II. Theoretical elaboration and measurement. Journal of Personality and Social Psychology, 75, 1586–1603.

40.

McCullough

M. E.

Worthington

E. L.

Jr. Rachal

K. C.

(1997). Interpersonal forgiving in close relationships. Journal of Personality and Social Psychology, 73, 321–336.

41.

Metts

(1994). Relational transgressions. In Cupach

W. R.

Spitzberg

B. H.

(Eds.), The dark side of interpersonal communication (pp. 217–239). Hillsdale, NJ: Erlbaum.

42.

Miller

D. T.

Ratner

R. K.

(1998). The disparity between the actual and assumed power of self-interest. Journal of Personality and Social Psychology, 74, 53–62.

43.

Miller

P. J. E.

Rempel

J. K.

(2004). Trust and partner-enhancing attributions in close relationships. Personality and Social Psychology Bulletin, 30, 695–705.

44.

Newby-Clark

I. R.

Ross

Buehler

Koehler

D. J.

Griffin

(2000). People focus on optimistic scenarios and disregard pessimistic scenarios while predicting task completion times. Journal of Experimental Psychology: Applied, 6, 171–182.

45.

Peetz

Buehler

(2009). Is there a budget fallacy? The role of savings goals in the prediction of personal spending. Personality and Social Psychology Bulletin, 35, 230–242.

46.

Peetz

Buehler

(2013). Different goals, different predictions: Accuracy and bias in financial planning for events and time periods. Journal of Applied Social Psychology, 43, 1079–1088.

47.

Peetz

Kammrath

(2011). Only because I love you: Why people make and why they break promises in romantic relationships. Journal of Personality and Social Psychology, 100, 887–904.

48.

Peetz

Kammrath

(2013). Folk understandings of self-regulation in relationships: Recognizing the importance of self-regulatory ability for others, but not the self. Journal of Experimental Social Psychology, 49, 712–718.

49.

Pinkus

R. T.

Lockwood

Marshall

T. C.

Yoon

H. M.

(2012). Responses to comparisons in romantic relationships: Empathy, shared fate, and contrast. Personal Relationships, 19, 182–201.

50.

Rusbult

C. E.

Verette

Whitney

G. A.

Slovik

L. F.

Lipkus

(1991). Accommodation processes in close relationships: Theory and preliminary empirical evidence. Journal of Personality and Social Psychology, 60, 53–78.

51.

Siegert

J. R.

Stamp

G. H

. (1994). “Our first big fight” as a milestone in the development of close relationships. Communication Monographs, 61, 345–360.

52.

Simpson

J. A.

Ickes

Grich

(1999). When accuracy hurts: Reactions of anxious–ambivalent dating partners to a relationship-threatening situation. Journal of Personality and Social Psychology, 76, 754–769.

53.

Simpson

J. A.

Orina

M. M.

Ickes

(2003). When accuracy hurts, and when it helps: A test of the empathic accuracy model in marital interactions. Journal of Personality and Social Psychology, 85, 881–893.

54.

Taylor

S. E.

Brown

J. D.

(1988). Illusion and well-being: A social psychological perspective on mental health. Psychological Bulletin, 103, 193–210.

55.

Van Boven

White

Kamada

Gilovich

(2003). Intuitions about situational correction in self and others. Journal of Personality and Social Psychology, 85, 249–258.

56.

Vorauer

J. D.

Sucharyna

T. A.

(2013). Potential negative effects of perspective-taking efforts in the context of close relationships: Increased bias and reduced satisfaction. Journal of Personality and Social Psychology, 104, 70–86.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.36 MB