In the current paper we test the hypothesis that individual differences in approach and avoidance social motivation are associated with how people evaluate their romantic partners. We first provide a brief overview of the well-established approach and avoidance motivation framework. We then turn our attention to work that has been done on approach and avoidance motivation in social interactions and relationships. We will then make a case for why approach and avoidance motivation would be associated with partner evaluations.

Approach and Avoidance Motivation

The approach and avoidance framework has been found to be an important perspective in the study of motivation. Approach motivation is the desire within an individual to pursue positive outcomes whereas avoidance motivation is the desire to avoid negative outcomes (Elliot & Gable, 2019; Monni et al., 2020). Research has found that individuals differ in the strength of these motives, and that approach and avoidance processes likely are rooted in separate neurological pathways, highlighting their distinctiveness and demonstrating that they are capable of independent effects (Elliot, 2008; Peterson et al., 2008; Slepian et al., 2017). Research on approach and avoidance motivation has not been limited to a single content area. Rather, the approach and avoidance distinction has been found to be important in variety of areas such as group dynamics, academic achievement, health behaviors, and emotion to name a few (Eder et al., 2013; Ryan, 2006).

Moreover, these motives have been associated with different patterns of information processing and judgments. For example, van Prooijen et al. (2006) have found that approach and avoidance motives can impact procedural justice judgments. Specifically, participants who were manipulated to adopt an approach framework or who measured high on approach motivation as an individual difference factor gave stronger procedural justice evaluations if they were given the opportunity to voice their opinion (compared to no voice conditions). Experimentally manipulated approach orientations have also been shown to be associated with viewing items as part of a larger whole (as opposed to thinking of them as belonging to smaller groups or units), with those primed to be in an approach orientation reporting more interdependent (as opposed to independent) self-construals (Nussinson et al., 2012). Additionally, Updegraff et al. (2004) found that approach motives can shape how information is processed when making judgments of daily well-being. Individuals high in approach motives tend to base well-being judgments more heavily on the presence of positive emotional experiences and are less swayed by the presence of negative emotional experiences. Together, these studies underscore the idea that approach and avoidance motives can change how we process and interpret information.

Social Approach and Avoidance Motivation

Of particular interest to the current study, the approach and avoidance framework has been extended to the social environment (Gable, 2015; Gable & Gosnell, 2013). As social relationships typically involve both incentives (intimacy, companionship) and threats (rejection, manipulation), both approach and avoidance goal systems should be employed. In addition, individuals vary in the strength of these motives, with some individuals focusing on incentives while others remain highly vigilant to threat. Past research supports the notion that approach and avoidance motives for social relationships differentially influence various aspects of social interactions and relationships. Measures of approach and avoidance motivation in the social domain have been developed and shown to predict a variety of social outcomes, such as satisfaction, loneliness, and the reporting of the frequency and impact of positive and negative social events (Elliot et al., 2006; Gable, 2006; Gable, 2015). In addition, in specific close relationships, such as a romantic relationship, approach and avoidance motivations have been shown to be associated with a variety of important outcomes, such as relationship satisfaction, sexual desire, responsive behaviors, stress communication, dyadic coping, commitment strategies, and relational aggression (Frank & Brandstätter, 2002; Impett et al., 2005; Impett et al., 2008; Impett et al., 2010; Kuster et al., 2017; Moron & Mandal, 2021; Nikitin & Freund, 2019; Strachman & Gable, 2006a). Past work has typically shown more favorable outcomes to be associated with stronger approach goals. For instance, recent work has shown that approach goals predict observed positive and constructive behaviors during conflict discussion between romantic partners, whereas higher avoidance goals predicted more negative behaviors (Bernecker et al., 2019).

Motivation and Relationship Processes

Although work has suggested that approach and avoidance motives are important in relationships, it is important to look for the mechanisms that may underlie the link between motives and outcomes in relationships. Some research has examined how motives may shape attention, cognitive biases, and the weighting of information (Gable, 2015). For example, Strachman and Gable (2006b) found that avoidance social goals were associated with increased memory for negative information and a tendency to interpret ambiguous cues negatively. Others have shown that approach and avoidance goals can influence the types of attributions made for acceptance or rejection in a speed-dating context and expectations regarding relationship formation (Nikitin et al., 2019). Other work has demonstrated how goals can influence what we see, with those high in avoidance goals “seeing” more negative emotions in an ambiguous facial expression (Nikitin & Freund, 2015). In addition, evidence for differential evaluation and weighting of information has also been found. Gable and Poore (2008) found that individuals high in approach goals based their relationship satisfaction ratings more on the presence or absence of positive relationship information, whereas individuals high in avoidance goals based their satisfaction ratings on the presence or absence of negative information.

Romantic Partner Perceptions

People seem to be fairly accurate in their partner perceptions, yet also have the capacity to be flexible in the way they think about their significant others and their relationships so as to better maintain those relationships (e.g., Fletcher, 2015; Karney et al., 2004). Sometimes this flexibility manifests in biases in the content of perceptions of a partner’s traits or behaviors (e.g., Fletcher & Kerr, 2010). For example, individuals may create more positive views of their partners than their partners have of themselves and these illusions can buffer them from the ups and downs of relationships (e.g., Murray et al., 1996) or even aid a partner’s progress toward becoming a more ideal version of him or herself (e.g., Rusbult et al., 2009).

Another way in which peoples’ thoughts about their partners are flexible is in the structure of their thinking. For example, Karney and Frye (2002) found that spouses’ recollections of how their relationship developed were shaped by their current satisfaction with the relationship. Specifically, in addition to biased recollection of their past levels of satisfaction, they also emphasized declines and gains in these levels in ways that were consistent with their current marital satisfaction, such that higher current satisfaction predicted a feeling of gain and vice versa, regardless of actual change (see also Sprecher, 1999). In addition, partners’ positive and negative traits can be cast in various shades. Neff and Karney (2002) showed that positive traits tend to be cast as more global while negative traits tend to be cast in more specific terms, and this is shown more by satisfied spouses than unsatisfied spouses. These studies demonstrate that not only do individuals play somewhat fast and loose with the content and structure of their thinking about their partners’ behaviors and traits, but that these patterns of thinking are linked to important relationship qualities, such as satisfaction.

Understanding how individuals evaluate and perceive their romantic relationships is important. A critical component of these perceptions rests on which traits and qualities one values as important and perceives to exist in a relationship partner (Eastwick & Neff, 2012; Fletcher et al., 2000; Fletcher & Simpson, 2000). However, there is limited work focused on examining individual differences that might influence these perceptions. We propose that individual differences in motivational style–specifically, approach and avoidance motivation—are key factors that may impact partner perceptions. In the present work we examine how approach and avoidance motives in relationships may shape the way in which individuals perceive their partners’ characteristics and how these perceptions are associated with relationship outcomes (such as relationship satisfaction).

Present Study

The present study was designed to extend the examination of mechanisms associated with approach and avoidance goals in relationships. Specifically, we sought to understand how approach and avoidance social goals influence the evaluation of a relationship partners’ traits and characteristics. We expected that due to the tendency for people with strong avoidance goals to be drawn to negative stimuli in their environment and to weigh negative information more heavily, they may be more likely to think of their partners in terms of the negative traits they lack, compared to those with weaker avoidance goals. In addition, we expected that individuals with stronger approach goals would be more likely to think of their partners in terms of the positive traits they possess, compared to those with weaker approach goals.

Although past research has examined individual difference variables (e.g., attachment, communal motives, self-esteem) that affect perceptions of partner behaviors (e.g., Fletcher, 2015; Pusch et al., 2020; Tucker & Anders, 1999), no previous work to our knowledge has examined how individual differences in motivation may shape perceptions of partner characteristics. Given that previous studies have suggested that individuals are often motivated to see their partners in a positive light, it is likely that the type of motivation and goal orientation an individual possesses is an important variable affecting exactly how he or she maintains a positive image of the partner. Because individuals high in approach goals are focused on approaching positive outcomes and incentives in their relationships, we predict that individuals high in approach goals will think of their partner in terms of the positive features they possess. Thus, we should see that individuals high in approach goals can more readily generate traits their partner possesses and can quickly confirm the presence of positive traits in their partner. On the other hand, we predict those high in avoidance goals will think of their partners in terms of their lack of negative features. Since individuals high in avoidance are vigilant to threat, we expect they will think of all the negative red flags their partner lacks and thus will more easily generate traits their partner lacks and will be faster to confirm that negative traits are not descriptive of their partner.

Study 1 Overview

In Study 1, participants completed a reaction time task in which they were asked to respond to the prompt “My partner is ____” or “My partner is not ____”. It was expected that those high in approach goals would be faster at generating positive characteristics that their partner possesses (faster responses to the “is” prompt) than those with weaker approach goals; whereas those high in avoidance goals would be faster at generating negative characteristics that their partner lacks (faster at the “is not” prompt) than those with weaker avoidance goals.

Method

Participants

Participants consisted of 73 individuals from a university community subject pool (50.7% of the participants were male and 49.3% were female) who indicated via a university pre-screening questionnaire that they were currently in romantic relationships. Participants were recruited for approximately 4.5 months with the goal of recruiting as many participants as possible with the community sample that fit the criteria prior to the end of the quarter. No interim data analysis was conducted during data collection.Footnote 1 The ethnicity distribution was as follows: 63.0% Caucasian, 13.7% Asian, 8.2% Hispanic, 2.7% Hawaiian or Pacific Islander, 1.4% Chicano, 1.4% other, and 9.6% claiming more than one race. Participants ranged in age from 18 to 24, with an average age of 19.7. Participants reported having been in their current relationship an average of 16.23 months (SD = 12.68). Upon completion of the study, participants were paid $10 for their participation.

Materials

Approach and Avoidance Social Goals Measure (Elliot et al., 2006)

This 8-item measure assesses levels of approach (e.g., “I will be trying to enhance the bonding and intimacy in my close relationships this quarter”) and avoidance (e.g., “I will be trying to make sure that nothing bad happens to my close relationships this quarter”) social goals. Items are responded to with a 7-item scale (1 = not at all true of me, 7 = very true of me). The alpha reliabilities of the social approach (M = 5.73, SD = 1.20) and avoidance scales (M = 5.33, SD = 1.30) were .86 and .75, respectively, and the correlation between approach goal and avoidance goal scores was 0.69 (p < .01).

Is/Is Not Reaction Time Task

Participants completed a reaction time task in which they alternated responding to the prompt “My partner is _____” and “My partner is not _____” on a computer (using the DirectRT program). The starting prompt was counterbalanced so that some participants began with the “is” prompt and others began with the “is not” prompt. They were told to fill in the blanks, but were not given specific direction as to what types of words or phrases should be typed into the blanks. Reaction times were measured from the time the prompt was displayed to the time the participants submitted an answer. All participants completed 30 blanks total (15 “My partner is _____” and 15 “My partner is not _____”). Average reaction time scores were calculated separately for the “is” (M = 9.56 s, SD = 5.86 s) and “is not” (M = 13.22 s, SD = 7.73 s) prompts and scores above or below 3 standard deviations from the mean response time were excluded from analysis (this included 21 (4 “is”, 17 “is not”) dropped scores across all participants, ~1% of trials). The “is” and “is not” reaction times were significantly correlated with each other (r = .496, p < .001).

Relationship Satisfaction

To assess relationship satisfaction, participants completed 5 items from the Rusbult et al. (1998) Investment Model Scale designed to measure satisfaction. All questions were answered using a 9-point (1 = do not agree at all, 9 = agree completely) rating scale (M = 7.69, SD = 1.35, α = .915) and included items such as “Our relationship makes me very happy”.

Procedure

Participants came into the lab and completed an informed consent form before theyFootnote 3 moved on to complete a series of questionnaires including the Approach and Avoidance Social Goals Measure (Elliot et al., 2006). After completion of these questionnaires, participants completed the Is/Is not reaction time task after receiving initial instructions.

Results and Discussion

Descriptive Statistics

Participants responded to 81% of the prompts with positive information (i.e. “is kind”, “is not mean”), 7% of the blanks with neutral partner information (i.e., “is a blonde”, “is not a brunette”), and 12% of blanks with negative information (i.e., “is lazy”, “is not motivated”). Thus, participants tended to point out the positive characteristics their partner possessed as well as the negative characteristics their partner lacked when completing this exercise. That is, most of the time participants were filling in the blank “My partner is____” with a positive trait and were filling in the blank “My partner is not____” with a negative trait, such that the overall meaning of the completed phrases was positive 81% of the time.

Approach and Avoidance Goals and Reaction Times

Partial correlations were computed between the “is” and “is not” reaction times and social approach goals, controlling for avoidance goals; and avoidance goals, controlling for approach goals (see Table 1). All correlations also control for the opposite “is” or “is not” reaction time (to account for individual differences in reaction time abilities). As predicted, approach goals were significantly negatively correlated with “is” reaction time; stronger approach goals were associated with faster reaction times on the “is” prompt (r = −.28, p = .018, 95% CI [−.432, −.128]). While the relationship between approach goals and the “is not” reaction time were in the predicted direction, the relationship was not significant (r = .21, p = .081, 95% CI [.036, .391]). Avoidance goals were significantly positively correlated with “is” reaction time; stronger avoidance goals were associated with slower reaction times on the “My partner is” prompt (r = .25, p = .036, 95% CI [.131, .380]). Although in the predicted direction, avoidance goals were not significantly correlated with the “is not” reaction time, (r = −.15, p = .208, 95% CI [−.382, .047]).

Table 1 Partial Correlations of Goals with Is and Is Not Reaction Times Controlling for Opposite Prompt Reaction Time
Full size table

These results provide some support for our initial hypotheses. Approach and avoidance goals seem to be associated with how individuals view current partners. Individuals high in approach goals are able to easily access information about the characteristics their partner possesses (usually positive characteristics). Individuals high in avoidance goals, on the other hand, are slower to think of the positive characteristics their partner possesses.

Is/Is Not Reaction Times and Relationship Satisfaction

We were interested in exploring whether the speed with which individuals can think of traits that either are true of their partner (“is” reaction times) or are not true of their partner (“is not” reaction times) was associated with relationship satisfaction. We utilized partial correlations that examined the relationship between the “is” or “is not” reaction times and relationship satisfaction, controlling for the opposite prompt reaction time. We found that neither the “is” (r = .09, p = .49, 95% CI [−.185, .303]) nor the “is not” (r = −.16, p = .20, 95% CI [−.410, .088]) reaction times were significantly associated with relationship satisfaction.

Study 2 Overview

The goal of Study 2 was to examine how approach and avoidance goals were associated with partner perceptions of positive and negative traits using a different methodology. In addition, we again examined how these differences may be associated with relationship satisfaction. We also utilized a more specific version of the social goals measure which focuses on the approach and avoidance goals for the romantic relationship of interest. Participants completed a reaction time task in which they responded to a series of traits and indicated whether each was or was not descriptive of their partner. It was expected that those high in approach goals would be faster to respond to positive partner traits (e.g., driven, considerate) as opposed to negative partner traits (e.g., lazy, mean). Those high in avoidance goals were expected to be faster to respond to negative partner traits.

Additionally, we wanted to examine whether there were any relationship characteristics associated with patterns of responses to certain types of traits (i.e., the accessibility of partners’ positive or negative traits). Thus, we examined whether faster responses to positively- vs. negatively-valenced words would be associated with relationship satisfaction.

Method

Participants

Participants included 96 individuals from a university student subject pool (20.8% of the participants were male and 79.2% were female) who were currently in romantic relationships (Mean length of relationship = 1.86 years, SD = 1.46) and were given partial class credit for participating in this lab study. Participants were recruited across a summer and fall quarter and sample size was based on the subject pool allocation at the university. No data were analyzed from this project until all data collection was complete and data collection ended at the end of the fall quarter.Footnote 4 Participants ranged in age from 18 to 48, with an average age of 19.53 (SD = 3.63). The ethnicity distribution was as follows: 55.2% Caucasian, 27.1% Hispanic, 13.5% Asian, and 4.1% Other.

Materials

Approach and Avoidance Relationship Goals Measure

In this study we modified the original approach and avoidance goals measure (Elliot et al., 2006) to focus specifically on one’s romantic relationship (replacing the term “close relationships” with “romantic partner” or “romantic relationship”). The scale measures levels of approach (e.g., “I will be trying to enhance the bonding and intimacy in my romantic relationship [over the next few months]”) and avoidance (e.g., “I will be trying to make sure that nothing bad happens in my romantic relationship [over the next few months]”) relationship goals. Items are responded to based on a 7-item scale (1 = not at all true of me, 7 = very true of me). This modified scale has been used in previous research (Impett et al., 2008). The alpha reliabilities of the relationship approach and avoidance scales were .88 and .72, respectively, and the correlation between relationship approach goal and avoidance goal scores was 0.66 (p < .01).

Relationship Satisfaction

To assess relationship satisfaction, participants completed the 7 item Hendrick (1988) Relationship Assessment Scale, which provided a new operationalization of relationship satisfaction compared to Study 1. All questions were answered using a 7-point (1 = very low satisfaction, 7 = very high satisfaction) rating scale (M = 5.94, SD = .85, α = .847) and included items such as “How well does your partner meet your needs?”

My Partner/Not my Partner Reaction Time Task

Using a procedure by Aron et al. (1991), participants were asked to respond to trait words as being descriptive of their partner (“my partner”) or not descriptive of their partner (“not my partner”) as fast as possible while still providing accurate responses. Participants completed this task for 103 traits (88 traits from the original Aron et al. (1991) task and 15 additional traits of interestFootnote 6). However, because not all words included were clearly “positive” or “negative” (e.g., frank, serious) all words were coded by three independent trained coders who rated each trait as “definitely a positive trait”, “definitely a negative trait”, or “neutral (neither positive nor negative or sometimes positive and sometimes negative). We focused on all traits that were rated by at least 2 of the 3 coders as definitely positive, definitely negative, or neutral. This resulted in the inclusion of 86 total traits in the analysis (40 positive, 35 negative, 10 neutral). In addition, because not all traits were suspected to be highly relevant to college relationships (e.g., “spendthrift”, “cordial”), we had coders rate the extent to which each trait was relevant and important to college romantic relationships (1 = Not at all important/Very unimportant, 4 = Somewhat Important, and 7 = Very Important). This allowed us to also focus on the top ten highest-rated positive and negative traits that were likely to be relevant to the relationships of the people in our sample. Reaction times were recorded through the DirectRT program. Reaction times that were more than three standard deviations away from the average were dropped from analyses (to control for lapses in attention). In addition, we calculated composite scores for use in the analyses that reflect the average reaction time of all positive trait words, the average reaction time of all negative trait words, and the average reaction time of all neutral words. The specific traits for each category were as follows (and traits that were selected as a top 10-relationship relevant word are marked with an *):

  • Positively-valenced words: active, ambitious, amusing*, appreciative*, attentive, cheerful, confident, congenial, considerate*, cordial, creative, driven*, faithful*, family-oriented, generous, good communicator, good-tempered*, good with kids, honest*, independent, inventive, loyal*, mature, neat, observant, productive, prompt, respectful*, responsible, self-reliant, sensible, social, sophisticated, sympathetic, tactful, tender, tolerant, trusting*, understanding, versatile

  • Negatively-valenced words: aggressive*, antisocial, argumentative*, boastful, cold, cowardly, crude, deceptive, domineering, dull, foolhardy, hot-headed*, ill-mannered*, inconsiderate*, irrational*, irritable, jealous, lazy, materialistic, mean*, nosey, prejudiced, scornful, self-centered*, showy, spendthrift, spiteful, superficial, suspicious, tactless, unfair*, unreliable*, vain, weak, worrier

  • Neutral words: foolish, frank, naïve, persistent, persuasive, sarcastic, serious, shy, submissive, systematic

Procedure

Participants were invited through a university subject pool to participate in a study on how people perceive romantic partners. Each participant first completed a series of initial questionnaires, which included the approach and avoidance relationship goals measure and relationship satisfaction measures. Next, they completed the partner/not my partner reaction time task.

Results and Discussion

Approach and Avoidance Relationship Goals and Positive Vs. Negative Trait Reaction Times

Partial correlations were computed between average positive and negative trait reaction times and relationship approach goals, controlling for avoidance goals; and relationship avoidance goals, controlling for approach goals. All correlations also control for the opposite prompt (either positive or negative) reaction time (to account for individual differences in reaction time abilities). Finally, after running the full trait analyses, we also focused on the more highly-relevant positive and negative traits as we expected the associations may be the strongest amongst these more relevant traits.

Consistent with our hypotheses, we found that approach goals were negatively correlated with positive trait words reaction time such that those who had stronger approach goals also tended to have faster reactions to the positive trait words when thinking about their partner (r = −.22, p = .036, 95% CI [−.438, −.060). The same pattern was seen when focusing on the top 10 relationship-relevant positive traits, though the association was even stronger (r = −.37, p < .001, 95% CI [−.585, −.016]). Approach goals were not significantly associated with reaction times for negative traits when looking at all negative traits (r = .15, p = .159, 95% CI [−.046, .309]). However, strong approach goals were positively correlated with negative trait reaction time when focusing on the most relationship-relevant negative traits (r = .28, p = .008, 95% CI [.001, .457]), suggesting individuals with high approach goals were slower at responding to the most relationship-relevant negative trait words.

Avoidance goals were significantly negatively correlated with negative trait reaction time and thus stronger avoidance goals were associated with faster reaction times on the negative trait word prompts (r = −.21, p = .045, 95% CI [−.438, .016]). A similar association was seen between avoidance goals and the top ten most relationship-relevant negative traits (r = −.23, p = .034, 95% CI [−.440, .047]). Avoidance goals were not significantly associated with positive trait reaction times—whether looking at all positive traits (r = .18, p = .101, 95% CI [−.076, .374]) or just the most relationship-relevant positive traits(r = .16, p = .13, 95% CI [−.143, .394]).

As a comparison, we also examined how approach and avoidance relationship goals would be associated with neutral trait ratings, as we would expect no relationships here. We again used partial correlations which controlled for the opposite goal as well as an average reaction time across the positive and negative trials. Consistent with our prediction, we found that approach goals showed no significant association to neutral word reaction times (r = .02, p = .893, 95% CI [−.187, .202]) and that avoidance goals showed no significant association to neutral word reaction times (r = .11, p = .301, 95% CI [−.108, .336]).

Partner Trait Reaction Time (Accessibility) and Relationship Satisfaction

We were interested in determining whether the speed with which individuals can make these trait decisions (an indicator to us of how accessible these traits are in their partner conceptualizations) was associated with an important relationship outcome: relationship satisfaction. We conducted a series of partial correlations in which we examined the association between average reaction time to positive (and then negative) trait reaction times and relationship satisfaction (controlling for general differences in reaction time abilities by controlling for the opposite reaction time).Footnote 7 Reaction times to positive traits were significantly negatively correlated with relationship satisfaction (r = −.30, p = .005, 95% CI [−.46, −.10]). This effect was even stronger when we focused on the top ten most relevant positive traits (r = −.34, p = .001, 95% CI [−.527, −.094]).) In other words, responding more quickly to positive words was associated with greater relationship satisfaction. Reaction times to negative words was not significantly associated with relationship satisfaction whether looking at all negative traits (r = .136, p = .20, 95% CI [−.036, .299]) or the top ten most relevant negative traits (r = .10, p = .34, 95% CI [−.098, .286]).

General Discussion

These studies suggest that approach and avoidance social goals play a role in shaping the perceptions of romantic partners and the accessibility of partner traits. Study 1 found that individuals high in social approach goals conceptualize their partner in terms of their positive traits (as these individuals were able to quickly name characteristics that their partner possessed). Individuals high in avoidance goals, on the other hand, have a harder time generating positive traits (as evidenced by slower reaction times). These differences in reaction time seem to tap into differences in cognitive accessibility. It is easier for those high in approach goals to think of all of the positive features a partner possesses, whereas for those high in avoidance goals this is a harder task. No relationship was found between the speed of responding to these prompts and relationship satisfaction.

Study 2 builds on the findings from the first study, again demonstrating the influence of goals in partner trait processing. We once again found support for the idea that individuals with strong relationship approach goals tend to view their partner in terms of positive traits they possess and they were faster to respond to questions of whether their partner possessed a variety of positive traits. Individuals higher in avoidance goals are slow to respond to these questions, and in Study 2, they were actually faster to respond to the prompts asking if their partner possessed negative traits. These results are especially interesting given past research into partner perception. Although previous work by Murray et al. (1996) suggests that individuals are motivated to form positive impressions of their partners, our results suggest that “positive impressions” can take different forms (which may lead to different outcomes). Although individuals high in both approach and avoidance motivation may view their partners in a largely positive light, strong approach goals seem to be associated with framing the partner in light of the possession of a number of positive traits whereas we found some evidence that avoidance goals may be associated with greater ease in thinking of the partner’s lack of a number of negative traits.

Study 2 also suggested a way in which processing partner-relevant information may be important. Faster responses to positive traits were found to be closely linked to relationship satisfaction. Although our results can’t speak directly to causal links, it seems plausible that being able to easily access information about the positive aspects of your partner would bolster your opinions of your relationship satisfaction (as you may more easily call to mind aspects of your partner that you like). It is also possible that having greater relationship satisfaction prompts you to more easily think about positive aspects of your partner. However, these results were different from those found in Study 1; while the pattern of correlations was the same, the effects were not significant. Although it is difficult to speculate on null findings, it may be that the methods of Study 2 (responding to actual positive vs. negatively valenced words that were presented) involved a more precise measure (compared to generating novel words to the is/is not prompt). Additional research will be needed to better understand how goal-driven partner trait processing may be important in the context of close relationships. Previous research (Murray et al., 1996; Neff & Karney, 2005) has demonstrated that the way individuals process and perceive information can have important influences on various outcomes (both within and outside of a romantic relationship). Thus, it is possible goal-driven trait processing could affect a number of outcomes (even outside of satisfaction). For instance, if high approach individuals tend to view their partner in terms of their positive traits, are they more willing to overlook shortcomings? If highly avoidance-oriented individuals are looking for partners who lack negative characteristics or character flaws, are minor mistakes or one-time instances of negative behavior likely to overshadow all of their positive features?

While this study focused on the processing of romantic partner information, we believe the findings could extend to several other areas. For instance, in a work setting, supervisors who are high in approach motivation may focus on positive aspects of a potential employee that signal that they can provide special assets and skills to the company. In contrast, supervisors high in avoidance motivation may be less likely to think of job candidates in association with positive traits and may focus more on evaluating any potential red flags or skill deficits that could be detrimental to job performance. These differences could have important consequences for the types of employees hired. In addition, the influence of goals could extend into how supervisors viewed employees (altering their perceptions and employee evaluations). Employees could potentially receive very different ratings if an evaluator was focusing on all of their achievements and positive characteristics vs. if he or she was looking for the presence of any faults or failures. Aside from work settings, these findings could likely generalize to other situations in which someone must evaluate other individuals or items (professors during admissions processes, doctors selecting medical treatments, etc.).

While our studies have suggested cognitive differences in how individuals high in approach vs. avoidance motivation process partner trait information, there are limitations of these studies. For one, we did not test “objective” levels of individual traits so it is hard to know if some of the differences observed could be due to differences in possession of positive and negative character traits as opposed to just how individuals perceive their partners. While our participants tended to report high relationship satisfaction and thus focused more on positive perceptions of their partner, it would be interesting to examine how goals might influence trait salience in a sample that reports lower relationship satisfaction and has negative partner perceptions (e.g., are those higher in avoidance goals more quick to nominate negative traits their partners possess whereas those higher in approach goals might be quick to nominate positive traits their partners lack?). Furthermore, both studies utilize a quasi-experimental design and thus it can still be hard to know if approach and avoidance motives are really driving the differences observed. Future work could seek to manipulate approach and avoidance goals to determine if manipulating these orientations leads to changes in trait response times. Other potential influences (such as the role of emotion or daily relationship satisfaction) could also be examined to determine if trait reaction times shift due to short-term changes in affect and relationship dynamics. In addition, it is possible that partner traits and behaviors might be able to alter approach and avoidance motivational states and future research could explore this idea experimentally. Both studies were conducted prior to current norms for power analysis, so our sample size was lower than might be ideal and included a fairly young and primarily college-aged sample. It is crucial for future studies to confirm causality and examine long-term implications. If avoidance motivation leads to less attention being placed on their positive attributes and perhaps greater attentiveness to the shortcomings of a partner, this could be setting individuals up for failure in relationships. A focus on the positive aspects of one’s partner and relationship, on the other hand, may prove to buffer approach-oriented individuals from everyday partner shortcomings that are likely to emerge over the course of a relationship.

These studies point to the power of motivation to shape romantic partner perceptions. Future work can build upon the present study by continuing to examine the influence of motivation on partner perception as well as other areas likely to be impacted by approach and avoidance motives and goals.