An Open Path to the Future: Perceived Financial Resources and School Motivation

Abstract

One contributing factor to gaps in academic achievement may be that some students perceive long-term educational goals, such as college, as financially out of reach, which can make schoolwork feel meaningless even several years before college. However, information that leads students to perceive that the financial path to college is open for them (i.e., need-based financial aid) can increase school motivation. Two classroom-based field experiments expand this area of theory and research. Early adolescent students who were randomly assigned to receive information about need-based financial aid (open path condition) showed greater school motivation than those who were randomly assigned to a control condition, specifically if they came from low-asset households. In a second exploratory experiment, the open path effect was mediated by an increased likelihood that students envision a future career that includes college (education-dependent identity). Implications for the study of identity and disparities in academic achievement are discussed.

Keywords

socioeconomic status assets school motivation identity

Students from families with more financial resources tend to outperform students from families with fewer financial resources, even when they attend the same schools (Baum, Ma, & Payea, 2010; Destin, Richman, Varner, & Mandara, 2012; Reardon, 2011). Socioeconomic disparities in achievement are especially pronounced between students from families that are low versus high in wealth or financial assets (Williams Shanks & Destin, 2009; Williams Shanks, Kim, Loke, & Destin, 2010). One possible reason that students from low-asset families are at risk of academic disengagement is that high college costs make the long-term goal of college feel out of reach, rendering current effort in school meaningless (Bettinger, Long, Oreopoulos, & Sanbonmatsu, 2012; Destin & Oyserman, 2009; Oreopoulos & Dunn, 2013). Identity-based motivation theory provides a framework to investigate a proposed pathway from financial assets to school outcomes for adolescents that is grounded in the premise that perceptions of financial barriers or resources, desired future identities, and intentions to engage in schoolwork are all malleable and interrelated (Oyserman & Destin, 2010).

According to identity-based motivation theory, people are motivated to engage in tasks that feel congruent with their future identities (i.e., the person that they feel likely to become). This is especially true during early adolescence, as young people begin to develop a meaningful sense of who they may become in the future (Dunkel, 2000; Meeus, Van De Schoot, Keijsers, Schwartz, & Branje, 2010; Oyserman, Bybee, & Terry, 2006). As a result, adolescents are more likely to work toward long-term academic goals if images of their future identity feel connected to careers that require a college education (education-dependent identity) rather than careers that do not require a college education (Destin & Oyserman, 2010). The proposed model builds upon the identity-based motivation theory framework to investigate the role of perceived financial resources. Students from low-asset families face many real financial barriers that separate them from college and close the pathway to their education-dependent identity, an identity that might otherwise be a source of motivation to engage in schoolwork. However, perceptions of financial circumstances as well as identities are dynamic and malleable meaning that they can shift from moment to moment depending upon available cues. Thus, socioeconomic disparities in school outcomes may be attenuated when contextual cues make college feel more financially accessible for students from low-asset families. In the proposed model, a cue that the pathway to college is “open,” which focuses on potential financial resources such as need-based financial aid, should increase the likelihood that adolescents from low-asset families feel motivated to engage in schoolwork, potentially by increasing the perceived accessibility of an education-dependent identity (see Figure 1). In the current research, two field experiments first evaluate whether subtle information that cues an open path can increase students’ intentions to engage in schoolwork, particularly for students from low-asset families. Second, an exploratory test of the proposed mechanism evaluates whether evidence supports that the increased school motivation comes from imagining a salient education-dependent identity when the path to college feels open.

Figure 1.

Proposed model of identity-based school motivation incorporating financial obstacles and resources.

School Motivation and Financial Assets

Motivation, or the self-directed urge to take action toward a goal (see Bandura, 2001; Higgins, 2012; Ryan & Deci, 2000), is a key determinant of academic engagement and achievement for adolescents. Several classic frameworks have demonstrated a number of important factors and constructs that facilitate motivation, such as self-efficacy, goals, expectations, and values (e.g., Bandura, 2001; Eccles & Wigfield, 2002; Grant & Dweck, 2003; Perreira, Fuligni, & Potochnick, 2010; Ryan & Deci, 2000). However, these dominant approaches are not fundamentally grounded in a nuanced understanding of an individual’s shifting perception of their context (e.g., perception of financial barriers or resources), which is likely to contribute to socioeconomic disparities in achievement and motivation. Identity-based motivation theory, on the other hand, not only asserts that people pursue goals that feel congruent with their future identities, as described above. It also incorporates insight from theories of situated cognition (e.g., Smith & Semin, 2007), which suggests that an individual’s perception of their social context, including financial barriers or resources, is inherently malleable (Oyserman & Destin, 2010). At the same time, identity-based motivation theory shows that a person’s own identity is also malleable and continually cued and re-created based on fluctuating contextual affordances. In other words, construals of both social contexts and identities are malleable, and together, the way that they are cued determines the types of behaviors that people choose. For example, in a series of lab experiments, Oyserman, Destin, and Novin (2015) showed that college students were more likely to engage in schoolwork after being randomly assigned to learn that college is a place where students are likely to succeed and considering their own possible success (or a place where students are likely to fail and considering their own possible failure).

Because of its focus on the perception of context, the identity-based motivation framework can be naturally extended to explore the effects of students’ perceptions of financial barriers or resources. Adolescents who grow up surrounded by plentiful financial assets encounter abundant contextual cues that continually bring to mind a high status and high achieving future identity, which is tied to school motivation. Adolescents in contexts with fewer financial assets, on the other hand, are less likely to encounter everyday signals of financial and academic success around them, decreasing the salience, accessibility, and motivational potential of their own successful future identities. When meaningful future identities related to achievement and college are not cued in context and become distant and abstract, academic motivation is likely to falter as students become less likely to envision pathways to success or make plans to engage in goal-directed behaviors.

In the prevailing academic culture, students often rely on education-dependent identities as a specific type of motivating future identity. However, adolescents from low-asset families encounter everyday contextual cues suggesting that college is less likely for them. Specifically, they realize that college will be very expensive, and they often even overestimate the costs of college tuition (e.g., Grodsky & Jones, 2007; Horn, Chen, & Chapman, 2003; McWhirter, Valdez, & Caban, 2013). Adolescents and parents must inevitably look beyond their household’s regular annual income to greater sources of financial assets as they begin to compare their family’s resources with the amounts necessary to pay for college. Despite fluctuations in the real estate market, housing wealth is most often a family’s greatest financial asset and an especially important predictor of whether or not children from families with fewer financial resources attend college (Lovenheim, 2011). So, particularly for adolescents from families who seem less likely to own their home, other cues may be important to encourage the perception that the pathway to college is open and to foster the development of motivating education-dependent identities.

Open Pathways to Future Success

As the financial pathway toward college begins to feel relatively open or closed, current schoolwork becomes more or less meaningful, respectively (Destin & Oyserman, 2009). Studies have provided initial evidence that small amounts of information can improve school motivation for students by leading them to perceive an open financial path to future college enrollment. In a study with over 1,600 high school students, Oreopoulos and Dunn (2013) observed an increase in expectations for college attendance among those who were randomly assigned to receive information related to need-based college financial aid. In another study, Destin and Oyserman (2009) increased school motivation of seventh-grade students who were randomly assigned to receive information about need-based college financial aid. Compared with students in control groups, where they received no information about college or only information about tuition costs, those who learned about financial aid made plans to spend more time on schoolwork.

Previous studies of open path interventions, however, focused exclusively on students from poor schools and neighborhoods and did not directly explore a proposed psychological process. Implications for theory, including the role of an adolescent’s socioeconomic context as a possible boundary condition and the role of an education-dependent identity as a possible mediator, remain unclear. If students become more motivated because perceived resources facilitate an identity shift, then wealthy students, who encounter everyday cues that resources are abundant, should normatively perceive an open path to college, and information about need-based financial aid should not influence their intentions to engage in schoolwork. Furthermore, a family’s apparent level of stable wealth and assets should matter more than less stable factors such as income when considering a motivational pathway that is related to facing the large financial barrier of college costs.

A competing hypothesis might expect open path information to increase motivation for all students regardless of financial circumstances. A universal effect for all students would suggest that certain prompts simply serve as a prime, increasing motivation by activating money, college, or achievement-related goals (e.g., Bargh, Gollwitzer, Lee-Chai, Barndollar, & Trötschel, 2001; Vohs, Mead, & Goode, 2006). A universal effect of open path information might also reflect a social desirability bias for all students to show more positive academic intentions after hearing about need-based financial aid opportunities. The proposed model, on the other hand, suggests an asset-specific hypothesis. If financial aid information does, in fact, open pathways to an education-dependent identity that would otherwise be closed, then a person-in-context interaction effect should emerge where only students from families that appear to be low in financial assets respond positively to need-based financial aid information. The school motivation of students from homes that seem abundant with financial assets, who do not regularly perceive evidence that paying for college will be difficult or impossible, should not be influenced by information intended to open a financial pathway to higher education.

Because the proposed motivational process depends on how students perceive financial resources and their own possible future identity when situated in their everyday school context, a classroom-based field study provides the most ecologically valid method of research. Furthermore, a controlled field experiment provides the most appropriate method to directly assess causation and determine whether cues that signal an open or closed path to college affect student identities and motivation. Two field experiments test the proposed effects of an open path experimental manipulation aiming to increase intentions to engage in schoolwork for middle school students from low-asset families. The samples match the socioeconomic criteria necessary to directly test two hypotheses. First, the asset-specific hypothesis suggests that adolescents from a socioeconomically diverse sample who are randomly assigned to an open path condition, where they receive information about need-based college financial aid, will show more school motivation only if they seem to come from households that are low in financial assets (i.e., low-asset). Second, an exploratory experiment will seek further evidence suggesting that low-asset adolescents will show more school motivation after an open path experimental manipulation because of the increased likelihood that they will subsequently imagine an education-dependent future identity.

Study 1

Seventh-grade students from two middle schools participated in a classroom-based field experiment, where they were randomly assigned, between-subjects, to an open path treatment condition or a control condition to examine effects on intentions to engage in schoolwork. The study sites were selected because of their within-school socioeconomic diversity, which was appropriate to test the asset-specific hypothesis: The open path treatment will lead to increased school motivation for students who appear to come from low-asset family backgrounds, but not for students who appear to come from high-asset family backgrounds. Apparent homeownership was the key indicator of wealth and assets, which is related to other indicators of socioeconomic status (SES) such as income, occupation, and education, but more relevant for financing costly endeavors such as a college education. Furthermore, homeownership is uniquely predictive of the development of future identities for children (Elliott, Choi, Destin, & Kim, 2011; Williams Shanks et al., 2010). As such, because students who live on neighborhood blocks with low rates of homeownership and in apartments (i.e., low-asset students) are more likely to perceive that their families are low in financial assets, the asset-specific hypothesis predicts that they should show a particularly strong benefit to receiving open path information about need-based financial aid. On the other hand, students in the same schools who live on neighborhood blocks with high rates of homeownership and in houses (i.e., high-asset students) should not be influenced by experimental condition.

Method

One hundred fifty-six seventh-grade students (53.9% male; 70.4% White, 17.1% Black, 3.9% Latino/a, 6.6% Asian/Asian American, 2.0% Other) from two socioeconomically diverse middle schools (average of 44.25% students across schools were eligible for free or reduced-price lunch) located just outside major Midwestern urban centers participated in a classroom-based field experiment. After obtaining parental consent, half of the participants were randomly assigned to an open path condition, and the other half of participants were randomly assigned to a control condition. During the normal school day, a member of the research team entered participating classrooms and was introduced by the teacher as a guest from a local university providing some information and distributing a short questionnaire to learn about the everyday lives of middle school students. Next, approximately half of the participants received a one-page information sheet containing real information about need-based financial aid for college (open path) followed by a short writing activity where students answered questions about the information to ensure that they processed the material (see Appendix A for materials). The other half of the participants, randomly assigned to the control condition, only received an introduction to the study of students’ everyday lives before beginning the questionnaire. In one school, only 1 day was allotted for data collection, and therefore, a smaller number of students were able to participate (26.3% of participants). Participants were randomized to condition by classroom, and all students in the same classroom were assigned to either the open path or control condition. In the second school (73.7% of participants), more days were designated for data collection. These participants received individualized packets, which randomly assigned students within the same classrooms to different conditions in order to ensure that participants were equally likely to be assigned to the control or open path condition regardless of their day of participation.¹

To evaluate the effect of experimental condition on intentions to engage in schoolwork, all participants completed a “general questionnaire of ordinary behaviors of middle school students,” which was designed to capture planned school behaviors after exposure to experimental treatment materials and has been used in other studies of school motivation for middle school students (see Destin & Oyserman, 2009). To reduce response bias, participants were instructed to answer as honestly as possible and that their responses would not be linked to their names, shared with their teachers, or used to influence their grades. The survey asked how much time students planned to spend that night engaged in everyday activities (0 = less than an hour, 7 = all night long). The mean score of two critical activities provided a measure of schoolwork (“How much time do you plan to spend doing homework tonight” and “How much time do you plan to spend reading or studying tonight?”; $\bar{X}$ = 1.60, SD = 1.24, α = .66). These two questions were embedded within questions about several filler activities to ensure that students fully considered their evening (e.g., watching television, listening to music, non-academic computer activities, playing sports).

Finally, participants provided their gender, race/ethnicity, and a subsample (35.3% of the participants) provided consent to link their questionnaire responses to their home address. Using the U.S. Census American FactFinder tool, home addresses were linked to block-level data and coded for whether each student lived in a community that was predominantly inhabited by homeowners (high-asset) or renters (low-asset). Participants from the rest of the sample did not provide permission to use identifying information, and instead, they self-reported whether they lived in a house or an apartment (76.9% of the total sample was high-asset and 23.1% was low-asset). Neither measure provides a perfect indicator of a family’s assets because blocks are not completely homogeneous, and living in a house rather than an apartment does not always indicate homeownership. However, the measures were selected because an early adolescent’s perception of their family’s assets is more likely to depend on their perception of the level of wealth on their block and the size of their home, rather than a family’s actual balance sheet. Thus, as described above, these measures that were selected to capture a young person’s subjective sense of their family’s assets are related to family income, education, and occupation, but they provide a more stable indicator of children’s perception of their family’s resources, and they uniquely predict many positive outcomes for children including the development of an education-dependent identity (Elliott, Destin, & Friedline, 2011; Sherraden, 1991; Williams Shanks et al., 2010). Immediately after completion of the questionnaire, all students in both conditions received an open path handout with information about need-based financial aid opportunities as a debriefing.

Results and Discussion

Because of the socioeconomically stratified nature of most metropolitan areas, it is rare to find a community that is inhabited with equal numbers of families from low- and high-asset backgrounds (Jargowsky, 1996; Quillian, 2012). The two communities that were selected for this study were chosen because they did contain a significant amount of socioeconomic diversity; however, the groups of low- and high-asset participants were still unbalanced by nature. Therefore, prior to conducting primary analyses, a nonparametric Levine test was conducted, which assesses homogeneity of variance when groups have different sample sizes due to the nature of the population and research question (Nordstokke & Zumbo, 2010). First, each participant received a ranking according to their response on the dependent variable. Next, an absolute difference was calculated for each participant indicating the difference of their ranking from the mean ranking within the group that they belonged (low- or high-asset). Finally, a one-way ANOVA tested whether there was a significant difference between groups in their average absolute difference in ranking from the mean. The test was not significant, indicating that the assumption of homogeneity of variance was supported, F(1, 154) = .01, p = .929.

Randomization to experimental condition was successful such that low- and high-asset participants were equally likely to be in the control or open path conditions (see Table 1). Participants also did not statistically differ across conditions in gender, race/ethnicity, or school. Even so, analyses statistically accounted for any potential effects of gender, race/ethnicity, and school on intentions to engage in schoolwork by including them as covariates in all analyses.² All effects were consistent whether covariates were included or excluded from analysis.

Table 1.

Distribution of Participants Across Experimental Conditions by Assets in Study 1.

	Control	Open path
Low-asset	15	21
High-asset	59	61

Note. χ²(1, N = 156) = .63, p = .429.

To assess the asset-specific hypothesis, a 2 × 2 ANCOVA tested the effect of experimental condition (open path, control) and assets (low, high) on schoolwork. There were no main effects of experimental condition, F(1, 143) = 1.36, p = .245, or assets, F(1, 143) = 2.35, p = .128, but rather the hypothesized asset-specific interaction effect was significant, F(1, 143) = 6.31, p = .013, d = .40, 95% confidence interval [CI] = [−2.02, −.24]. As shown in Figure 2, tests of simple effects supported the hypothesis. Students from low-asset households showed significantly greater intentions to engage in schoolwork in response to open path, financial aid information ( $\bar{X}$ = 2.48, SD = 1.59) compared with low-asset students in the control condition ( $\bar{X}$ = 1.63, SD = 1.13; p = .037, d = .62). As predicted, experimental condition did not affect students from high-asset families (control $\bar{X}$ = 1.50, SD = 1.22; open path $\bar{X}$ = 1.33, SD = 1.02; p = .175). Furthermore, low- and high-asset students did not significantly differ in intentions to engage in schoolwork in the control condition (p = .583), but information about need-based financial aid did influence students differently based on their families’ financial background. Low-asset students in the open path condition were more motivated than high-asset students in the open path condition (p = .003, d = .86) or control condition (p = .052, d = .69). Students from low-asset families who were randomly assigned to the open path condition subsequently planned to engage in more schoolwork than did any other students in the study.³

Figure 2.

Asset-specific effect of open path treatment condition on intentions to engage in schoolwork in Study 1.

Additional analyses incorporating planned leisure activities

The open path manipulation was not expected to influence the amount of time that students planned to spend on a mean composite of planned leisure activities (i.e., television, music, videogames, computer, sports; $\bar{X}$ = 1.71, SD = 1.21, α = .64) due to a wide range of individual variation in how students spend their time outside schoolwork. An ANCOVA was conducted substituting planned leisure activities for planned school activities as the dependent variable, which, as expected, did not find an effect of experimental condition, F(1, 143) = .87, p = .354, assets, F(1, 143) = 2.38, p = .126, or the interaction between condition and assets, F(1, 143) = .002, p = .967.

Even so, an alternative composite measure of school motivation was created, which took into account leisure behaviors by subtracting mean planned leisure activities from mean planned school activities ( $\bar{X}$ = −.14, SD = 1.35). In further support of the asset-specific hypothesis, the same interaction effect between experimental condition and assets that was observed on the primary measure of schoolwork was also found on the alternative composite measure, which incorporated leisure activities, F_condition(1, 143) = 3.27, p = .073; F_assets(1, 143) = .00, p = 1.000; F_interaction(1, 143) = 4.86, p = .029, d = .40, 95% CI = [−2.18, −.12]. Low-asset students scored significantly higher on the alternative composite measure, which took into account planned leisure activities, if they were randomly assigned to the open path condition rather than the control condition ( $\bar{X}$ _diff = 1.02, p = .023). Other simple effects were not significant (condition effect for high-asset students p = .695; comparison of low-asset students in the open path condition to high-asset students in open path p = .109, or control p = .208 conditions).

Additional analyses with an alternative socioeconomic moderator

A final set of additional analyses was conducted to investigate whether the measure of family financial assets was indeed a better moderator of the open path effect than a more general measure of SES, such as family household income. For the subsample of participants whose data were linked to home addresses, the median household income of their census tract was collected from the American FactFinder tool, and participants were coded as high- or low-income based on whether their census tract’s household income was above or below the median of the sample (Mdn = 52,923.00, SD = 18,331.90). An analysis that substituted income for homeownership did not find any main effects nor an interaction effect on intentions to engage in schoolwork, condition F(1, 51) = .83, p = .367; income F(1, 51) = .78, p = .381; interaction F(1, 51) = .66, p = .421.⁴

To ensure that the lack of effect with income as a moderator was not due primarily to the reduced power of the subsample, a parallel analysis with the measure of assets included only on the same subset of participants whose data were linked to home addresses (36.4% low-asset, 63.6% high-asset). As expected, a diminished yet consistent interaction effect emerged that matched the full sample, F_condition(1, 51) = 1.66, p = .203; F_assets(1, 51) = .73, p = .396; F_interaction(1, 51) = 3.03, p = .088, d = .37, 95% CI = [−2.89, .21]. Again, low-asset students showed more school motivation if they were randomly assigned to the open path condition rather than the control condition, at a marginal level ( $\bar{X}$ _diff = 1.17; p = .058, d = .76), or compared with high-asset students in the open path condition ( $\bar{X}$ _diff = 1.00.; p = .037, d = .67). Other simple effects were not significant (condition effect for high-asset students p = .718; comparison of low-asset students in the open path condition with high-asset students in the control condition p = .140).

In summary, students from low-asset families showed intentions to spend more time on schoolwork when they were randomly assigned to the open path condition where they learned about need-based financial aid for college, rather than a control condition. Study 1 supported the asset-specific hypothesis, that open path information improves school motivation for adolescents who live in contexts where they perceive fewer family financial assets. For higher asset adolescents within the same schools, however, financial aid information does not necessarily influence perceived opportunities to pay for college, which should already feel open due to home and neighborhood influences. The open path was especially important for students from low-asset rather than low-income families illuminating a potentially efficient and effective approach to reducing disparities in achievement tied to persistent and staggering disparities in wealth and assets. As predicted, an experimental prime that reframed students’ perceptions of potential economic barriers facilitated the drive to perform relevant goal-directed behaviors. These behavioral intentions matter because student motivation and homework completion consistently predict actual school achievement (Cooper, Robinson, & Patall, 2006; Zimmerman & Kitsantas, 2005). In a controlled field experiment, the study illustrated that adolescents who live in contexts with the least financial assets benefit most from information that leads students to perceive an open path to college.

Next, Study 2 drew from an exclusively low-asset sample to replicate the effects of an open path experimental manipulation and to explore whether they are mediated by increasing the likelihood that low-asset adolescents envision an education-dependent identity and describe a future that is connected to education. Although Study 1 supported the idea that need-based financial aid information helps to motivate low-asset adolescents in particular, studies have not yet explored the psychological mechanism or shown the proposed direct connection to an education-dependent identity as the active link to school motivation. Study 2 also sought to strengthen the conclusions of Study 1 in three other ways in addition to replicating the open path effect with an exclusively low-asset sample and exploring the proposed mechanism. First, Study 2 expanded the scope of open path effects by exploring a more conventional dependent measure related to specific academic grade goals that students set for themselves (see Zimmerman, Bandura, & Martinez-Pons, 1992) rather than behavioral intentions. Second, Study 2 compared the effects of an open path condition to a closed path condition that also led students to think about college, rather than a no information control, in order to verify that effects depend on the perception of an open path rather than simply an increased salience of college in general. Last, participants in Study 2 provided consent to collect a measure of their prior academic achievement from school administration to statistically control for prior achievement, which is more closely connected to academic grade goals than the Study 1 covariates of race/ethnicity, gender, and school.

Study 2

The purpose of Study 2 was to replicate the effect of an open path manipulation on school motivation with a small sample of exclusively low-asset students and to further explore the proposed psychological process. An exploratory field experiment tested whether the effect of an open path manipulation on low-asset students’ motivation was mediated by a salient education-dependent identity. In a study of identity and achievement, Destin and Oyserman (2010) showed that middle school students who spontaneously described their futures in a way that included a college education subsequently spent more time on schoolwork and earned higher grades than did students who generally envision an education-independent identity, including future careers that are more independent of education and may not require college (e.g., entertainers, actors, athletes). Open path information may function through the same identity-based mechanism, but studies have not directly explored whether an open path increases the likelihood that students from low-asset households envision an education-dependent identity for themselves.

The financial realities of students from low-asset families may normatively impair school motivation by reducing the likelihood that they imagine college as part of their future. In other words, the future careers that readily come to mind for students in low-asset contexts are less likely to be careers that are believed to require a college education. However, according to the proposed model, an open path manipulation, which draws attention to possible financial resources, should increase the likelihood that low-asset students imagine themselves in future career paths that include college. If college financial aid information actually influences the imagined future identities of certain students, the effects of an open path on school motivation should be mediated by an increased likelihood that students describe an education-dependent identity, including college as part of their imagined future. The current study explores whether providing information that promotes an open path to college for low-asset adolescents enhances school motivation because it leads students to meaningfully imagine an education-dependent identity (i.e., future careers that require college) when asked to consider their future.

Method

Thirty-six rising eighth-grade students were recruited at the beginning of a summer math enrichment program in a diverse Midwestern school district outside a major urban center (56% male, 44% female; 11.8% White, 55.9% Black, 17.6% Latino/a, 2.9% Asian/Asian American, 11.8% mixed or other). The program exclusively targeted and admitted average achievement level students from financially strained, low-asset neighborhoods and households. During the first week of the program, after providing parental consent, teachers distributed packets to participants, which randomly assigned them, between-subjects, to receive one of two college information sheets. Exactly 50% of the participants were randomly assigned to the open path condition, and the second page of their packet (after the introduction page) included information describing sources of need-based financial aid, similar to Study 1. The other 50% of the participants were randomly assigned to the closed path condition, and the second page of their packet included information about the costs of major colleges and universities in the area (see Appendix B). According to existing research, low-asset adolescents are readily aware of the costs of college and have been found to respond to closed path information in the same way that they respond to a no-information control (Destin & Oyserman, 2009; Grodsky & Jones, 2007; Horn et al., 2003). The closed path condition also provided a comparison group that discounted the possibility that the open path treatment effect in Study 1 simply reflected an increased salience of college. Both conditions contained a short writing activity where students answered questions about the treatment information to ensure that they processed the material (see appendices).

Next, to assess salient future identity, students completed an open-ended prompt, asking “What do you plan to be doing in 10 years?” (see Destin & Oyserman, 2010). All responses were independently rated by two research assistants blind to condition and hypotheses. Responses that did not mention and were unlikely to require a college education were coded “education-independent identities” (e.g., I would like to be a waiter for a restaurant; 66.7%), and responses that did mention or were likely to require college were coded “education-dependent identities” (e.g., I want to go to college for a business degree; 33.3%). Most students provided only one response to the prompt (52.8%), and of those who provided two or more responses, the majority wrote about multiple responses that were all education-dependent or all education-independent. The 10% of the participants who provided both education-dependent and education-independent responses were coded as having an education-dependent identity because they did indicate at least one salient idea about the future that included a college education. The raters agreed on 95% of the responses, and the two disputed responses were resolved by the author (see Table 2). This open-ended measure of future identity allows students to indicate their most salient future identity without explicit reference to college or jobs within the prompt. Closed-ended and guided measures of future identity (e.g., expected educational attainment or occupational expectations; see Mello, Anton-Stang, Monaghan, Roberts, & Worrell, 2012) have a much stronger demand effect and are designed to capture a student’s more stable sense of his or her likely future. The open-ended prompt, on the other hand, is less transparent and captures a student’s salient future identity, which could be influenced from moment to moment by information in their context making it much more likely to predict immediate school-related outcomes (Destin & Oyserman, 2010).

Table 2.

Percentage of Participant Responses Coded Into Each Category by Experimental Condition in Study 2.

Coding categories	Sample responses	Closed path %	Open path %
	I would like to be a waiter for a restaurant
Education-independent identity		88.9	44.4
	I’m a cheerleader now, so I hope to continue, but I also love doing hair
	I want to go to college for a business degree
Education-dependent identity		11.1	55.6
	I want to become an engineer and a part-time artist.

Next, students completed a measure of school motivation that differed from Study 1 in order to expand the range of motivational outcomes that may be influenced by perceiving an open path to college and to focus on an indicator relevant to the particular sample in a program aiming to enhance students’ goals in math. Students reported the grade that they aimed to earn in their math class during the upcoming school year (1 = F, 13 = A+; $\bar{X}$ = 9.92, B/B+; SD = 1.42). Finally, they completed demographic items, and all participants in both conditions were immediately debriefed with information regarding resources available to help pay for college. Participants also provided permission to release their percentile rank on their most recently completed standardized achievement test in math ( $\bar{X}$ = 61.94, SD = 20.70) as a measure of prior achievement to be included in analyses as a covariate.

Results and Discussion

A preliminary ANOVA tested for existing differences in achievement between participants who were randomly assigned to the open path or closed path condition. A difference was found, and students who were randomly assigned to the open path condition actually scored lower on previous achievement tests ( $\bar{X}$ = 55.11, SD = 25.18) than students in the closed path condition ( $\bar{X}$ = 68.78, SD = 12.23), F(1, 34) = 4.29, p = .046, d = .69, 95% CI = [.26, 27.08]. So, the test of the open path treatment on motivation was a more conservative test than anticipated because students in the closed path condition were higher achievers; however, the following analyses statistically controlled for prior achievement to eliminate bias.

An ANCOVA tested the effect of experimental condition on student grade goals. As expected, participants who were randomly assigned to the open path condition set significantly higher grade goals ( $\bar{X}$ = 10.06, SD = 1.55) than did students who were randomly assigned to the closed path condition ( $\bar{X}$ = 9.78, SD = 1.31), F(1, 33) = 4.62, p = .039, d = .19, 95% CI = [−1.71, −.05]. Using an exclusively low-asset sample, Study 2 conceptually replicated the pattern observed among low-asset students in the socioeconomically diverse Study 1 sample.

Next, to explore the effect of the open path treatment on the proposed mediator, a binary logistic regression analysis tested the effect of experimental condition on future identity, again controlling for prior achievement. As predicted, participants who were randomly assigned to the open path condition were then more likely to report an education-dependent identity than participants who were randomly assigned to the closed path condition, b = 2.41, standard error (SE) = .93, p = .010, odds ratio (OR) = 11.171, 95% CI = [1.80, 69.41]. In fact, students who were randomly assigned to consider an open path were over 5 times more likely to imagine a future identity that included college than students who were randomly assigned to consider a closed path (55.6% vs. 11.1%).

In order to explore whether a salient education-dependent identity was responsible for the effect of the open path treatment on grade goals, the indirect effect was tested using a model generated by the INDIRECT macro for SPSS Version 23 (Preacher & Hayes, 2008). As shown in Figure 3, the proposed indirect effect model was supported. Path “a” was significant, and random assignment to the open path condition led to a significantly higher chance of generating an education-dependent identity (b = .47, p = .005). Path “b” was also significant, and an education-dependent identity predicted higher grade goals (b = .95, p = .038). At the same time, Path “c’”, the direct effect of condition on school motivation was not significant (b = .43, p = .331). Most important, the test of the indirect effect of the open path manipulation on grade goals through education-dependent identity using 5,000 bootstrap samples was significant because the 95% CI did not include zero, 95% CI = [0.01, 1.37]. A test of the alternative indirect effect model, which modeled the opposite direction between education-dependent identity and motivation (effect of open path treatment on education-dependent identity, mediated by increased school motivation), was not significant, 95% CI = [−.22, 6.51].

Figure 3.

Identity-Mediated Effect of Open Path Treatment Condition on Grade Goals in Study 2.

In an exploratory experiment, Study 2 conceptually replicated and extended the effect of perceiving an open path to college. In an exclusively low-asset sample of adolescents, those who were randomly assigned to receive information about need-based financial aid were more likely to subsequently imagine education-dependent future identities than were participants who instead received information about the costs of college. As suggested by identity-based motivation theory, the increased salience of an education-dependent identity mediated effects on the grade goals that students set for themselves, controlling for their prior achievement. This initial exploration suggests that financial aid information actually encourages young students from low-asset family backgrounds to visualize a future that is likely to include education and in turn set higher current school goals.

General Discussion

Two field experiments provided evidence of a psychological pathway toward school motivation for early adolescents that is dependent upon the perception that financial resources are available to reach long-term educational goals, such as college. When young students from families with few financial assets are provided with a possible strategy to confront high college costs, they become more likely to imagine higher education as a part of their future, and they set higher grade goals and plan to exert more effort in school. As expected, open path information matters most for students from households that seem to have fewer financial assets, who know that paying for college will be difficult. Despite the chronic home and neighborhood cues that may make their own education-dependent identity feel less salient and out-of-reach, a simple exercise that draws attention to possible resources inspires motivation to set higher goals and engage in schoolwork. Low-asset students who are not guided to consider possible resources for college are less motivated to set higher goals and engage in schoolwork because they are more likely to generate salient images of their future identity that are detached from higher education.

The current approach builds upon theory that advances the understanding of how socioeconomic contexts and financial resources influence student outcomes. Psychological approaches to achievement motivation rarely address students’ perceptions of financial resources directly, instead focusing primarily on the development of values and goal orientations with the secondary acknowledgment of the role of contextual factors, in general. Identity-based motivation theory provides valuable contrast to these approaches by fundamentally describing identities as dynamically constructed within contexts, and even contexts themselves are construed differently based upon available cues. By building from a model that starts with construal of context, the current research directly manipulates and assesses how young people’s perceptions of specific financial considerations lead to school motivation. Continued research may further advance theory by investigating how a similar approach intersects with more stable, individual-level factors related to achievement motivation and cultural factors such as race and ethnicity. In addition to implications for theory, an increasingly complex psychological understanding of socioeconomic disparities in achievement has significant practical implications.

As students from certain backgrounds continue to face persistent structural challenges that impede academic success and social mobility (Aikens & Barbarin, 2008; Burkhauser, Feng, Jenkins, & Larrimore, 2009; Piketty & Saez, 2003), it is increasingly important to understand the subtle psychological factors that can drive students to persist in or disengage from school. Early adolescence provides a window where students are particularly likely to imagine a fluid range of possible futures (Klimstra, Hale, Raaijmakers, Branje, & Meeus, 2009; Kroger, Martinussen, & Marcia, 2010; Meeus et al., 2010), and carefully distributed information that is relevant to their future identities can carry lasting motivational implications (Yeager & Walton, 2011). The economically diverse school settings sampled in Study 1 show that information intended to open opportunity pathways matters most for students whose everyday lives are more likely to reinforce the idea that resources to pay for college will be limited. For students from backgrounds where resources feel more abundant, pathways to future goals feel open on a regular basis. More generally, information about an open path to a future identity that aligns with messages that are transmitted on a regular basis by everyday contextual cues does not induce a change in students’ behavioral plans. Instead, when information about available resources is received by students who normally inhabit a home where assets are limited, it seizes attention and inspires the corresponding motivation to work toward reaching a desired future identity.

Apparent homeownership provided the most relevant and accessible measure of wealth and assets, which are necessary to finance a college education (Sherraden, 1991). Analyses of nationally representative data have consistently illustrated a relationship between indicators of wealth, such as homeownership, and children’s educational outcomes that is independent of the effects of parental income and education (e.g., Boehm & Schlottmann, 1999; Haurin, Parcel, & Haurin, 2002; Williams Shanks, 2007; Williams Shanks et al., 2010; Yeung & Conley, 2008; Zhan & Sherraden, 2003). Assets such as homeownership have also been uniquely linked to the development of an education-dependent identity (Elliott, Choi, et al., 2011), which was found to mediate effects of an open path on school motivation among the low-asset sample in Study 2. In a large, longitudinal study of household wealth, increases in home equity strongly predicted children’s actual college enrollment, showing that homeownership does indeed help to make college aspiration a financial reality (Lovenheim, 2011).

Given the normative role of household wealth in how students perceive college possibilities, it is appropriate that a measure of apparent homeownership moderated the effect of open path information on school motivation in a socioeconomically diverse school setting. Students who were less likely to perceive that their family owned their home and had significant financial assets benefitted most from a cue that drew attention to other possible resources. A parallel test with available data on family household income did not show the same pattern, providing more evidence that wealth and assets are especially relevant to the development of an education-dependent identity and school motivation.

Limitations and Future Directions

Although the current studies did test and provide evidence supporting the proposed psychological process of focus among early adolescents, the scope of the studies was limited in certain ways and future research is necessary to build upon the developments. First, school motivation was assessed in the form of students’ plans to spend time at home engaged in schoolwork and the grades that they aimed to earn in upcoming classes. These types of outcomes were appropriate to test whether an open path brings to mind an education-dependent identity that can be drawn upon to shape low-asset students’ everyday choices and goals (Oyserman & Destin, 2010), and whether school-related behaviors and achievement become more important to young students when they can imagine futures that are connected to education. Although a demand effect may lead some participants to exaggerate their school plans and goals, these measures remain predictive of actual school behaviors and achievement, especially when students know that their responses will not be evaluated by their teachers (e.g., Oyserman, Bybee, Terry, & Hart-Johnson, 2004; Zimmerman et al., 1992). Furthermore, by randomly assigning participants to the open path or control condition, students who are more prone to demand characteristics should be randomly distributed across conditions. Even so, larger scale intervention studies remain necessary to assess the effects of an open path over time or to pair such information with other effective strategies for longer term learning (e.g., implementation intentions Gollwitzer, 1999; learning goals Grant & Dweck, 2003; self-directed learning Gureckis & Markant, 2012) to assess effects on diverse indicators of school motivation and achievement over time. While future and related work may focus on the longer term applications (e.g., Bettinger et al., 2012), the current studies were foundational and aimed to expand theory by providing controlled, shorter scale illustrations of an important socioeconomic boundary condition of the effect and the mediating psychological pathway. Additionally, for community partners, it was important that all participating students, including those in control conditions, were debriefed quickly and provided with the open path information that was found to be beneficial. As such, the studies were designed to assess immediate psychological and motivational outcomes rather than attempting to maintain some students’ fidelity to a control condition over longer periods of time.

Taken together, two studies provide a consistent pattern of findings despite relatively modest sample sizes. When conducting classroom-based experiments, many unique constraints emerge that do not influence laboratory-based research or even traditional educational survey research. The need for administrative and parental support often limits the scope and number of eligible participants for each individual study. As such, smaller experiments in field settings can provide insight and expand theory beyond that which can be ascertained from more convenient and traditional methods in psychology or education. Students from low-asset family backgrounds can be particularly difficult to recruit, and de-facto socioeconomic segregation limits the number of low-asset students who attend mixed-income schools such as those sampled in Study 1. However, the current studies were able to include sufficient samples to illustrate effects consistent with hypotheses to show the socioeconomic boundary condition of a reliable open path effect.

Study 2 provided promising findings regarding the malleability and importance of a salient education-dependent identity. Future research might draw from more qualitative approaches (e.g., Behnke, Piercy, & Diversi, 2004) to further disentangle distinct aspects of students’ future identities, such as their expected careers and their expected education, in order to expand upon these encouraging initial effects. Future studies might also expand on the current measures of assets with a range of other indicators that reflect students’ perceptions of their families’ levels of wealth and financial resources. Finally, additional research might draw from these findings to investigate how differing policies toward college admission and funding at the state, regional, and national levels could influence the academic motivation and achievement of early adolescent students from backgrounds with limited financial assets.

Broadly, the field experiments provide evidence that efforts to improve school outcomes can consist of small amounts of targeted information that connect meaningfully to students’ future identities and address their actual current financial circumstances. The open path effect differs for students from different backgrounds within the same classrooms, suggesting that efforts to reduce inequality require more than a reduction of socioeconomic segregation in education. Instead, interventions and public policy that take into account the family and household factors influencing how students imagine their own futures are most likely to be successful.

Footnotes

Appendix A

Appendix B

Acknowledgements

I would like to acknowledge Daphna Oyserman, David Rapp, and Jennifer Cromley for their contributions to the development of this research.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Mesmin Destin is an assistant professor at Northwestern University in the School of Education & Social Policy and the Department of Psychology. He is also a fellow of the Institute for Policy Research. He received his PhD in social psychology from the University of Michigan.

References

Aikens

N. L.

Barbarin

(2008). Socioeconomic differences in reading trajectories: The contribution of family, neighborhood, and school contexts. Journal of Educational Psychology, 100, 235-251.

Bandura

(2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology, 52, 1-26.

Bargh

J. A.

Gollwitzer

P. M.

Lee-Chai

Barndollar

Trötschel

(2001). The automated will: Nonconscious activation and pursuit of behavioral goals. Journal of Personality and Social Psychology, 81, 1014-1027.

Baum

Payea

(2010). Education pays: The benefits of higher education for individuals and society. Washington, DC: The College Board.

Behnke

A. O.

Piercy

K. W.

Diversi

(2004). Educational and occupational aspirations of Latino youth and their parents. Hispanic Journal of Behavioral Sciences, 26, 16-35.

Bettinger

E. P.

Long

B. T.

Oreopoulos

Sanbonmatsu

(2012). The role of application assistance and information in college decisions: Results from the H&R Block FAFSA experiment. The Quarterly Journal of Economics, 127, 1205-1242.

Boehm

T. P.

Schlottmann

A. M.

(1999). Does home ownership by parents have an economic impact on their children, 8, 217-232.

Burkhauser

R. V.

Feng

Jenkins

S. P.

Larrimore

(2012). Recent trends in top income shares in the United States: reconciling estimates from March CPS and IRS tax return data. Review of Economics and Statistics, 94(2), 371–388.

Cook

T. D.

(2002). Randomized experiments in educational policy research: A critical examination of the reasons the educational evaluation community has offered for not doing them. Educational Evaluation and Policy Analysis, 24, 175-199.

10.

Cooper

Robinson

J. C.

Patall

E. A.

(2006). Does homework improve academic achievement? A synthesis of research, 1987-2003. Review of Educational Research, 76, 1-62.

11.

Destin

Oyserman

(2009). From assets to school outcomes: How finances shape children’s perceived possibilities and intentions. Psychological Science, 20, 414-418.

12.

Destin

Oyserman

(2010). Incentivizing education: Seeing schoolwork as an investment, not a chore. Journal of Experimental Social Psychology, 46, 846-849.

13.

Destin

Richman

Varner

Mandara

(2012). “Feeling” hierarchy: The pathway from subjective social status to achievement. Journal of Adolescence, 35, 1571-1579.

14.

Dunkel

C. S.

(2000). Possible selves as a mechanism for identity exploration. Journal of Adolescence, 23, 519-529.

15.

Eccles

J. S.

Wigfield

(2002). Motivational beliefs, values, and goals. Annual Review of Psychology, 53, 109-132.

16.

Elliott

Choi

E. H.

Destin

Kim

K. H.

(2011). The age old question, which comes first? A simultaneous test of children’s savings and children’s college-bound identity. Children and Youth Services Review, 33, 1101-1111.

17.

Elliott

Destin

Friedline

(2011). Taking stock of ten years of research on the relationship between assets and children’s educational outcomes: Implications for theory, policy and intervention. Children and Youth Services Review, 33, 2312-2328.

18.

Gollwitzer

P. M.

(1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54, 493-503.

19.

Grant

Dweck

C. S.

(2003). Clarifying achievement goals and their impact. Journal of Personality and Social Psychology, 85, 541-553.

20.

Grodsky

Jones

M. T.

(2007). Real and imagined barriers to college entry: Perceptions of cost. Social Science Research, 36, 745-766.

21.

Gureckis

T. M.

Markant

D. B.

(2012). Self-directed learning: A cognitive and computational perspective. Perspectives on Psychological Science, 7, 464-481.

22.

Haurin

D. R.

Parcel

T. L.

Haurin

R. J.

(2002). Does homeownership affect child outcomes? Real Estate Economics, 30, 635-666.

23.

Hedges

L. V.

Hedberg

E. C.

(2007). Intraclass correlation values for planning group-randomized trials in education. Educational Evaluation and Policy Analysis, 29, 60-87.

24.

Higgins

E. T.

(2012). Motivation science in social psychology: A tale of two histories. In Kruglanski

A. W.

Stroebe

(Eds.), Handbook of the history of social psychology (pp. 199-218). New York, NY: Psychology Press.

25.

Horn

L. J.

Chen

Chapman

(2003). Getting ready to pay for college: What students and their parents know about the cost of college tuition and what they are doing to find out. Washington, DC: National Center for Education Statistics.

26.

Jargowsky

P. A.

(1996). Take the money and run: Economic segregation in U.S. metropolitan areas. American Sociological Review, 61, 984-998.

27.

Klimstra

T. A.

Hale

W. W.

III Raaijmakers

Q. A. W.

Branje

S. J. T.

Meeus

W. H. J.

(2009). Identity formation in adolescence: Change or stability? Journal of Youth and Adolescence, 39, 150-162.

28.

Kroger

Martinussen

Marcia

J. E.

(2010). Identity status change during adolescence and young adulthood: A meta-analysis. Journal of Adolescence, 33, 683-698.

29.

Lovenheim

M. F.

(2011). The effect of liquid housing wealth on college enrollment. Journal of Labor Economics, 29, 741-771.

30.

McWhirter

E. H.

Valdez

Caban

A. R.

(2013). Latina adolescents’ plans, barriers, and supports: A focus group study. Journal of Latina/o Psychology, 1, 35-52.

31.

Meeus

Van

Schoot

Keijsers

Schwartz

S. J.

Branje

(2010). On the progression and stability of adolescent identity formation: A five-wave longitudinal study in early-to-middle and middle-to-late adolescence. Child Development, 81, 1565-1581.

32.

Mello

Z. R.

Anton-Stang

H. M.

Monaghan

P. L.

Roberts

K. J.

Worrell

F. C.

(2012). A longitudinal investigation of African American and Hispanic adolescents’ educational and occupational expectations and corresponding attainment in adulthood. Journal of Education for Students Placed at Risk, 17, 266-285.

33.

Nordstokke

D. W.

Zumbo

B. D.

(2010). A new nonparametric Levene test for equal variances. Psicológica, 31, 401-430.

34.

Oreopoulos

Dunn

(2013). Information and college access: Evidence from a randomized field experiment. Scandinavian Journal of Economics, 115, 3-26.

35.

Oyserman

Bybee

Terry

(2006). Possible selves and academic outcomes: How and when possible selves impel action. Journal of Personality and Social Psychology, 91, 188-204.

36.

Oyserman

Bybee

Terry

Hart-Johnson

(2004). Possible selves as roadmaps. Journal of Research in Personality, 38, 130-149.

37.

Oyserman

Destin

(2010). Identity-based motivation: Implications for intervention. The Counseling Psychologist, 38, 1001-1043.

38.

Oyserman

Destin

Novin

(2015). The context-sensitive future self: Possible selves motivate in context, not otherwise. Self and Identity, 14, 173-188.

39.

Perreira

K. M.

Fuligni

Potochnick

(2010). Fitting in: The roles of social acceptance and discrimination in shaping the academic motivations of Latino youth in the U.S. southeast. Journal of Social Issues, 66, 131-153.

40.

Piketty

Saez

(2003). Income inequality in the United States, 1913-1998. The Quarterly Journal of Economics, 118, 1-39.

41.

Preacher

K. J.

Hayes

A. F.

(2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879-891.

42.

Quillian

(2012). Segregation and poverty concentration: The role of three segregations. American Sociological Review, 77, 354-379.

43.

Raudenbush

S. W.

Martinez

Spybrook

(2007). Strategies for improving precision in group-randomized experiments. Educational Evaluation and Policy Analysis, 29, 5-29.

44.

Reardon

S. F.

(2011). The widening academic achievement gap between the rich and the poor: New evidence and possible explanations. In Duncan

G. J.

Murnane

R. J.

(Eds.), Whither opportunity? Rising inequality, schools, and children’s life chances (pp. 91-116). New York, NY: Russell Sage Foundation.

45.

Ryan

R. M.

Deci

E. L.

(2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55, 68-78.

46.

Sherraden

(1991). Assets and the poor: A new American welfare policy. Armonk, NY: M.E. Sharpe.

47.

Smith

E. R.

Semin

G. R.

(2007). Situated social cognition. Current Directions in Psychological Science, 16, 132-135.

48.

Stephens

N. M.

Hamedani

Destin

(2014). Navigating the social class divide: A diversity education intervention improves first-generation students’ academic performance and all students’ college transition. Psychological Science, 25, 943-953.

49.

Vohs

K. D.

Mead

N. L.

Goode

M. R.

(2006). The psychological consequences of money. Science, 314, 1154-1156.

50.

Williams Shanks

T. R

. (2007). The impacts of household wealth on child development. Journal of Poverty, 11, 93-116.

51.

Williams Shanks

T. R.

Destin

. (2009). Parental expectations and educational outcomes for young African American adults: Do household assets matter? Race and Social Problems, 1, 27-35.

52.

Williams Shanks

T. R.

Kim

Loke

Destin

. (2010). Assets and child well-being in developed countries. Children and Youth Services Review, 32, 1488-1496.

53.

Yeager

D. S.

Walton

G. M.

(2011). Social-psychological interventions in education: They’re not magic. Review of Educational Research, 81, 267-301.

54.

Yeung

W. J.

Conley

(2008). Black-white achievement gap and family wealth. Child Development, 79, 303-324.

55.

Zhan

Sherraden

(2003). Assets, expectations, and children’s educational achievement in female-headed households. Social Service Review, 77, 191-211.

56.

Zimmerman

B. J.

Bandura

Martinez-Pons

(1992). Self-motivation for academic attainment: The role of self-efficacy beliefs and personal goal setting. American Educational Research Journal, 29, 663-676.

57.

Zimmerman

B. J.

Kitsantas

(2005). Homework practices and academic achievement: The mediating role of self-efficacy and perceived responsibility beliefs. Contemporary Educational Psychology, 30, 397-417.