High School Students' Tenacity and Flexibility in Goal Pursuit Linked to Life Satisfaction and Achievement on Competencies Tests

Acknowledgement: This article uses data from the National Educational Panel Study (NEPS): Starting Cohort 4–9th grade (School and Vocational Training), https://doi.org/10.5157/NEPS:SC4:10.0.0. The NEPS data collection is part of the Framework Program for the Promotion of Empirical Educational Research, funded by the German Federal Ministry of Education and Research and supported by the Federal States. We have no conflict of interest to disclose.

The purpose of education is not just academic achievement but also the development of life skills that allow students to flourish both socially and economically (Ciarrochi et al., 2016; Heckman, 2006). Young people face many challenges, including those related to education, relationships, and career pathways (Ciarrochi et al., 2017; Dietrich et al., 2012; Erikson, 1968). The school years are crucial for them in forming and refining their life goals. Healthy adolescent development includes the formation, maintenance, and adjustment of goals (Bühler et al., 2019). Tenacious goal pursuit and flexibility in the face of barriers to goals are important for a healthy self-regulation (Brandtstädter, 2009; Brandtstädter & Renner, 1990). Yet, there is a dearth of research on these skills in school students. Understanding the role of tenacity and flexibility regarding goals among students can help guide interventions aimed at promoting student achievement and flourishing.

Young people often feel a lot of self-imposed pressure to perform well academically and fit in socially, but also pressure from important others such as parents, teachers, and peers (Hayes & Ciarrochi, 2015). They may deal with such pressures by striving for goals, which can help them alter their context or adapt to it (Little, 2000; McAdams & Olson, 2010). But youth may also experience barriers to realizing their goals; for instance, lack of time, resources, support, or skill, which can compromise their well-being (Basarkod, 2019). Healthy adolescence development, therefore, requires students to not only determine and strive for goals, but also navigate through the difficulties they face while pursuing their goals (Massey et al., 2008).

The dynamics of goal striving often involve a discrepancy between actual and desired outcomes for the goals (Brandtstädter & Renner, 1990). According to the dual-process framework (Brandtstädter & Rothermund, 2002), there are two coping tendencies for reducing this discrepancy: (a) assimilative coping, which involves transforming one’s circumstances in accordance with personal preferences to have a better chance of achieving the goal; and/or (b) accommodative coping, which involves adjusting personal preferences in the face of negative consequences of goal pursuit, which can involve letting go of the goal. Assimilative coping implies that when encountering barriers to goals, young people can try to redouble their efforts and tenaciously alter themselves or the environment (TEN). Alternatively, in accommodative coping, youth could flexibly adjust to the negative consequences of certain goals by altering their standards (FLEX). The present study examines the links of these two forms of coping tendencies in young people to their satisfaction with life and achievement on competencies tests.

Prior research suggests that the tendency to relentlessly pursue goals (TEN) as well as to accept goal failure when it is encountered (FLEX) may be beneficial for both life satisfaction and achievement (Bailly et al., 2012, 2016; Brandtstädter & Renner, 1990). The key question of our research is: are these two tendencies complementary (i.e., do students benefit from using both TEN and FLEX together) or oppositional (i.e., is it beneficial to use one or the other but not both simultaneously)? Some researchers suggest that TEN and FLEX are complementary, and people with high levels of both are able to adapt most effectively to changing circumstances (Kelly et al., 2013). However, others have suggested that the two present a regulatory dilemma between holding on to a goal or letting it go (Bailly et al., 2016).

Although there has been a substantial amount of research on the dual process framework (Brandtstädter & Rothermund, 2002; Schaeper, 2020; Zhang, 2020); hardly any is focused on young people. The majority has focused on older people (e.g., Brandtstädter & Renner, 1990; Bailly et al., 2016; Kelly et al., 2013). The reason behind this focus may be that accommodative processes are particularly relevant to people who have aging related reduced capacity (Siltanen et al., 2019). However, there is no reason to assume that accommodative coping would be any less important for young people. Youth, too, encounter barriers that are either difficult or impossible to overcome. Their choices and potential pathways are often prescribed by people and forces outside of their control. For example, in Germany, the context of the current study, children are assigned to university or vocational school tracks at an early age. The focus on older adults in past research on TEN and FLEX limits our ability to make conclusions about the dual processes model in young people, whose goals are often different from older adults (Ogilvie et al., 2001). Further, as past research has not focused on youth, achievement on tests of academic competencies as an outcome of TEN and FLEX has not been explored. Still, past research can provide hints about how these processes may operate in youth and affect their well-being and test performance.

Before we delve into past empirical research on TEN and FLEX, it would be helpful to clarify the meaning of TEN and FLEX to distinguish these constructs from other apparently similar sounding popular constructs in the broader literature. TEN seems to have some conceptual overlap with the widely studied construct of grit or perseverance in long-term goal pursuits (Duckworth et al., 2007); which has been associated with youth academic achievement (Clark & Malecki, 2019; Muenks et al., 2017). And FLEX sounds similar to the popular construct of psychological flexibility (Williams et al., 2012). However, we must be careful of the “jingle-jangle fallacy” that seems to plague educational psychology: constructs that sound the same do not necessarily mean the same thing (Gonzalez et al., 2020; Marsh et al., 2019).

Grit comprises two facets, perseverance of effort and consistency of interests, which have been examined with respect to academic achievement (e.g., Guo et al., 2019). TEN does not focus on consistency of interests. Although there appears to be a conceptual overlap of TEN with persistent effort, this is not the same as the perseverance facet of grit, which is highly similar to conscientiousness (e.g., “I am diligent”). Neither conscientiousness nor the perseverance facet of grit focus on difficulties in goal pursuits, which is a crucial theoretical aspect of TEN. Along the same lines, conscientiousness involves good time management skills and dependability, none of which are measured by the TEN or FLEX items. Instead, TEN subsumes clear representations of actions as under one’s control and implementation of assimilative intentions using a concrete plan of action (Brandtstädter et al., 1999), which do not feature in the theory of grit or the personality trait of conscientiousness.

As mentioned previously, TEN involves persistent effort in the service of goals. However, unlike other related constructs, TEN has an explicit focus on a relentless pursuit of goals (e.g., “the more difficult the goal, the more I think it is worth doing”), which is not necessarily adaptive like grit, conscientiousness, persistence or resilience (Brandtstädter et al., 1999). It is this relentless aspect of TEN that seems to be the key for the regulatory dilemma hypothesis. Specifically, TEN may undo the adaptive effects of FLEX during performance on standardized tests of academic competences, even though TEN is expected to be helpful for school grades in general that require long-term cumulative efforts as well as flexibility in adapting to day-to-day academic challenges. In a specific situation of a one-off test of competencies, having high levels of both TEN and FLEX are expected to work against each other. For instance, flexible standards or adjusting one’s perspective regarding tractable versus intractable test questions (i.e., high levels of FLEX) would help students better manage their emotional state during the test, whereas rigidly continuing to work on the difficult problem (i.e., high levels of TEN) may cost them valuable time that would be better spent on other more tractable questions. A conscientious pupil would show better time management skills whereas a student high in TEN is expected to pay a price for the relentless aspect of TEN.

FLEX may also be likened to psychological flexibility, which focuses on people’s adjustment to life in general, particularly through their willingness to endure distress for what they find meaningful (Arewasikporn et al., 2019). Distress endurance is a key component of psychological flexibility. In contrast, FLEX focuses on adjustments to negative consequences of goal pursuit, which sometimes may involve abandoning specific goals but ultimately has to do with adjustments of personal standards. Psychological flexibility emphasizes acceptance of distress regarding the status quo and clarification of personal values (Williams et al., 2012), whereas the accommodative processes of FLEX involve devaluation of blocked goals in a reappraisal of the status quo and change in goals to orient oneself toward new developmental options (Brandtstädter et al., 1999).

From a theoretical perspective, TEN and FLEX should not be seen as enduring personality traits such as conscientiousness. Rather, they focus on how individuals respond to goals at certain phases of their lives, depending on the constellation of resources at their disposal, including their physical and mental capabilities (Brandtstädter et al., 1999). The early theorizing of TEN and FLEX (e.g., Rothermund & Brandtstädter, 1997) emphasized that these constructs need to be conceptualized in a life span developmental perspective such that TEN and FLEX could be orthogonal at some stages of life (e.g., in youth) but complementary at other stages (e.g., in old age). Therefore, TEN and FLEX are importantly different from general personality traits that are relatively stable across the life span.

A major focus of past research on TEN and FLEX is on coping with blocked goals due to aging and/or disability (e.g., multiple sclerosis; Van Damme et al., 2019). With age, goal pursuit is expected to exceed resources, and switching from tenacity to flexibility mode helps to regain an overall sense of efficacy, despite functional declines and losses (Bailly et al., 2012). Past research on older adults suggests that TEN and FLEX are both linked to well-being: they have been shown to be associated with lower levels of stress and depression, fewer internalizing and externalizing symptoms, and greater levels of subjective well-being, life satisfaction, and self-esteem (Brandtstädter & Greve, 1994; Brandtstädter et al., 1993, Wrosch et al., 2011, 2003).

Even though research on these coping strategies have predominantly been done with older adults, adolescents too face blocked goals and insoluble problems at times. FLEX could help young people cope with situations they have little control over and buffer the possible detrimental impacts of barriers on well-being. In terms of research with youth, there have been only a few studies exploring FLEX (none were found using TEN). A few studies have focused on the predictors of FLEX or accommodative coping (e.g., cognitive abilities, experiences in life, goal disengagement, and parental influences; Thomsen & Greve, 2013; Greve & Thomsen, 2013); with even fewer exploring the outcomes associated with this coping strategy (Greve & Enzmann, 2003).

Greve and Enzmann (2003) explored the impact of FLEX (accommodative coping tendency) on the self-esteem of German incarcerated male youths (aged between 14 and 24) serving their first prison sentence. Their primary finding was that self-esteem increased during imprisonment and that this increase depended on the level of accommodative coping; the increase in self-esteem was earlier and faster for highly accommodative individuals. They also found that accommodative coping was negatively correlated with depression (r = −.21), and positively correlated with wellbeing (r = .19) and self-efficacy (r = .36). These correlations of flexible goal pursuit with wellbeing related outcomes are similar to the ones reported in other studies with incarcerated youths (Greve et al., 2001) and older adults (e.g., Brandtstädter & Greve, 1994; Brandtstädter et al., 1993).

Much of the past research on TEN and FLEX has assumed that old age demands individuals to accommodate to their diminishing functioning. Can we say the reverse is true of adolescents, that youth is a time to be tenacious and inflexible, given that young people are arguably at the peak of their functioning while also having had insufficient time to find the “limitations” of life? People in later adulthood face a number of goal obstacles that young people are relatively less likely to encounter (or at least be aware of): loss of future opportunities, increased physical pain, and diminished mental and physical functioning (Bozkurt & Yılmaz, 2016; Heckhausen et al., 2001; Kelly et al., 2013). Most youth in schools may not be aware of such limits and may view their life as full of possibilities.

Such a difference in outlook could translate into differences in the goals people have across the life span. For instance, Ogilvie et al. (2001) examined goal contents of adolescents, middle-age individuals, and older adults, and found clear age differences. Adolescents were more focused than middle-aged and older adults on goals related to acquiring or obtaining positive outcomes (looking good, making money, getting good grades). In contrast, older adults had relatively higher levels of goals that focused on keeping what they already had and avoiding the loss of currently existing positions (e.g., reading more to stay up-to-date). Thus, in general, youth appear to focus more on goals that build capacity and resources while older adults focus on goals that hold on to capacities and resources.

The only study (that we are aware of) examining the impact of FLEX on wellbeing in school-aged students within an academic setting was conducted by Thomsen et al. (2015). They found that FLEX positively predicted well-being 1 year later in a sample of 12–15-years-old German students. In a second sample of 9–12-years-old students, they found a moderation effect of stress and FLEX in predicting well-being. That is, for students who were high on FLEX, the negative impact of stress on well-being was reduced. In this case, stress was operationalized as change in academic achievement. While Thomsen et al. did not explore the impact of FLEX on change in achievement (as they were both used as predictor variables), they reported a statistically nonsignificant correlation between accommodative coping at the two time points with the change in achievement scores.

As mentioned previously, there are two competing hypotheses of how tenacity and flexibility might interact. The complementary hypothesis suggests that tenacious goal pursuit (TEN) while also being able to let go of hopeless goals to flexibly adjust one’s standards (FLEX) is ideal. Tenacious people who are low in flexibility may experience problems such as goal conflict and persisting with a behavior even when it ceases to be effective. Both of these possibilities have been linked to lower levels of well-being (Emmons & King, 1988; Serpell et al., 2009). The complementary hypothesis, therefore, suggests that the interaction between TEN and FLEX would have a positive association with positive outcomes; simultaneous high levels of both tendencies would be associated with high levels of positive outcomes such as students’ life satisfaction and achievement on competencies tests.

In contrast, the regulatory dilemma hypothesis suggests that people face a conflict between holding on to a goal and letting it go (Bailly et al., 2016; Bak & Brandtstädter, 1998). Thus, while individuals may find it helpful to either persist with their goals tenaciously or to let it go, having both tendencies simultaneously may not be helpful or practical. Thus, the regulatory dilemma hypothesis suggests that the interaction between TEN and FLEX would have a negative association with positive outcomes; simultaneous high levels of both tendencies would be associated with low levels of positive outcomes of student life satisfaction and achievement on competencies tests.

Past studies have found support for both these hypotheses in samples of elderly individuals. In line with the complementary hypothesis, Bailly et al. (2016) found that older adults (Mage ∼ 78) who identified as having high levels of both TEN and FLEX were more likely to report having better health and life satisfaction than older adults who either had moderate levels of TEN with low levels of FLEX and moderate levels of FLEX with low levels of TEN. Similarly, Kelly et al. (2013) found that for older people (aged about 55 to 65), being high in both TEN and FLEX predicted the greatest reduction in self-reported depression, hostility, and physical ill-health. Arewasikporn et al. (2019) found that having both tendencies predicted less interference of pain in daily life among individuals who had a disability (Mage = 58). In contrast, two studies have found evidence for the regulatory dilemma hypothesis. Heyl et al. (2007) found that participants (aged 55–98) who scored high in both tendencies experienced higher negative affect. Similarly, Coffey et al. (2014) found that adults (Mage = 62; lower limb amputation) who tended to be high in TEN and FLEX experienced greater negative affect.

In sum, past research on the complementary versus regulatory dilemma has yielded mixed findings. The conditions under which the two hypotheses would hold true remain somewhat equivocal. Further, given the past focus has primarily been on older samples and self-reports, it is also unclear which of the two hypotheses is likely to be supported in samples of adolescents, especially in an academic setting.

We tested the two competing hypotheses of complementary versus regulatory dilemma with respect to tenacious and flexible goal pursuits in a sample of German high school students (N = 10,957) from the National Educational Panel Study (NEPS). The NEPS employed a 10-item measure of TEN and FLEX adapted from the popular measure of Brandtstädter and Renner (1990), which has primarily been used with older adults. We first examined the factor structure of the 10-item measure in our youth sample. The measure showed a poor fit to the data. To improve the measure, we employed a state-of-the-art machine learning based approach using genetic algorithms (GA), which avoids the pitfalls of relying on researchers’ subjective judgments about the items (Sahdra et al., 2016; Yarkoni, 2010). The GA-derived measure was then used in latent moderated structural equation models (LMS) for testing the complementary versus regulatory dilemma hypotheses regarding the links of TEN and FLEX in predicting the outcomes. Specifically, we tested the link of the latent interaction of TEN and FLEX with students’ self-reported life satisfaction and achievement in academic competencies tests.

We used a cross-sectional design with the data from the Starting Cohort 4 of the German National Educational Panel Study (NEPS; Blossfeld & von Maurice, 2019). The NEPS sampling design and weighting are discussed in detail elsewhere (Aßmann et al., 2011; Skopek et al., 2013; Steinhauer & Zinn, 2016). Briefly, a multilevel sampling design was used. Schools were sampled based on school lists, then classes from those schools were sampled and then students from the classes. A stratified sampling design was used for school sampling based on school type, region and other school characteristics. A professional institute collected data from schools in small groups of students. The NEPS had ethics clearance. The Starting Cohort 4 started in 2010/2011.

The subsample of the current study consisted of the students in Grade 10 who provided at least one valid response on the primary measures of interest, tenacious goal pursuit and flexible goal adjustment, which were administered in 2011/2012: N = 10,957; 52% female; 454 schools. The degree of missing data in the 10 items was low, ranging from .4 to 2% missing. The mean age of our sample in 2011 was M = 15.65 (SD = .68) years. The German school system has various school tracks. In our sample, 12% attended Hauptschule, 27% attended Realschule (both of these are vocational high school tracks), 47% attended Gymnasium (university high school track), and 14% attended other kinds of schools, including comprehensive.

Self-Reports

We used a 10-item self-report measure of the tendencies of tenacious goal pursuit (TEN) and flexible goal adjustment (FLEX), adapted from a popular measure of these construct first introduced by Brandtstädter and Renner (1990) and widely used in research since then (e.g., Bailly et al., 2016; Coffey et al., 2014; Heyl et al., 2007). In the NEPS subsample of our study, participants were provided with the following instruction: “The following statements refer to situations where your wishes, goals or plans might not be realized the way you would like to,” and asked to indicate the extent to which subsequent statements applied to them. Five statements measured TEN (e.g., “When there are difficulties in my path, I normally try harder,” “I tend to keep on fighting, even if the situation seems hopeless”). Another five items measured FLEX (e.g., “I often am still able to find meaning in major disappointments,” “If I do not get what I want, I see that as an opportunity to learn how to deal with things”). Participants responded to these items using a scale ranging from 1 (does not apply at all) to 5 (applies completely). The ωtotal reliability estimate for the TEN items was .70, and for the FLEX items was .68 (McNeish, 2018).

Satisfaction with life was measured using a six-item scale previously validated in school children (Tomyn & Cummins, 2011). Students were asked, “How satisfied are you . . . ” with regards to the following domains: “… with your life overall at the present,” “… with what you have? Think of money and things that you own,” “… with your health,” “… with your family life,” “… with your group of friends and acquaintances,” and “… with your situation at school?” Participants responded to the six items on a scale ranging from 0 (completely dissatisfied) to 10 (completely satisfied). The ωtotal reliability of the 6 items was .82.

Standardized Achievement Tests

The NPES Study included standardized achievement tests. For the 2011/2012 Grade 10 students in the current study, standardized achievement scores were available for German reading (M = .00, SD = .76), mathematics (M = −.04, SD = .84) and English (M = .02, SD = 1.60). For each of the three tests, we used the weighted likelihood estimate (WLE) from item response theory (IRT) models provided in the NEPS scientific use file (see details in the NPES technical report by Pohl & Carstensen, 2012).

Reasoning Ability

To control for students’ general intelligence in our structural equation models, we used the scores of the NEPS matrices test. The matrices test developed and validated specifically for use in the NEPS was based on Raven’s Standard Progressive Matrices (for further details, see Haberkorn & Pohl, 2013). Matrices tests are regarded as good proxies for reasoning or fluid intelligence (Nisbett et al., 2012); which is among the strongest predictors of academic performance (e.g., Deary et al., 2007). The NEPS matrices test included 12 items for which students had to identify a missing element from the response options that best completed a geometrical pattern. Each item was scored as correct or incorrect. The measure had a reliability of categorical ωtotal of .67 in our sample.

Machine Learning-Based Psychometrics

Preliminary analyses showed that a confirmatory factor analysis (CFA) of two correlated factors of TEN and FLEX items fit the data poorly (goodness-of-fit indices are reported in the Results section). Since the main goal of the study was to test the interaction of TEN and FLEX in predicting the outcomes, the first critical step was to ensure that the measurement model of the TEN and FLEX was sound. We therefore sought to identify the best fitting items of TEN and FLEX using a novel machine-learning approach based on genetic algorithms (GA), and examined their factor structure in a CFA.

The GA were first introduced by Holland (1975) as optimization tools for game theory and pattern recognition problems, but they have recently gained popularity in psychometrics for being highly convenient optimization tools for efficiently finding best fitting items in a measure without having to depend on researchers’ subjective judgments about the items (Sahdra et al., 2016; Yarkoni, 2010). The GA implement the principles of biological evolution (e.g., mutation, crossover, and selection based on fitness) in a computational framework to select the items for a new abbreviated measure, which tends to fit better and has less noise than the original measure (Sahdra et al., 2016; Yarkoni, 2010). The GA have been employed to create reliable and valid short forms of several psychological constructs, including personality traits (Yarkoni, 2010), psychopathy (Eisenbarth et al., 2015), experiential avoidance (Sahdra et al., 2016), body image (Basarkod et al., 2018), mindfulness in sports (Noetel et al., 2019), and multidimensional well-being (Marsh et al., 2020).

We implemented the GA method using the GAabbreviate package (Scrucca & Sahdra, 2015) in R Version 3.6.3 (R Core Team, 2020). The technical details of the GA procedure for scale abbreviation are described in Yarkoni (2010); and those of the GAabbreviate package can be found in Sahdra et al. (2016). Briefly, the GAabbreviate aims to minimize the “cost” of an item in the abbreviated scale based on the “fitness function” below, as described by the following equation in Sahdra et al. (2016):

Here, I is the item cost, k is the number of items to be retained, s is the number of subscales in the measure (if applicable), wi are the weights associated with each subscale (if applicable), and R2 is the variance that a linear combination of individual item scores can explain in the ith subscale.

Structural Equation Modeling (SEM)

All SEMs were run in Mplus Version 8.4 (Muthén & Muthén, 2019). To examine the factor structure of TEN and FLEX items, we ran several CFA models. Model 1 was a CFA to test the measurement model of our four variables of interest: flexibility, tenacity, life satisfaction, and achievement on competencies tests. Model 2 was based on Model 1 but included a structural model where life satisfaction and achievement were regressed on FLEX and TEN. Additionally, we controlled for sex, school track, and reasoning ability in Model 2. Our final Model 3 was based on Model 2 but included an interaction term of FLEX and TEN.

To test the interaction of TEN and FLEX in predicting students’ self-reported life satisfaction and achievement in NEPS competencies tests, we employed latent moderated structural equation models (LMS). We chose LMS because, unlike conventional approaches to testing interactions such as ordinary least squares regression, LMS yields estimates of interactions less affected by measurement error, thus increasing power and reducing bias in estimation (Little et al., 2006; Marsh et al., 2004; Trautwein et al., 2012). LMS models were estimated with the XWITH command in Mplus. We used cluster robust standard errors (MLR estimator), with school as a cluster variable (TYPE = COMPLEX in Mplus), and NEPS cross-sectional Wave 3 sampling weights in all SEMs.

In accordance with commonly accepted criteria, the Tucker-Lewis index (TLI) and comparative fit index (CFI) ≥ .90, and root mean square error of approximation (RMSEA) < .08 were considered to provide evidence of model fit (Chen, 2007). We used full information maximum likelihood (FIML) to handle missing data in all SEMs because FIML has smaller estimation bias than traditional approaches of listwise or pairwise deletion, even under violations of the assumptions of multivariate normality and missing-at-random (Enders, 2010; Graham, 2009).

Open Data and Analysis Code

Starting Cohort 4 data can be downloaded from the NEPS website (https://doi.org/10.5157/NEPS:SC4:10.0.0) after signing their data use agreement. We used R Version 3.6.3 (R Core Team, 2020) for processing the NEPS scientific use files for preparing the data for our study. In the interest of transparency and reproducibility (Nosek et al., 2015); the R code and Mplus input files for the analyses reported in this paper can be downloaded from Open Science Framework (https://osf.io/97yuv/?view_only=963667e17cc0414aa19d89c7e76c98f8).

As shown in Table 1, the fit indices of a CFA testing the factor structure of the 10-item TEN and FLEX measure was poor. We therefore employed the GA to improve the measurement model of the measure. As mentioned earlier, the GA emulates biological evolution in silico by starting with a “population” of items that are winnowed down through a process of “evolution” involving “successive generations” in which “mutations” (whereby a selected item is rejected) and “cross-over” (in which a selected item is swapped with a previously unselected item) are implemented with preset probabilities. We used the default settings of the R package GAabbreviate (Scrucca & Sahdra, 2015) for these parameters.

Consistent with the cross-validation recommendations for machine learning applications aimed to minimize overfitting (James et al., 2014), we employed cross-validation by training the GA on 75% of the sample (n = 7,877), and testing the variance-explained criterion on the remaining 25% of the sample (n = 2,626). Our goal was to find three “fittest” items each for TEN and FLEX subscales to allow at least three items per construct in subsequent SEM models. We therefore set the argument in GAabbreviate for the maximum number of items per subscale to be three. We used the default settings for item cost and population size in the GAabbreviate package but doubled the maximum iteration (from 100 to 200) to ensure a stable solution (see the R code in the link provided earlier).

After deriving a six-item measure using the GA on the training subsample, we tested the correlation of the new measure with the original measure using the test subsample. The three items of the TEN subscale of the GA-derived measure correlated with the original five items at .93, and the three items of the GA-derived FLEX subscale correlated with the original five items of the FLEX subscale at .91 in the validation subset. As shown in Table 1, a CFA of the new six-item measure in the validation subsample fit the data well. Table 1 also reports the fit indices of the same CFA in the full sample, which also shows that the model fit the data well, much better than the fit of the CFA of the 10-item measure. Tables S1 and S2 in the online supplementary materials show the standardized factor loadings and their standard errors of the CFAs of the 10-item and six-item measures, respectively, using the full sample. The wording of the six items can also be found in Table S2. Given the substantially better fit of the six-item measure compared to the 10-item measure, in all subsequent analyses, we used the six-item measure of TEN and FLEX.

Table 2 reports the means, SD and correlations of all the study variables. As shown in Table 2, the TEN and FLEX were positively linked with self-reported satisfaction with life. They also showed small but negative correlations with objective tests of competencies. Life satisfaction was largely unrelated with competencies scores, the only exception being a small positive correlation with math competence.

Table S3 in the online supplementary materials compares the correlations of the 10-item TEN and FLEX measure with the key study variables with the correlations of the GA-derived six-item measure with the same key study variables. The table also includes additional measures that were not part of our research question regarding the regulatory dilemma of TEN and FLEX with respect to performance on standardized tests of competencies; but these additional measures are nevertheless relevant for the construct validity of the TEN and FLEX measure used in our study. Performance on one-off standardized tests is importantly different from achievement in school grades, which requires consistent persistence and flexibility in navigating academic goals. Consistent with this reasoning, school grades and standardized test scores can have only a small positive correlation; particularly in Germany given the tracking system (e.g., see Marsh et al., 2005), which is what we found in our sample as well. Consequently, we would expect positive associations of TEN and FLEX with school grades and other measures of academic involvement. Indeed, as shown in Table S3, we observed positive associations of TEN with self-reported German and math grades in the previous semester. FLEX showed a small and positive association with math grades, but no association with German grades. Further, TEN and FLEX were positively associated with additional relevant variables: students’ attitude towards education, clarity about their future profession, familiarity with school qualification requirements, and the extent to which students think about school qualifications. These associations provide preliminary evidence for the construct validity of the TEN and FLEX measure of our study.

Importantly though, the GA-derived TEN and FLEX measure was associated with other variables similarly to the long-form TEN and FLEX. Specifically, the mean of absolute values of the differences in the zero-order correlations of the long-form and short-form TEN with other measures was .011. Similarly, the mean of absolute values of the differences in the correlations of the original FLEX measure and the GA-derived FLEX measure with other variables was .014. Both the mean values are very small. These results show that the GA-derived TEN and FLEX measure preserved the pattern of correlations of the original 10-item measure with other variables. That is, there was little loss of information in the process of abbreviating the TEN and FLEX measure using the GA method, but as shown earlier in the results of the CFA models, the GA improved the structural validity of the measure.

We employed LMS modeling to test the effects of an interaction of TEN and FLEX on satisfaction with life and achievement on competencies tests. As our first model, we tested a CFA of tenacity, flexibility, life satisfaction and test performance that fit the data well: χ2(84) = 628; RMSEA = .02; 90% CI [.023, 026]; CFI = .96; TLI: .95. We then ran Model 2 including the covariates of sex, German school track, and reasoning ability. Model 2 also fit the data well: χ2(117) = 1244; RMSEA = .03; 90% CI [.028, .031]; CFI = .94; TLI: .92. The R2 for life satisfaction was 49.3% and for test performance was 14.6%. This model served as a baseline for the final Model 3, which included the paths from the interaction of TEN and FLEX to the outcome variables, as shown in Figure 1 (Tables S4 to S6 in the online supplementary materials report the standardized estimates of the paths of Models 1 to 3, respectively).

Model 3 with the interaction paths fit the data significantly better than Model 2 without the interaction paths: D = 119.97, df = 1, p < .001. The ΔR2 for life satisfaction was 2.8%, but there was no change in R2 for achievement on competencies tests. Table 3 contains the standardized estimates of the regression paths of Model 3. To visually explore the interaction effect, we imported the Mplus factor scores from Model 3 into R. Figure 2 shows the region of significance plot (top) and an interaction plot (bottom) for life satisfaction.

The region of significance plot (Figure 2 top) for life satisfaction shows that the slope of flexibility is positive and statistically different from zero for most of the range of the observed data except at high levels of tenacity. The interaction plot (Figure 2 bottom) shows that the slope of flexibility at high (+2 SD) levels of tenacity is relatively flat (simple slope: −.01, SE = .03, p = .83); whereas the slope of flexibility at low (−2 SD) levels of tenacity is positive (simple slope: .51, SE = .02, p < .001). Flexibility does not seem to add much to satisfaction with life for highly tenacious students, but it seems to be linked to improved life satisfaction for those who are less tenacious.

For achievement on competencies tests, Figure 3 (top panel) shows that the slope of flexibility is negative and statistically different from zero for most of the range of the observed data except at very low levels of tenacity. The interaction plot in Figure 3 (bottom panel) shows that the slope of flexibility is negative but very small at low (−2 SD) levels of tenacity (simple slope: −.04, SE = .01, p = .01); but more negative at high (+2 SD) levels of tenacity (simple slope: −.17, SE = .01, p < .001). The effect size of −.04 is very small for explanation of single events but potentially consequential in the not-very-long run, and an effect size of −.17 is considered small at the level of single events but potentially more ultimately consequential (Funder & Ozer, 2019). High flexibility in adjusting goals seems to be costly for achievement on competencies tests for almost all students, but especially so for those who are highly tenacious in pursuing goals (see the online supplemental materials for exploratory person-centered analyses, which complement the variable-centered analyses reported here).

Overall, the results show that having high levels of both TEN and FLEX does not necessarily have any benefit for life satisfaction particularly for highly tenacious individuals and seems to be counterproductive for achievement on competencies tests. The findings are consistent with the theory that having high levels of both would represent a regulatory dilemma in which the two tendencies would work against each other.

We sought to test two competing hypotheses regarding the links of tenacious goal pursuit and flexible goal adjustment with high school students’ life satisfaction and achievement on academic competencies tests. The complementary hypothesis suggests that having the tendency of tenacious goal pursuit (TEN) in conjunction with the tendency to flexibly adjust one’s standard regarding goals (FLEX) would be most beneficial, whereas the regulatory dilemma hypothesis suggests that the two tendencies of TEN and FLEX pose a conflict because it is hard to let go of goals while flexibly changing one’s standards while also doggedly pursuing one’s goals (Bailly et al., 2016; Bak & Brandtstädter, 1998). We used an improved measure of TEN and FLEX in latent moderated structural equation models to test the association of the latent interaction of TEN and FLEX in predicting students’ self-reported satisfaction with life and achievement on the NEPS competencies tests. We controlled for gender, German school track, and reasoning ability while testing the interaction. Of the two competing hypotheses, our results supported the regulatory dilemma hypothesis, and were in line with some of the past research with older adults (Coffey et al., 2014; Heyl et al., 2007), which found that people who used both TEN and FLEX tended to experience greater negative affect. We found that having high levels of both tendencies was negatively associated with students’ life satisfaction and achievement on competencies tests.

When first setting up our latent models, the results revealed that the 10-item measure of TEN and FLEX used in the NEPS fit the data poorly. The 10 items were adapted from a longer 30-item measure developed by Brandtstädter and Renner (1990). The 30-item measure is widely used in studies on coping with goals, particularly in cases where goals may be blocked due to the aging process (e.g., Coffey et al., 2014; Heyl et al., 2007). But its measurement model is rarely explicitly tested before the investigators test their substantive hypotheses. A few researchers who did test the measurement model of the original 30-item measure identified issues with it: Heckhausen (1997) reported that the original German version of the TEN scale consisted of two subscales of positively and negatively worded items. Mueller and Kim (2004) tested an English version of the measure and found that the direction of the wording of the items of TEN and FLEX explained more variance in scores than did the two constructs themselves. Henselmans et al. (2011) tested the fit of the 30-item measure in three CFAs in a Dutch sample (1-factor model, 2-target-factors model, and 2-keying-factors model), and found that all three models fit the data poorly.

To the best of our knowledge, the fit of the 10 items of TEN and FLEX used in the NEPS has never been tested in either an adult sample or a youth sample before. In our sample of youth, a model of two correlated factors of TEN and FLEX using the 10 items did not fit the data. Therefore, it was necessary to improve the TEN and FLEX measure first, which we did using a state-of-the-art machine learning method using genetic algorithms (Sahdra et al., 2016; Scrucca & Sahdra, 2015; Yarkoni, 2010). That itself is an important contribution of the study to the literature. To the best of our knowledge, the six-item measure of TEN and FLEX we used in this study is the only well-fitting measure of these constructs that has been used with youth in the literature. Very few studies have used the FLEX measure with youth, and none, as far as we can tell, have used TEN with youth. Greve and colleagues (Greve & Enzmann, 2003; Greve et al., 2001) have used the 15-item measure of FLEX to examine its links with the self-esteem of incarcerated youth, but they used regression models with manifest variables. Thomsen and Greve (2013) used latent path modeling, but the focus of their study was to test a set of predictors of FLEX. They did not use TEN, so could not test the factor structure of the TEN and FLEX items in their sample of adolescents. Future studies with youth (and adults) could benefit from using the improved brief measure of TEN and FLEX from our study.

Past research testing the interaction of TEN and FLEX for well-being-related outcomes has largely focused on older adult samples and manifest variables (e.g., Arewasikporn et al., 2019; Coffey et al., 2014; Heyl et al., 2007). Our study is the first that we know of that tests the complementary and regulatory dilemma hypotheses in a school students’ sample, and uses SEM methods, which correct for measurement error. Both TEN and FLEX had statistically significant positive associations with life satisfaction, which is in line with prior research showing that these constructs are positively associated with subjective well-being-related outcomes in older samples (Brandtstädter & Greve, 1994; Brandtstädter et al., 1993, Wrosch et al., 2011, 2003). Importantly though, the main effects of TEN and FLEX on life satisfaction were moderated by an interaction of the two in our study; FLEX was positively associated with life satisfaction especially among those who scored low in TEN. This finding is interesting in the context of early theorizing about the relative importance of FLEX and TEN in younger versus older individuals. Brandtstädter and Renner (1990), when they originally developed the TEN and FLEX measure, suggested that people tend to shift from TEN to FLEX with age. Other research on TEN and FLEX with older adults (e.g., Heyl et al., 2007) suggests that FLEX may be particularly important for the elderly or disabled, who need to accommodate themselves in the face of reduced functioning. Importantly, our results suggest that FLEX is also important for youth and may be particularly helpful for students who find it hard to persist in their goals.

Preliminary analyses regarding the construct validity of TEN and FLEX showed that these constructs were positively associated with past semester’s grades in German and mathematics and other indicators of academic involvement such as students’ attitude toward education, clarity about their future profession, familiarity with school qualification requirements, and the extent to which students think about school qualifications. Therefore, TEN and FLEX seem to be generally beneficial for students. For our key outcome measure though, we focused on performance in competencies tests. TEN and FLEX showed small but negative associations with achievement on competencies tests, though the main effects were moderated by an interaction, which also had a small effect size.

One reason for the small interaction effect on achievement on competencies tests could be that the items of FLEX and TEN measure coping strategies regarding difficulties with goals in general, which for students might encompass much more than the difficulties related to academic achievement goals per say. Students face many challenges during school years, including those related to education as well as relationships with peers, family members, and teachers (Ciarrochi et al., 2017). The goal of education is not just academic achievement but preparing students for the challenges of life more broadly and enabling them to live a fulfilling life. As Heckman (2006) notes, the development of noncognitive skills is becoming increasingly important for the long-term economic success of young people. The 21st century desirable traits of the labor force include self-regulatory capacities, such as TEN and FLEX, which can help students thrive during and beyond their school years despite the regulatory dilemma regarding specific tests situations such as the one observed in the present study.

As was found with older adults’ samples (Coffey et al., 2014; Heyl et al., 2007); our results revealed strong evidence for the regulatory dilemma hypothesis. Thus, it does not seem to be beneficial for students’ life satisfaction, nor their achievement on competencies tests, to have both tendencies of TEN and FLEX, despite the fact that on their own, these tendencies showed positive effects, at least for life satisfaction. Having high levels of TEN and FLEX appear to be counterproductive for students. The reason for that may be that the two tendencies cancel each other out, at least in the short run. Past studies on TEN and FLEX in youth (e.g., Greve & Enzmann, 2003; Thomsen & Greve, 2001) have largely focused on FLEX and neglected to test TEN alongside. But accommodative tendency (FLEX), as important as it seems to be for students, is only one aspect of the dual process framework (Brandtstädter & Rothermund, 2002), which also emphasizes assimilative tendency (TEN). FLEX and TEN appear to be opposing tendencies, at least among youth, and measuring them both is important to capture the regulatory dilemma in students.

Promoting TEN and FLEX in school students may be important for their life satisfaction and long-term success. Research on goals-related interventions in educational settings has shown the benefits of improving the contents of students’ goals (e.g., Latham & Brown, 2006; Poppes et al., 2002), but also how students reflect on their goal pursuit; for example, by imagining negative and positive outcomes of the goals and considering how they might overcome barriers to the goals (Duckworth et al., 2013). To the best of our knowledge, there is no published research on interventions in education that directly target TEN and FLEX as conceptualized in the dual process theory (Brandtstädter & Rothermund, 2002). There is some evidence that learning about the struggles of great scientists can improve high school students’ motivation to learn science (Lin-Siegler et al., 2016), but this intervention seems to target persistence indirectly and seems to be specific to science education. The interaction effects in our data suggest the need for targeted interventions that build on students’ personal strengths, their predominant tendency in tenacity or flexibility in general. Interventionists need to be cautious about the regulatory dilemma inherent in combining the tendencies of tenacious goal pursuit and flexible adjustment of personal standards.

Teachers, parents, and mentors of students need to be aware that it is not necessarily beneficial to expect students to be tenacious and flexible about goal pursuit at the same time, particularly in the context of specific tests. Students who exhibit a tendency to do both may be responding to mixed messages from others about the best way to respond to the challenges of goal striving. Our results suggest that in an intervention targeting TEN and FLEX, it may be important to first identify students’ strength in TEN or FLEX. Those who find it easier to persist than to flexibly adjust one’s standards may be encouraged to focus on TEN, while those who find it harder to persist than to flexibly adapt may be encouraged to focus on FLEX. At least during school years, focusing on one coping tendency or the other to capitalize on students’ general tendencies of TEN and FLEX may serve students better than trying to promote both at the same time.

Our study has several strengths, such as the use of a large sample and sophisticated statistical methods. Furthermore, as mentioned earlier, our study is the first to use a well-fitting measure of TEN and FLEX in a high school student sample. Nevertheless, we acknowledge that a key limitation of our study is its cross-sectional design. Another limitation of the study is that some of the items of the TEN and FLEX are framed broadly and do not specifically refer to goals. However, when placed in the context of other items and the instructions of the measure focusing on goals, these items form a part of TEN and FLEX regarding goals. Also, at this stage, it remains unclear if the results of our study would generalize to non-German samples of youth. Still, the study fills a vacuum in the literature by showing that the regulatory dilemma regarding goal pursuit is just as important for youth as it has been shown in past research with older adults. We hope that the GA-derived brief measure of TEN and FLEX we used in our study will prove fruitful for future research testing the interactive effects of TEN and FLEX for student outcomes in diverse samples.

Future studies should also directly examine the correlations of TEN and FLEX with grit and conscientiousness, something we were not able to do in the NEPS subsample of our study. The perseverance facet of grit is highly similar to conscientiousness (Guo et al., 2019), both of which imply dependability and time management skills, which are not measured by the items of our TEN-FLEX measure. Conscientiousness or perseverance are generally adaptive traits, whereas TEN has a potentially maladaptive element of the relentless pursuit of goals at almost all cost. Also, conscientiousness is a relatively enduring personality trait, whereas TEN and FLEX are theorized in a life span developmental context where their levels are expected to change through age (Rothermund & Brandtstädter, 1997). Given that TEN-FLEX are theoretically different from conscientiousness/perseverance, we would expect TEN and FLEX to have at best weak positive associations with grit and conscientiousness.

Our findings suggest that having high levels of both TEN and FLEX appear to be counterproductive for students’ performance on academic competencies tests. As we mentioned earlier, one reason for that could be that TEN and FLEX operate in an orthogonal manner during the early years of life, particularly during adolescence (Rothermund & Brandtstädter, 1997). Another possibility is that teachers and/or parents could be encouraging both TEN and FLEX in students who might be struggling academically, hoping that one or the other might help such students. It is also possible that both of these processes might be occurring such that adults around students may be promoting both TEN and FLEX, but such efforts might be exacerbating the difficulties of some low achieving students. Further research is needed to disambiguate these processes.

Future research with students might also benefit from a different measure of TEN and FLEX that is specific to academic goals. The items of the TEN and FLEX in our study were worded generally, that is, they were not specific to any particular life goal. We observed weak negative associations of TEN and FLEX with achievement on the NEPS competencies tests. Importantly though, the main effects of TEN and FLEX on achievement were moderated by an interaction, which showed a small but negative association with achievement. High levels of FLEX were costly for achievement on competencies tests, especially for students with high levels of TEN. It is possible that a more targeted measure of TEN and FLEX specific to students’ regulatory strategies with respect to difficulties in achieving academic goals might show stronger associations with academic performance measures. Perhaps, the regulatory dilemma effect with respect to academic achievement might be stronger when assessed using a measure of TEN and FLEX focused on academic goals.

The fact that we examined both TEN and FLEX in our youth sample is an important strength of this study, and importantly adds to past research that focused only on FLEX in youth (e.g., Greve & Enzmann, 2003; Thomsen et al., 2015). But the fact that we only had a youth sample limits the conclusion we can draw about the interactive effects of TEN and FLEX in the larger life span. In its original conception, the dual process framework (Brandtstädter & Rothermund, 2002) was meant to describe self-regulation processes with respect to goal pursuit over the entire life span. Future research would benefit from cross-sectional samples with wide age ranges and longitudinal samples covering as much life span as possible, to examine for whom and under what conditions would the regulatory dilemma hypothesis or the complementary hypothesis may be true. Based on the results of our study with students, taken together with past research with older individuals, we speculate that TEN and FLEX may be orthogonal in high school years, perhaps even university education years, but may become complementary as students move beyond the school/university environment and have more experience with TEN and FLEX with age.

Academic achievement is an important goal of education, but it is not the sole purpose of education. The larger goal of education is to facilitate students’ development of desirable traits that help them to flourish socially and economically in the long run. Tenacity and flexibility regarding life goals are such capacities, which may be cultivated in a school context. In line with the theoretical model of dual process framework (Brandtstädter & Rothermund, 2002) and past research supporting the regulatory dilemma hypothesis in older adults (Coffey et al., 2014; Heyl et al., 2007), our results showed that having a high tendency of tenacity in conjunction with a high tendency of flexibility was negatively associated with achievement on competencies tests and life satisfaction of high school students. The results suggest that it is crucial to take both tenacity and flexibility into account while examining the role of coping tendencies regarding goal pursuit in students’ wellbeing and performance on standardized tests. More broadly, our findings suggest the need for targeted interventions that build on students’ personal strengths in tenacity or flexibility and try to minimize their regulatory dilemma of being tenacious and flexible simultaneously.

Arewasikporn, A., Roepke, A. M., Turner, A. P., Alschuler, K. N., Williams, R. M., Amtmann, D., & Molton, I. R. (2019). Goal pursuit, goal adjustment, and pain in middle-aged adults aging with physical disability. Journal of Aging and Health, 31(Suppl. 10), 214S–240S. 10.1177/0898264319827142

Aßmann, C., Steinhauer, H. W., Kiesl, H., Koch, S., Schönberger, B., Müller-Kuller, A., Rohwer, G., Rässler, S., & Blossfeld, H.-P. (2011). 4 sampling designs of the National Educational Panel Study: Challenges and solutions. Zeitschrift für Erziehungswissenschaft, 14(S2), 51–65. 10.1007/s11618-011-0181-8

Bailly, N., Hervé, C., Joulain, M., & Alaphilippe, D. (2012). Validation of the French version of Brandtstädter and Renner’s Tenacious Goal Pursuit (TGP) and Flexible Goal Adjustment (FGA) scales. European Review of Applied Psychology, 62(1), 29–35. 10.1016/j.erap.2011.09.005

Bailly, N., Martinent, G., Ferrand, C., Gana, K., Joulain, M., & Maintier, C. (2016). Tenacious goal pursuit and flexible goal adjustment in older people over 5 years: A latent profile transition analysis. Age and Ageing, 45(2), 287–292. 10.1093/ageing/afv203

Bak, P. M., & Brandtstädter, J. (1998). Flexible Zielanpassung und hartnäckige Zielverfolgung als Bewältigungsressourcen: Hinweise auf ein Regulationsdilemma [Flexible goal adjustment and tenacious goal pursuit as coping resources: Hints to a regulatory dilemma]. Zeitschrift fur Psychologie mit Zeitschrift fur Angewandte Psychologie, 206(3), 235–249.

Basarkod, G. (2019). The six ways to well-being (6W-WeB): A new measure of valued action that targets the frequency and motivation for six behavioural patterns that promote well-being [Doctoral thesis, Australian Catholic University]. ACU Research Bank. 10.26199/5dcc7d5fab0e8

Basarkod, G., Sahdra, B., & Ciarrochi, J. (2018). Body Image-Acceptance and Action Questionnaire-5: An abbreviation using genetic algorithms. Behavior Therapy, 49(3), 388–402. 10.1016/j.beth.2017.09.006

Blossfeld, H.-P., & von Maurice, J. (2019). Education as a lifelong process. In H.-P.Blossfeld & H.-G.Roßbach (Eds.), Education as a lifelong process: The German National Educational Panel Study (NEPS) (Vol. 3, pp. 17–33). Springer. 10.1007/978-3-658-23162-0_2

Bozkurt, U., & Yilmaz, M. (2016). The determination of functional independence and quality of life of older adults in a nursing home. International Journal of Caring Sciences, 9(1), 198–210.

Brandtstädter, J. (2009). Goal pursuit and goal adjustment: Self-regulation and intentional self-development in changing developmental contexts. Advances in Life Course Research, 14(1–2), 52–62. 10.1016/j.alcr.2009.03.002

Brandtstädter, J., & Greve, W. (1994). The aging self: Stabilizing and protective processes. Developmental Review, 14(1), 52–80. 10.1006/drev.1994.1003

Brandtstädter, J., & Renner, G. (1990). Tenacious goal pursuit and flexible goal adjustment: Explication and age-related analysis of assimilative and accommodative strategies of coping. Psychology and Aging, 5(1), 58–67. 10.1037/0882-7974.5.1.58

Brandtstädter, J., & Rothermund, K. (2002). The life-course dynamics of goal pursuit and goal adjustment: A two-process framework. Developmental Review, 22(1), 117–150. 10.1006/drev.2001.0539

Brandtstädter, J., Wentura, D., & Greve, W. (1993). Adaptive resources of the aging self: Outlines of an emergent perspective. International Journal of Behavioral Development, 16(2), 323–349. 10.1177/016502549301600212

Brandtstädter, J., Wentura, D., & Rothermund, K. (1999). Intentional self-development through adulthood and later life: Tenacious pursuit and flexible adjustment of goals. In J.Brandtstadter & R. M.Lerner (Eds.), Action and self-development: Theory and research through the life span (pp. 373–400). SAGE. 10.4135/9781452204802.n13

Bühler, J. L., Weidmann, R., Nikitin, J., & Grob, A. (2019). A closer look at life goals across adulthood: Applying a developmental perspective to content, dynamics, and outcomes of goal importance and goal attainability. European Journal of Personality, 33(3), 359–384. 10.1002/per.2194

Calinski, R. B., & Harabasz, J. (1974). A dendrite method for cluster analysis. Communications in Statistics, 3(1), 1–27.

Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling, 14(3), 464–504. 10.1080/10705510701301834

Ciarrochi, J., Atkins, P. W. B., Hayes, L. L., Sahdra, B. K., & Parker, P. (2016). Contextual positive psychology: Policy recommendations for implementing positive psychology into schools. Frontiers in Psychology, 7, 1561. 10.3389/fpsyg.2016.01561

Ciarrochi, J., Morin, A. J. S., Sahdra, B. K., Litalien, D., & Parker, P. D. (2017). A longitudinal person-centered perspective on youth social support: Relations with psychological wellbeing. Developmental Psychology, 53(6), 1154–1169. 10.1037/dev0000315

Clark, K. N., & Malecki, C. K. (2019). Academic Grit Scale: Psychometric properties and associations with achievement and life satisfaction. Journal of School Psychology, 72, 49–66. 10.1016/j.jsp.2018.12.001

Coffey, L., Gallagher, P., Desmond, D., & Ryall, N. (2014). Goal pursuit, goal adjustment, and affective well-being following lower limb amputation. British Journal of Health Psychology, 19(2), 409–424. 10.1111/bjhp.12051

Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational achievement. Intelligence, 35(1), 13–21. 10.1016/j.intell.2006.02.001

Dietrich, J., Parker, P., & Salmela-Aro, K. (2012). Phase-adequate engagement at the post-school transition. Developmental Psychology, 48(6), 1575–1593. 10.1037/a0030188

Duckworth, A. L., Kirby, T., Gollwitzer, A., & Oettingen, G. (2013). From fantasy to action: Mental contrasting with implementation intentions (MCII) improves academic performance in children. Social Psychological & Personality Science, 4(6), 745–753. 10.1177/1948550613476307

Duckworth, A., Peterson, C., Matthews, M. D., & Kelly, D. R. (2007). Grit: Perseverance and passion for long-term goals. Journal of Personality and Social Psychology, 92(6), 1087–1101. 10.1037/0022-3514.92.6.1087

Eisenbarth, H., Lilienfeld, S. O., & Yarkoni, T. (2015). Using a genetic algorithm to abbreviate the Psychopathic Personality Inventory-Revised (PPI-R). Psychological Assessment, 27(1), 194–202. 10.1037/pas0000032

Emmons, R. A., & King, L. A. (1988). Conflict among personal strivings: Immediate and long-term implications for psychological and physical well-being. Journal of Personality and Social Psychology, 54(6), 1040–1048. 10.1037/0022-3514.54.6.1040

Enders, C. K. (2010). Applied missing data analysis. Guilford Press.

Erikson, E. H. (1968). Identity: Youth and crisis. Norton.

Funder, D. C., & Ozer, D. J. (2019). Evaluating effect size in psychological research: Sense and nonsense. Advances in Methods and Practices in Psychological Science, 2(2), 156–113. 10.1177/2515245919847202

Gonzalez, O., MacKinnon, D. P., & Muniz, F. B. (2020). Extrinsic convergent validity evidence to prevent jingle and jangle fallacies. Multivariate Behavioral Research. Advance online publication. 10.1080/00273171.2019.1707061

Graham, J. W. (2009). Missing data analysis: Making it work in the real world. Annual Review of Psychology, 60, 549–576. 10.1146/annurev.psych.58.110405.085530

Greve, W., & Enzmann, D. (2003). Self-esteem maintenance among incarcerated young males: Stabilisation through accommodative processes. International Journal of Behavioral Development, 27(1), 12–20. 10.1080/01650250143000562

Greve, W., Enzmann, D., & Hosser, D. (2001). The stabilization of self-esteem among incarcerated adolescents: Accommodative and immunizing processes. International Journal of Offender Therapy and Comparative Criminology, 45(6), 749–768. 10.1177/0306624X01456008

Greve, W., & Thomsen, T. (2013). Developmental conditions of adaptive self-stabilization in adolescence: An exploratory study. International Journal of Developmental Science, 7(3-4), 119–131. 10.3233/DEV-120101

Guo, J., Tang, X., & Xu, K. M. (2019). Capturing the multiplicative effect of perseverance and passion: Measurement issues of combining two grit facets. Proceedings of the National Academy of Sciences of the United States of America, 116(10), 3938–3940. 10.1073/pnas.1820125116

Haberkorn, K., & Pohl, S. (2013). Cognitive basic skills in the scien- tific use file. University of Bamberg, National Educational Panel Study. https://www.neps-data.de/Portals/0/NEPS/Datenzentrum/Forschungsdaten/SC3/com_cogbasic2013_en.pdf

Hayes, L. L., & Ciarrochi, J. V. (2015). The thriving adolescent: Using acceptance and commitment therapy and positive psychology to help teens manage emotions, achieve goals, and build connection. New Harbinger Publications.

Heckhausen, J. (1997). Developmental regulation across adulthood: Primary and secondary control of age-related challenges. Developmental Psychology, 33(1), 176–187. 10.1037/0012-1649.33.1.176

Heckhausen, J., Wrosch, C., & Fleeson, W. (2001). Developmental regulation before and after a developmental deadline: The sample case of “biological clock” for childbearing. Psychology and Aging, 16(3), 400–413. 10.1037/0882-7974.16.3.400

Heckman, J. J. (2006). Skill formation and the economics of investing in disadvantaged children. Science, 312(5782), 1900–1902. 10.1126/science.1128898

Hennig, C., & Liao, T. (2013). How to find an appropriate clustering for mixed-type variables with application to socio-economic stratification. Journal of the Royal Statistical Society, Series C: Applied Statistics, 62(3), 309–369. 10.1111/j.1467-9876.2012.01066.x

Henselmans, I., Fleer, J., van Sonderen, E., Smink, A., Sanderman, R., & Ranchor, A. V. (2011). The tenacious goal pursuit and flexible goal adjustment scales: A validation study. Psychology and Aging, 26(1), 174–180. 10.1037/a0021536

Heyl, V., Wahl, H.-W., & Mollenkopf, H. (2007). Affective well-being in old age. European Psychologist, 12(2), 119–129. 10.1027/1016-9040.12.2.119

Holland, J. H. (1975). Adaptation in natural and artificial systems. The University of Michigan Press.

James, G., Witten, D., & Hastie, T. (2014). An introduction to statistical learning: With applications in R. Springer.

Kelly, R. E., Wood, A. M., & Mansell, W. (2013). Flexible and tenacious goal pursuit lead to improving well-being in an aging population: A ten-year cohort study. International Psychogeriatrics, 25(1), 16–24. 10.1017/S1041610212001391

Latham, G. P., & Brown, T. C. (2006). The effect of learning vs. outcome goals on self-efficacy, satisfaction and performance in an MBA program. Applied Psychology, 55(4), 606–623. 10.1111/j.1464-0597.2006.00246.x

Lin-Siegler, X., Ahn, J. N., Chen, J., Fang, F.-F. A., & Luna-Lucero, M. (2016). Even Einstein struggled: Effects of learning about great scientists’ struggles on high school students’ motivation to learn science. Journal of Educational Psychology, 108(3), 314–328. 10.1037/edu0000092

Little, B. R. (2000). Persons, contexts, and personal projects: assumptive themes of a methodological transactionalism. In S.Wapner, J.Demick, T.Yamamoto, & H.Minami (Eds.), Theoretical perspectives in environment-behavior research: Underlying assumptions, research problems, and methodologies (pp. 79–88). Springer U.S. 10.1007/978-1-4615-4701-3_8

Little, T. D., Bovaird, J. A., & Widaman, K. F. (2006). On the merits of orthogonalizing powered and product terms: Implications for modeling interactions among latent variables. Structural Equation Modeling, 13(4), 497–519. 10.1207/s15328007sem1304_1

Marsh, H. W., Huppert, F. A., Donald, J. N., Horwood, M. S., & Sahdra, B. K. (2020). The well-being profile (WB-Pro): Creating a theoretically based multidimensional measure of well-being to advance theory, research, policy, and practice. Psychological Assessment, 32(3), 294–313. 10.1037/pas0000787

Marsh, H. W., Pekrun, R., Parker, P. D., Murayama, K., Guo, J., Dicke, T., & Arens, A. K. (2019). The murky distinction between self-concept and self-efficacy: Beware of lurking jingle-jangle fallacies. Journal of Educational Psychology, 111(2), 331–353. 10.1037/edu0000281

Marsh, H. W., Trautwein, U., Lüdtke, O., Köller, O., & Baumert, J. (2005). Academic self-concept, interest, grades, and standardized test scores: Reciprocal effects models of causal ordering. Child Development, 76(2), 397–416. 10.1111/j.1467-8624.2005.00853.x

Marsh, H. W., Wen, Z., & Hau, K. (2004). Structural equation models of latent interactions: Evaluation of alternative estimation strategies and indicator construction. Psychological Methods, 9(3), 275–300. 10.1037/1082-989X.9.3.275

Massey, E., Gebhardt, W., & Garnefski, N. (2008). Adolescent goal content and pursuit: A review of the literature from the past 16 years. Developmental Review, 28(4), 421–460. 10.1016/j.dr.2008.03.002

McAdams, D. P., & Olson, B. D. (2010). Personality development: Continuity and change over the life course. Annual Review of Psychology, 61, 517–542. 10.1146/annurev.psych.093008.100507

McNeish, D. (2018). Thanks coefficient alpha, we’ll take it from here. Psychological Methods, 23(3), 412–433. 10.1037/met0000144

Mueller, D. J., & Kim, K. (2004). The Tenacious Goal Pursuit and Flexible Goal Adjustment scales: Examination of their validity. Educational and Psychological Measurement, 64(1), 120–142. 10.1177/0013164403258456

Muenks, K., Wigfield, A., Yang, J. S., & O'Neal, C. R. (2017). How true is grit? Assessing its relations to high school and college students’ personality characteristics, self-regulation, engagement, and achievement. Journal of Educational Psychology, 109(5), 599–620. 10.1037/edu0000153

Muthén, L. K., & Muthén, B. O. (2019). Mplus user’s guide (8th ed.).

Nisbett, R. E., Aronson, J., Blair, C., Dickens, W., Flynn, J., Halpern, D. F., & Turkheimer, E. (2012). Intelligence: New findings and theoretical developments. American Psychologist, 67(2), 130–159. 10.1037/a0026699

Noetel, M., Ciarrochi, J., Sahdra, B., & Lonsdale, C. (2019). Using genetic algorithms to abbreviate the Mindfulness Inventory for Sport: A substantive-methodological synthesis. Psychology of Sport and Exercise, 45, 101545. 10.1016/j.psychsport.2019.101545

Nosek, B. A., Alter, G., Banks, G. C., Borsboom, D., Bowman, S. D., Breckler, S. J., Buck, S., Chambers, C. D., Chin, G., Christensen, G., Contestabile, M., Dafoe, A., Eich, E., Freese, J., Glennerster, R., Goroff, D., Green, D. P., Hesse, B., Humphreys, M., . . .Yarkoni, T. (2015). Promoting an open research culture. Science, 348(6242), 1422–1425. 10.1126/science.aab2374

Ogilvie, D. M., Rose, K. M., & Heppen, J. B. (2001). A comparison of personal project motives in three age groups. Basic and Applied Social Psychology, 23(3), 207–215. 10.1207/S15324834BASP2303_7

Pohl, S., & Carstensen, C. H. (2012). NEPS technical report—Scaling the data of the Competence Tests(NEPS working paper No. 14). Otto-Friedrich-Universität, Nationales Bildungspanel.

Poppes, P., Vlaskamp, C., de Geeter, K. I., & Nakken, H. (2002). The importance of setting goals: The effect of instruction and training on the technical and intrinsic quality of goals. European Journal of Special Needs Education, 17(3), 241–250. 10.1080/08856250210162149

R Core Team. (2020). R: A language and environment for statistical computing (Version 3.6.3) [Computer software]. R Foundation for Statistical Computing.

Rothermund, K., & Brandtstädter, J. (1997). Entwicklung und Bewaltigung: Festhalten und Preisgeben von Zielen als Formen der Bewaltigung von Engwicklungsproblemen. In C.Tesch-Romer, C.Saleswski, & G.Schwarz (Eds.), Psychologie der Bewaltigung (pp. 120–133). Psychologie Verlags Union.

Sahdra, B. K., Ciarrochi, J., Parker, P., & Scrucca, L. (2016). Using genetic algorithms in a large nationally representative American sample to abbreviate the multidimensional experiential avoidance questionnaire. Frontiers in Psychology, 7, 189. 10.3389/fpsyg.2016.00189

Schaeper, H. (2020). The first year in higher education: The role of individual factors and the learning environment for academic integration. Higher Education, 79(1), 95–110. 10.1007/s10734-019-00398-0

Scrucca, L., & Sahdra, B. K. (2015). GAabbreviate: Abbreviating questionnaires (or other measures) using genetic algorithms (Version 1.0) [Computer software] .CRAN.

Serpell, L., Waller, G., Fearon, P., & Meyer, C. (2009). The roles of persistence and perseveration in psychopathology. Behavior Therapy, 40(3), 260–271. 10.1016/j.beth.2008.07.001

Siltanen, S., Rantanen, T., Portegijs, E., Tourunen, A., Poranen-Clark, T., Eronen, J., & Saajanaho, M. (2019). Association of tenacious goal pursuit and flexible goal adjustment with out-of-home mobility among community-dwelling older people. Aging Clinical and Experimental Research, 31(9), 1249–1256. 10.1007/s40520-018-1074-y

Skopek, J., Pink, S., & Bela, D. (2013). Starting Cohort 4: 9th Grade (SC4). SUF Version 1.1.0 data manual(NEPS research data paper). National Educational Panel Study (NEPS).

Steinhauer, H. W., & Zinn, S. (2016). NEPS technical report for weighting: Weighting the sample of Starting Cohort 4 of the National Educational Panel Study(Wave 1 to 6; NEPS Survey Paper No. 2). Leibniz Institute for Educational Trajectories.

Thomsen, T., & Greve, W. (2013). Accommodative coping in early adolescence: An investigation of possible developmental components. Journal of Adolescence, 36(5), 971–981. 10.1016/j.adolescence.2013.08.003

Thomsen, T., Fritz, V., Mößle, R., & Greve, W. (2015). The impact of accommodative coping on well-being in childhood and adolescence. International Journal of Behavioral Development, 39(5), 467–476. 10.1177/0165025414551762

Tomyn, A. J., & Cummins, R. A. (2011). The subjective wellbeing of high-school students: Validating the Personal Wellbeing Index—School Children. Social Indicators Research, 101(3), 405–418. 10.1007/s11205-010-9668-6

Trautwein, U., Marsh, H. W., Nagengast, B., Lüdtke, O., Nagy, G., & Jonkmann, K. (2012). Probing for the multiplicative term in modern expectancy–value theory: A latent interaction modeling study. Journal of Educational Psychology, 104(3), 763–777. 10.1037/a0027470

Van Damme, S., Kindt, S., Crombez, G., Goubert, L., & Debruyne, J. (2019). The relation between goal adjustment, goal disturbance, and mental well-being among persons with multiple sclerosis. Psychology & Health, 34(6), 645–660. 10.1080/08870446.2018.1556272

Williams, K. E., Ciarrochi, J., & Heaven, P. C. L. (2012). Inflexible parents, inflexible kids: A 6-year longitudinal study of parenting style and the development of psychological flexibility in adolescents. Journal of Youth and Adolescence, 41(8), 1053–1066. 10.1007/s10964-012-9744-0

Wrosch, C., Amir, E., & Miller, G. E. (2011). Goal adjustment capacities, coping, and subjective well-being: The sample case of caregiving for a family member with mental illness. Journal of Personality and Social Psychology, 100(5), 934–946. 10.1037/a0022873

Wrosch, C., Scheier, M. F., Miller, G. E., Schulz, R., & Carver, C. S. (2003). Adaptive self-regulation of unattainable goals: Goal disengagement, goal reengagement, and subjective well-being. Personality and Social Psychology Bulletin, 29(12), 1494–1508. 10.1177/0146167203256921

Yarkoni, T. (2010). The abbreviation of personality, or how to measure 200 personality scales with 200 items. Journal of Research in Personality, 44(2), 180–198. 10.1016/j.jrp.2010.01.002

Zhang, J. (2020). Tenacious goal pursuit, flexible goal adjustment, and life satisfaction among Chinese older adult couples. Research on Aging, 42(1), 13–22. 10.1177/0164027519876125

Submitted: June 26, 2020 Revised: December 4, 2020 Accepted: January 7, 2021