Identifying the nature of personal goals offers a window into a person’s life and helps us understand who they are. In McAdams’ (1995) terms, personal goals are part of what we know when we know a person, at the level of characteristic adaptations (McAdams & Pals, 2006). The role of personal goals in human functioning and growth has been described as “the linchpin of psychological organization” (Klinger, 1998, p. 44). Goals infuse individual lives with a purpose for living, as they reflect a person’s values, interests and priorities (Carver & Scheier, 2005). Whether studied as current concerns (Klinger, 1975, 1998), personal projects (Little, 1983), personally expressive activities (Waterman, 1993), personal strivings (Emmons, 1986), possible selves (Markus & Ruvolo, 1989), aspirations (Kasser & Ryan, 1993), or life tasks (Cantor & Sanderson, 1999), the literature is generally aligned in that goals contribute in varied ways to well-being (Carver & Scheier, 1998; Emmons & McCullough, 2003; Heckhausen et al., 2010, 2019; McAdams, 1995). We conceptualize personally expressive activities as the building blocks of pursuing personally meaningful goals, which are the foundation of eudaimonic well-being. Goal pursuit is consistently linked to well-being, such as a person’s subjective evaluation of how life is going (Brunstein, 1993; Diener et al., 2002; Hill et al., 2023a; Ryan & Deci, 2017) or enjoyment of the activities themselves (Waterman & Schwartz, 2013). In general, people prosper when they engage in valued activities that are inherently interesting and important to them (Waterman, 1993), and when activities are congruent with their personality strengths (Diener & Seligman, 2002), such as achievement striving (Hill et al., 2023a). Some researchers suggest the most effective route to increasing one’s long-term well-being is through the selection and pursuit of personally expressive activities (Emmons & McCullough, 2003; Waterman, 1993) which is by definition, shaped by the personality traits of a person. Indeed, people who strive for achievement tend to organize their behaviors in ways that help them pursue goals and experience well-being at a single point in time, and feeling competent is one reason why achievement strivers tend to enjoy a sense of well-being (Hill et al., 2023a). However, goal pursual is temporal in nature, requiring organization of one’s efforts and actions over time to achieve an outcome. Thus, the purpose of this paper is to test the processes through which achievement striving leads to well-being.

1 Theoretical Framework: Self-Determination Theory

Self-determination theory (SDT) is a meta-theory of personality and human motivation, containing six mini-theories that address facets of personality and human motivation. In particular, basic psychological need mini-theory proposes that every person has three innate psychological needs (autonomy, competence, and relatedness) which must be satisfied to experience optimal psychological health (Ryan & Deci, 2001). These needs are essential for psychological growth and wellness, are an inherent part of human functioning, and are universal (e.g., across culture, gender, and age; Ryan & Deci, 2001; Vansteenkiste et al., 2020). The satisfaction of these needs is necessary to enjoy positive psychological health, while thwarted needs may lead to negative outcomes such as psychological ill-being. SDT describes person-environment interactions which is the basis for theoretical predictions about people’s motivation, personality, and behaviors. In other words, people experience social environments or contexts which can either support or thwart the fulfilment of the three needs (Legault, 2017).

When people pursue activities that support the fulfilment of their BPNs, effective pursual (i.e., need satisfaction) is associated with positive life outcomes (Adams et al., 2017; Hill et al., 2023a). For example, when people feel competent (i.e., use of one’s skills and expertise while completing an activity), they may experience feelings of effectiveness and mastery. These social contexts are essentially need-relevant conditions that help facilitate improved well-being, such as activities that draw on one’s personality strengths. Those whose’ personality strengths involve having ambitious goals and the drive to pursue them are particularly inclined to experience well-being boosts from feelings of competence. Recently, Hill et al. (2023a) showed that personal projects can act as competence-supportive environments (which offer challenge and allow for skills and abilities to develop) which are important well-being boosters for those with achievement-oriented personality traits. Personally expressive activities may serve as a social environment that supports basic psychological need satisfaction and well-being. Moreover, with personality shaping the types of activities pursued and the likelihood of satisfying needs shaped by social contexts, there is a pathway from personality to activity pursual to psychological needs to well-being that is worth investigating.

2 Literature Review

2.1 Achievement Striving

Achievement striving is a personality trait that describes a disposition motivated to work hard and succeed (Drasgow et al., 2012). As a facet of conscientiousness, achievement striving entails setting high goals and having the drive to pursue and attain them (Dudley et al., 2006). Achievement-oriented individuals tend to spend their time in a way that promotes attainment of their ambitious goals (e.g., working hard to obtain a prestigious career). Making progress on personal goals elicits feelings of efficacy and achievement (Little et al., 1992), which is one reason why achievement strivers tend to experience high well-being. Hill et al. (2023a) proposed that satisfaction of the basic psychological need for competence may be so important to achievement strivers’ happiness that the relationship will hold across many dimensions or types of well-being. Indeed, meta-analytic research has highlighted a moderately high correlation between achievement striving’s broader factor conscientiousness and well-being (r =.36; Anglim et al., 2020). Taken together, the literature shows that being conscientiousness may be ‘good for’ one’s psychological health, fostering empirical investigation into the mechanistic pathways of these well-being benefits.

2.2 Well-Being

Psychological research on well-being is divided by two traditions: hedonia (usually studied in terms of satisfaction and enjoyment of one’s life) and eudaimonia (e.g., positive functioning, including personal growth, authenticity, and meaning in life; Ryan & Deci, 2001; Waterman, 1993). Hedonia and eudaimonia also influence how individuals generally orient their lives and motives for the activities they engage in (Huta & Waterman, 2014; Peterson et al., 2005). Assessing how and when eudaimonic and hedonic well-being differ is a primary theme in positive psychology research. Although conceptually distinct, self-report measures of hedonic and eudaimonic well-being tend to be positively and strongly correlated (Gallagher et al., 2009; Hill et al., 2023b; Joshanloo, 2016; Keyes et al., 2002) leading some to question the validity (or value) of differentiating these two forms of well-being (Kashdan et al., 2008). Nonetheless, there remains strong interest in understanding their similarities and differences (Huta & Waterman, 2014), not only in terms of experiences of well-being, but also with respect to how individuals live their lives (Ryan & Huta, 2009). Engaging in activities that are personally valued is an indicator of eudaimonic well-being (Scheier et al., 2006). More broadly, activities are a key component of many conceptualizations of eudaimonic well-being, such as worthwhile activities (White et al., 2017), personally expressive activities (Waterman, 1993), and eudaimonic motives for activities (Sheldon, 2016). Personal expressiveness refers to intense experiences of feeling more complete, fulfilled, or alive when participating in some activities compared to others. As described by Waterman (1993), the personality-activity fit is a feeling of a special fit between the activity and the person’s inherent abilities and is also reflected in an individual’s sense that actions within the activity reflect who one really is and what one was meant to do. For example, when achievement strivers engage in activities that draw on their personality strengths (i.e., encourage their expression of competence), they experience increased well-being (Hill et al., 2023a).

One mechanism that may explain these well-being benefits is the feeling of efficacy derived from pursuing activities which optimally utilize achievement strivers’ need for competence. In general, when people feel fully absorbed in moderately simple activities, they tend to report higher levels of enjoyment (Hill et al., 2024). This relationship is likely more nuanced for achievement strivers, who may thrive at relatively challenging activities because it optimally balances their skillset with the difficulty of completing the task at hand. When this balance is struck, people report feeling absorbed to the point of entering a flow state, another key indicator of eudaimonic well-being (Vittersø & Søholt, 2011).

2.3 The Flow State

Flow is a temporary state characterized by challenge-skills balance, merging of action and awareness, clear goals, unambiguous feedback, concentration, sense of control, loss of self-consciousness, time transformation, and feelings of intensity (Csikszentmihalyi, 1975 in Norsworthy et al., 2021). In a flow state, a person feels that work is effortless, they are in control, performance concerns disappear, and time seems to stop still (Norsworthy et al., 2021). Nakamura & Csikszentmihalyi (2002, 2014) argue that the ideal conditions for flow are challenge-skills balance, clear goals, and unambiguous feedback. Being in a flow state is an enjoyable experience (Diener & Seligman, 2002; Ghani & Deshpande, 1994; Moneta, 2004; Vittersø & Søholt, 2011) and reflecting on flow experiences can increase well-being. Norsworthy et al. (2021) recently argued that a challenge-skills balance is the precondition or antecedent for flow, while enjoyment is a fundamental characteristic of flow. One reason for considering well-being as an outcome of flow may be that self-focused attention (which is minimized during a flow state) is what creates negative affect; when self-focused attention is absent, self-defeating processes like rumination do not occur (Nakamura & Roberts, 2016). Further, flow may contribute to positive development (e.g., well-being, increased motivation) through a broaden-and-build process. For example, flow improves previous learning satisfaction and future performance (Wang & Hsu, 2013) and actively contributes to effective learning (Andersen, 2016). Nakamura & Csikszentmihalyi (2002) suggested that when flow is achieved, new intrinsic motives emerge that facilitate further engagement, a cycle fitting Fredrickson’s (1998) broaden-and-build theory. While flow is considered emotionless in the moment (as self-consciousness temporarily disappears), reflecting on the experience may bring feelings of happiness, particularly for achievement strivers. People high in achievement striving tend to be concerned with occupational goals (e.g., completing a university degree with good grades) and feel competent at activities associated with pursuing their goals (e.g., intensive study; Hill et al., 2023a), both of which promote their well-being. Achievement strivers are characteristically inclined to pursue personally valued activities that provide a sense of competence and feeling competent boosts well-being. More broadly, being conscientious involves emotional and motivational mechanisms that make an individual likely to engage in flow promoting activities (i.e., they may be more likely to intentionally spend time to master a challenging task; Kappe & van der Flier, 2010), which is why conscientiousness is the personality trait most strongly linked to flow proneness (Ullén et al., 2012).

Earlier work by Fredrickson (1998) showed that positive states such as flow increase well-being through broadening attentional, behavioral, and cognitive abilities and building intellectual and social resources. Although flow has traditionally been studied in people who hold elite skills and engage in performative settings (e.g., ballerinas, artists, rock climbers, and chess players; Csikszentmihalyi, 2014), flow is a state which can also be achieved in everyday activities (Baumann, 2012; Olčar et al., 2019). Flow is most likely to occur when individuals are engaged in activities where they feel challenged in their task but have the resources (e.g., skills) to adequately deal with the challenge. This challenge-skills balance is considered a primary dimension of flow (Csikszentmihalyi, 1975 in Norsworthy et al., 2021). Feeling you have the necessary skills in order to succeed at a task (i.e., feeling competent) may be a first, foundational step, in achieving a flow state, and experiencing a sense of competent should have stronger effects on well-being for individuals who strive for achievement (vs. those low in achievement striving) because feelings of being skilled to succeed at a task would match their dispositional valuation of achievement.

Overall, the literature on activities, competence, flow, and well-being, suggests that achievement-strivers will experience boosts in well-being when they pursue activities that give them a sense of competence and facilitate feelings of flow. This study aims to identify the order in which this process occurs, contributing to the scholarly debate on the role of flow and well-being mediators versus outcomes.

2.4 The Present Study

Much of the foundational knowledge of flow and well-being is drawn from samples characterized by elite skills and unique situations, such as athletes, musicians, and artists. Comparatively, much less is known about how flow can be experienced in everyday life, and particularly, through feeling competent at activities that are personally expressive. Thus, our objective is to assess the degree to which personally expressive activities boost achievement strivers’ well-being through feelings of competence and flow.

We assess this pathway both cross-sectionally and longitudinally to explore the temporal nature of achievement strivers’ experience of well-being through feelings of competence and flow during personally expressive activities. Our first hypothesis is as follows:

H1: Achievement striving will have a serial indirect effect on well-being through competence and flow.

This can be tested first through five hypothesized pathways in one cross-sectional serial mediation model using Time 1 data, as depicted in Fig. 1. Additionally, this hypothesis can be tested through a cross-lagged panel model incorporating both Time 1 and Time 2 data, as depicted in Fig. 2. Though longitudinal models have long been considered superior to cross-sectional tests (Cole & Maxwell, 2003), if the time lag (i.e., distance between the measurement occasion) does not match the true data generating process, results may be misleading. Thus, we present both cross-sectional and longitudinal models to compare and contrast results.

As personally expressive activities are specific to an individual’s personality strengths, understanding the types of activities that are most likely to improve well-being can inform positive psychology interventions, we also aim to answer two exploratory research questions:

Research Question 1: What types of personally expressive activities do individuals engage in?

Research Question 2: Which types provide the highest sense of competence and flow?

Fig. 1
figure 1

Proposed cross-sectional mediation model

Full size image
Fig. 2
figure 2

Proposed cross-lagged panel model. Note. Solid lines represent covariance between variables over time; covariance between variables within one wave not shown for simplicity. Bolded, colored, and dashed lines represent hypothesized paths for the indirect effect, according to the color key at bottom. T1 = Time 1 and T2 = Time 2

Full size image

3 Method

3.1 Sample Size Determination

Because the overall study was designed to answer multiple research questions in addition to the results presented in the present paper, the initial sample size was determined with a precision analysis. A precision analysis for bivariate correlations indicated 352 participants would be sufficient to produce a 95% confidence interval half-width of ±0.10 for correlations, assuming similar effect sizes to the average in social psychology research (r =.21; Richard et al., 2003). The present study came close to this target for Time 1 (N = 346), but attrition was high at Time 2 (N = 244; 29.5% attrition).

Because indirect effects are generally smaller than their constituent bivariate correlations, we also calculated a series of sensitivity statistical power analyses for a serial indirect effect in the cross-sectional model using Monte Carlo simulations (Schoemann et al., 2017), using 5000 replications, 20,000 Monte Carlo draws per rep (seed 1234), assuming a sample size of N = 346 and assuming equal correlations between all variables. Figure S1 shows we had sufficient (i.e., 80%) statistical power to detect indirect effects if correlations are around r =.25 or larger. Supplementary Figure S2-3 shows simulations for smaller samples, suggesting we have 80% statistical power to detect indirect effects if correlations are around 0.30 (N = 244) or 0.35 (N = 225) when considering Time 2 sample sizes after attrition and exclusions.

3.2 Participants

Participants were drawn from a larger study on personality and well-being in academic settings and were postsecondary students who were taking at least one statistics course at their university (i.e., Dalhousie or York University). There were recruitment inclusion and exclusion criteria for the purposes of the larger study studying anxiety in statistics students but are unrelated to the hypotheses tested in this paper. A participant flow diagram is depicted in Figure S4. The final sample size for this study was 346 participants at Time 1 and 244 participants at Time 2 (four months later). See Table 1 for a full summary of socio-demographic characteristics.

Table 1 Socio-demographic characteristics
Full size table

3.3 Measures

Copies of all materials and measures used in this study, including measures not examined in the present paper, can be found on our OSF page [https://osf.io/gn4tq/].

3.4 Achievement Striving

We used 10 items from the International Personality Item Pool to measure achievement striving (Jackson et al., 1996), from the broader factor conscientiousness of the Five Factor Model (McCrae & John, 1992). Participants rated their agreement on items such as “go straight for the goal.” This scale has previously shown good internal consistency (α = 0.89; Hill et al., 2023a), and moderate test-retest reliability (α = 0.78). In the current study, internal consistency is strong (α = 0.86).

3.5 Well-Being

Well-being was measured with two single-item measures: life satisfaction (i.e., hedonic or evaluative well-being) and life worth (i.e., eudaimonic well-being). The 10-point life satisfaction measure asks, “How satisfied are you with your life in general?” and provides two anchor labels (1 = very dissatisfied, 10 = very satisfied). The 10-point life worth measure asks, “To what extent you feel the things you do in your life are worthwhile?” and provides two anchor labels (1 = not at all, 10 = completely). Though the reliability and validity of single-item well-being measures have been challenged, research suggests they are psychometrically sound (Lucas & Donnellan, 2007; Moldovan, 2017), and they are feasible for inclusion in multi-purpose surveys. For example, single-item well-being measures perform similarly to multiple-item well-being scales and do not produce systematically different correlations compared to multiple-item well-being measures on theoretically relevant variables (Cheung & Lucas, 2014). The reliability of single-item measures has been deemed moderate to acceptable (Anusic & Schimmack, 2016; Krueger & Schkade, 2008; Lucas & Donnellan, 2007; Schimmack & Oishi, 2005). As these two well-being items are single-items, there is no internal consistency value to report.

3.6 Personally Expressive Activities

We used one item from the Personally Expressive Activities Questionnaire – Standard form (PEAQ-S) (Waterman, 1993, 2004) to gather open-ended information about personally expressive activities. Participants were asked to name one activity of personal importance that they “would use to describe themselves to another person.” The activities listed were then piped into the remaining questionnaires (competence and flow) with the items phrased as completions of a common stem: “When I engage in…”.

3.7 Competence

The Perceived Competence Scale was originally developed to measure people’s feelings of competence during medical school (Williams & Deci, 1996) and in patients managing their glucose levels during diabetes (Williams et al., 1998). The 4-item scale is scored on a 7-point scale (1 = not at all true, 7 = very true) and was measured in relation to the activity described by each participant using branching logic in the survey software; a sample item is: “I am capable to engage in this activity.” On average, Cronbach’s alpha has been good in past studies (a > 0.80; Williams et al., 1998; Williams & Deci, 1996). In the current study, internal consistency is strong (α = 0.87).

3.8 Flow

The experience of flow was measured using the Absorption subscale (4 items) of the Flow Short Scale (Rheinberg et al., 2003) and in relation to the previously described activity, which was piped in using branching logic in the survey software. The item completions for this scale were: (a) I feel just the right amount of challenge, (b) I do not notice time passing, (c) I am totally absorbed in what I am doing, (d) I am completely lost in thought. Each item was measured on a 7-point scale ranging from ‘‘not at all” to ‘‘very much.’’ Cronbach’s alpha for the overall scale was previously high (.90; Rheinberg et al., 2003). In the current study, internal consistency is moderately strong (α = 0.76).

4 Procedure

The research was approved by the Institutional Research Ethics Board at University Y (2022–6038) and University X (e2022-187). The survey was administered through SurveyMonkey, an online survey platform and took about 45 min to complete. Participants were collected through two separate methods (a) students involved in the undergraduate participation pool at two universities; (b) through flyers and online advertisements to students enrolled in statistics classes; (c) when permitted by course instructors, short presentations or videos presented to students inviting them to participate; and (d) email notifications for Time 2. They had to participate in the first wave of the study in the first month of classes.Footnote 1 No other inclusion/exclusion criteria were applied to maximize generalizability and feasibility. During the first survey, students were able to choose between cash or bonus points (or a combination), through the following compensation models: (1) A $25 Amazon gift card ($10 for completing the first survey and $15 for completing the second survey), or (2) bonus credit points for an eligible psychology class using the undergraduate participant pool system and an Amazon gift card (1 bonus point for the first survey and a $15 gift card for completing the second survey). Overall, recruitment began September 6, 2022 and finished on June 24, 2023). Because Time 1 surveys were administered at the beginning of an academic term, there were two cohorts (Fall 2022 and Winter 2023). More specifically, the Fall 2022 cohort (N = 313) filled out their first survey in the first five weeks of the Fall term (between September 6 and October 14, 2022) and their second survey in the first five weeks of the Winter term (between January 11 and February 16, 2023). The Winter 2023 cohort (N = 110) filled out their first survey in the first four weeks of the Winter term (between January 18 and February 20, 2023) and second survey in the four weeks following the Winter term (between May 25, 2023 and June 26, 2023). Because the processes under study likely generalize to Canadian university students more broadly, samples from both universities (Dalhousie University and York University) and both cohorts (Fall 2022 vs. Winter 2023) were merged into a single dataset.

5 Analytic Plan

5.1 Quantitative

The data, syntax, codebook, and questionnaires used in this study can be found on our OSF page (https://osf.io/gn4tq/]). Cross-sectional serial mediation models were tested using the lavaan package in R (Rosseel, 2012). Significance for indirect effects were calculated using bootstrapping using 5000 resamples. First, hypotheses were tested with two separate cross-sectional serial mediation models at Time 1 only (see Fig. 1 for a conceptual model). School and cohort were entered as covariates. The two models differed by dimension of well-being outcome (hedonic vs. eudaimonic), and all data are Time 1. Indirect effects using bootstrapping use unstandardized coefficients; the remaining coefficients are standardized to improve interpretation. See also supplementary Tables S1-S4 for coefficients not reported in-text.

The hypothesized indirect effect was also analyzed using a two-wave cross-lagged panel model (depicted in Fig. 2) to predict Time 2 well-being while controlling for Time 1 variables. The two-wave cross-lagged panel model allows for a longitudinal mediation model without requiring three waves of data (Little et al., 2007). The cross-lagged panel model requires the assumption that the variables themselves and the relationships between them remain stationary throughout the time. Additionally, the assumption of synchronicity requires that the data for each time point was truly collected at approximately the same time. This is reasonable, as all participants responded to the survey within a limited time range (about one month) for both waves of data collection. Model fit was assessed using multiple fit indices.

5.2 Missing Data

After exclusions, there were 346 participants at Time 1, but attrition was high at Time 2 (N = 244; 29.5% attrition). Nineteen participants are missing well-being outcome data at Time 2; thus the models use the 225 participants who have data for each variable in the specified model (35% missing). At the item level, missing data ranged from 0.58% (a competence item) to 3.18% (life worthwhileness). Scale totals were calculated by averaging all items; thus, if some items were missing for a given participant, their total score would be the average of items completed. Predictors of missingness were investigated; participants recruited from Dalhousie University had a greater proportion of missingness than York University (38.3% vs. 29.1%), and the Winter 2023 cohort had more missing data than the Fall 2022 cohort (45.1% vs. 30.2%). Thus, school and cohort were included as covariates in the serial mediation models and auxiliary variables in the cross-lagged panel model; no other variables in the model predicted missingness. For the cross-lagged panel models, we used auxiliary variables (package semTools; Jorgensen et al., 2022) and handled missing data using full information maximum likelihood approach for hypothesis testing and using listwise deletion for descriptive statistics. Auxiliary variables are variables that can help to make estimates on incomplete data, but are not part of the main analysis (Collins et al., 2001). Including auxiliary variables has the most impact when their correlation with missingness is high (greater than 0.4) and when the amount of missing values is large (greater than 25%; Collins et al., 2001; Graham, 2003). Missing data was relatively minor for the ANOVA analyses (0.5%) so listwise deletion was used.Footnote 2

5.3 Content Coding of Open-Ended Data

To answer the exploratory research questions, the personally expressive activities were first thematically categorized into six project types, based on: (a) common types of activities reported in positive psychology literature (Hill et al., 2023a) and (b) a pilot coding of project types on 100 participant’s listed activity. First, two coders (TH and JL) independently reviewed the listed activities and drafted a list of possible activity types. After a comparison and discussion of the potential activities categories, the coders both independently coded the first listed 100 activities into the categories. Then, the categories were discussed, and some refinements were made. After the final seven categories were agreed on, the two coders categorized the remaining activities. Discrepancies were resolved through discussion to arrive at a final categorization for each activity. For all activities, inter-rater reliability, measured as percentage of agreement before consensus, was 92.5%. Cohen’s kappa was calculated using the package irr (Gamer et al., 2019) which showed κ = 0.901, a near perfect level of inter-rater reliability (McHugh, 2012). To assess differences in levels of flow by activity type, we used a one-way ANOVA using a Welch F-test and the Games-Howell method for post-hoc tests which are appropriate for unequal variances (Field et al., 2012).

6 Results

Descriptive statistics on both demographics and on key study variables are displayed in Tables 1 and 2, respectively. Both the cross-sectional serial mediation models and longitudinal cross-lagged panel models test indirect effects leading from achievement striving to competence need satisfaction, through to flow, then to well-being. The serial mediation models use only Time 1 measures, and control for sample recruitment characteristics (i.e., school and cohort). The cross-lagged panel models includes measures at both time points, and includes school and cohort as auxiliary variables. Across both statistical techniques, Model 1 represents the hedonic well-being outcome of life satisfaction and Model 2 represents the eudaimonic well-being outcome of life worth. Some coefficients are omitted from the results for clarity (e.g., paths for covariates), but supplementary material includes all standardized serial mediation model coefficients (Tables S1-S2) and all standardized cross-lagged panel model coefficients (Tables S3-S4).

Table 2 Scale descriptives and internal consistency
Full size table

Correlations between key study variables are presented in Table 3. Achievement striving at Time 1 was positively correlated with other Time 1 variables except for flow; the relationship with age (r =.13) was weak, and the relationships with competence (r =.39) and well-being (average between well-being types: r =.33) were moderately high. Thus, compared to those who were lower in achievement striving, the achievement strivers tended to be slightly older, experience more competence when engaging in personally expressive activities, and feel happier. These relationships were generally similar for Time 2 variables, although eudaimonic well-being was more strongly related to other Time 2 variables than hedonic well-being and was related to flow. Competence at Time 1 was strongly, positively correlated to flow at Time 1 (r =.51), and weakly, positively correlated to both Time 1 well-being (average r =.22) and Time 2 life worth (but not life satisfaction; r =.15). Flow at Time 1 was weakly, positively correlated to well-being at Time 1 (average r =.14), but not at Time 2.

Table 3 Bivariate correlations between study variables
Full size table

6.1 Quantitative Analyses

6.1.1 Model 1: Life Satisfaction

The cross-sectional serial mediation model’s indirect effects are displayed in Fig. 3, with all coefficients in Tables S1. We did not find a significant serial indirect effect, wherein achievement striving indirectly predicted life satisfaction through competence and flow. Achievement striving had a moderate positive total effect on life satisfaction, β = 0.29, 95% CI [0.19, 0.39]; neither competence nor flow significantly predicted well-being. Achievement striving was moderately, positively related to competence, β = 0.39, 95% CI [0.26, 0.52], which in turn was strongly, positively related to flow, β = 0.57, 95% CI [0.46, 0.67]. The relationship between achievement striving and flow was weak and negative, but significant, β = − 0.14, 95% CI [-0.24, − 0.04].

Fig. 3
figure 3

Results of serial mediation model testing the indirect effect of achievement striving on life satisfaction (hedonic well-being) through competence and flow. Note. Bolded lines are significant, dashed are insignificant. Standardized path coefficients shown with 95% CI (bootstrapped 5000 resamples)

Full size image

We next tested across-lagged panel model (Fig. 4), which tested the same hypothesized paths but while controlling for Time 1 variables and using Time 2 mediators. Fit indices were as follows: χ2(6) = 11.02, p =.09; CFI = 0.99; TLI = 0.97; RMSEA = 0.05 (90% CI [0.00, 0.09]). All autoregressive paths with T1 variables predicting the same variable at T2 were large and statistically significant; however, none of the hypothesized cross-lagged paths were statistically significant.

Fig. 4
figure 4

Results of cross-lagged panel model testing the indirect effect of achievement striving on life satisfaction (hedonic well-being) through competence and flow. Note. Standardized coefficients with 95% CI. T1 = Time 1 and T2 = Time 2. Cohort and school included as auxiliary variables to predict missingness. Only selected coefficients shown, see Table S3 for all coefficients

Full size image

6.1.2 Model 2: Life Worth

The cross-sectional serial mediation model’s indirect effects are displayed in Fig. 5, with all coefficients in Table S2. We did not find a serial indirect effect, wherein achievement striving indirectly predicted life worth through competence and flow. Achievement striving had a moderate, positive total effect on life worth, β = 0.36, 95% CI [0.26, 0.46], and a moderate, positive effect on competence, β = 0.39, 95% CI [0.26, 0.52]. Additionally, achievement striving was weakly, negatively related to flow, β = − 0.14, 95% CI [-0.24, − 0.04], which in turn significantly, albeit weakly, predicted well-being (β = 0.12, 95% CI [0.01, 0.25).

Fig. 5
figure 5

Results of serial mediation model testing the indirect effect of achievement striving on life worth (eudaimonic well-being) through competence and flow. Note. Bolded lines are significant, dashed are insignificant. Standardized path coefficients shown with 95% CI (bootstrapped 5000 resamples)

Full size image

Next, we tested a cross-lagged panel model (Fig. 6), which tested the same hypothesized paths but while controlling for Time 1 variables and using Time 2 mediators. Fit indices were as follows: χ2(6) = 13.37, p = 04; CFI = 0.99; TLI = 0.95; RMSEA = 0.06 (90% CI [0.01, 0.10]). All autoregressive paths with T1 variables predicting the same variable at T2 were large and statistically significant; however, none of the hypothesized cross-lagged paths were statistically significant.

Fig. 6
figure 6

Results of cross-lagged panel model testing the indirect effect of achievement striving on life satisfaction (hedonic well-being) through competence and flow. Note. Standardized coefficients with 95% CI. T1 = Time 1 and T2 = Time 2. Cohort and school included as auxiliary variables to predict missingness. Only selected coefficients shown, see Table S4 for all coefficients

Full size image

6.2 Overall Cross-Lagged Panel Model

Achievement striving, competence, flow, and well-being at T1 accounted for about a substantial amount of the variance in well-being at T2, while considering all other variables at T2 as covariates (life satisfaction: 40%; life worth: 30%).

6.3 Exploratory Analyses

The personally expressive activities were categorized in one of seven types (Table 4). The most common types of activity fell in the physical fitness category (30.06%); the least common category was connection to people or nature (9.25%). The overall F-test suggested that there was a difference in flow across the 7 activity types, F(6,125.68) = 6.52, p <.001, ω2 = 0.20. Results were further probed with post-hoc tests. Overall, 2 of 21 post-hoc tests were statistically significant after adjusting for familywise error. The Time 1 data showed that flow was significantly higher in reading and writing activities than personal/self-care (Mdifference = 1.04, 95% CI [0.33, 1.75]) and occupational activities (Mdifference = 1.20, 95% CI [0.48, 1.92]; Fig. 7). Levels of competence did not differ by activity type (Fig. 8), F(6,124.68) = 1.27, p =.28, ω2 = 0.01. Visually, the pattern of means by activity type share one key difference across flow and competence (personal/self-care activities tended to have the lowest level of competence and flow reported) but two key differences: people rated academic/occupational and connections as higher in competence than in flow. Means and standard deviations are depicted in Figs. 7 and 8.

Table 4 Types of personally expressive activities
Full size table
Fig. 7
figure 7

Level of flow at time 1 by activity type

Full size image
Fig. 8
figure 8

Level of competence at time 1 by activity type

Full size image

7 Discussion

The purpose of this study was to test the processes through which achievement striving leads to both hedonic and eudaimonic well-being over time. Feeling competent and in flow during personally expressive activities did not increase achievement strivers’ well-being concurrently or four months later. In fact, competence did not directly influence well-being in the short or long term. Although achievement strivers tended to feel competent and happy (significant direct effects and bivariate correlations), the hypothesized mediating pathway to well-being was not supported.

7.1 Achievement Strivers Tend to Feel Competent and Happy

Broadly, research shows that achievement strivers have characteristic emotional and motivational mechanisms that predispose them for feeling competent and well-being. Though the present study did not provide evidence of a mechanism, we can speculate about alternative mechanisms. The pursuit of ambitious goals may be associated with an upward spiral of motivational resources and progress in goal-directed behaviours. Conscientious people tend to be high in trait inspiration, in addition to having high goal inspiration (Milyavskaya et al., 2012); when they feel effective and proud of their accomplishments, achievement strivers’ emotional and motivational resources may be utilized and strengthened. As Milyavskaya et al. (2012) theorized, goal progress and goal inspiration may have a reciprocal relationship which creates the upward spiral of successful goal pursuit. The overall beneficial outcome of the upward spiral is that individuals are transformed, such that they become “more creative, knowledgeable, resilient, socially integrated, and healthy” (Fredrickson, 2004, p. 153). Sheldon & Houser-Marko (2001) used a five-wave panel design to test if the upward spiral of broaden-and-build theory holds for goal striving and motivation. More specifically, if initial self-concordant motivation (i.e., goals that are aligned with one’s values and beliefs; Milyavskaya et al., 2014; Sheldon & Elliot, 1999) would indirectly predict increased well-being through goal attainment, creating a self-reinforcing cycle (i.e., increased motivation for future striving, even better attainment, and then further increases in well-being). When initiated, maintaining the upward spiral is a different story. Sheldon & Houser-Marko (2001) further found that while increasing one’s level of well-being is possible, few participants were able to further increase their well-being after the first upward spiral. These findings suggested that self-concordant motivation may be the key driver of this process, in that people need to ‘strive for the right reasons’ (p. 152).

7.2 Flow and Well-Being

Experiencing flow directly predicted life worth (but not life satisfaction) concurrently, but we found no longitudinal relationships between flow and outcomes. Research suggests people who are naturally persistent and intrinsically motivated in everyday life are particularly flow prone, which increases well-being both indirectly (Tse et al., 2021) and directly (Peterson et al., 2007; Tse et al., 2020). That is, daily diary and cross-sectional evidence suggest that flow increases well-being (Tse et al., 2020, 2021). In fact, Tse et al. (2020) suggest that well-being is shaped by the ease with which one can engage in, enjoy, and become absorbed in activities. In particular, sustainable increases in well-being is possible by engaging in and enjoying a variety of activities, rather than narrowly focusing in on one specific activity. The null indirect findings may be due to measurement error. By asking participants to name just one personally expressive activity, they named the activity that is most important to their current life goals, but did not have an opportunity to list the potentially rich and varied activities that add quality to their lives. Further, this study is drawn from a sample of post-secondary students who tend to experience high stress (American College Health Association, 2019) that can be severe (Linden & Stuart, 2019), which may increase rumination tendencies, particularly for those high in neuroticism (Zuo et al., 2024). Neuroticism is linked to low flow proneness (Ross & Keiser, 2014; Ullén et al., 2012) and to decreased well-being (Liu et al., 2023; Steel et al., 2008). Thus, our participants could be less likely to engage fully in personally expressive activities and instead, devote their effort into their studies, thus diminishing their proneness to experiencing flow and the associated psychological benefits.

7.3 Types of Personally Expressive Activities

Personally expressive activity types include those that are productive, social, athletic, arts-related, values-related, and media-related (Waterman, 1993), which can further be classified as low effort (hedonically motivated) activities or high effort (intrinsically motivated) activities (Waterman, 2005). Physical fitness and occupational activities (our two most commonly reported personally expressive activities) are intrinsically motivated, according to Waterman (2005)’s description of high effort enjoyable activities. Creative arts, personal/self-care, general interest hobbies, reading and writing, and connecting with people and places would be considered hedonically motivated activities, in that they require little effort but are enjoyed. Our participants’ high effort enjoyable activities were only associated with mid-range levels of flow and did not provide higher feelings of competence than other activity types. Flow activities are those with conditions such as goal and feedback structures that make flow more likely (Nakamura & Csikszentmihalyi, 2002). One can experience flow in virtually any activity; as Nakamura & Csikszentmihalyi (2002) stated, ‘a museum visit, a round of golf, a game of chess’ (p. 242) can all be experienced with boredom or anxiety, it is the challenge-skills balance condition which is key to flow. Levels of flow were highest in reading and writing activities, which has supports previous reports that flow is attainable when engaged in activities such as literary writing (Perry, 1999). As described by Coatsworth et al., 2006, people “typically engage in a wide range of activities that they can use to define themselves” (p. 165), and being a student is considered a social identity (White et al., 2011), experiencing flow during reading and writing may be attributable to the nature of students’ everyday lives.

7.4 Flow During Reading and Writing

Research on flow has been blossoming within language and learning research because flow-facilitating conditions are present in reading and writing (Czimmermann & Piniel, 2016; Liu et al., 2022), as well as general learning processes (Payant & Zuniga, 2022), likely due to the presence of interest. The body of literature surrounding the study of flow in additional language learning (Aubrey, 2017a, b; Cho, 2018; McQuillan & Conde, 1996; Zare-ee, 2013; Zuniga & Payant, 2021) highlights that intrinsically interesting and collaborative tasks which offer clear goals and feedback, present appropriate challenge, and support learning autonomy can create conditions for the experience of flow within the classroom. There has even been a reading-specific flow scale developed (Thissen et al., 2018). Being flow-prone (e.g., being high in trait absorption (Rheinberg et al., 2003) not only increases levels of intrinsic interest in activities, but also strengthens positive emotions (Li et al., 2019; Özhan & Kocadere, 2020). For example, after experiencing flow, people have reported increased positive achievement emotions (e.g., pride and satisfaction) in addition to a higher sense of well-being (Pekrun, 2006). Indeed, flow during reading has been linked to increased motivation for reading and learning (Piniel & Albert, 2018; Shernoff et al., 2003; Liu & Ma, 2021). On the other hand, Fink & Drake (2016) found that flow may not be achievable after a single writing session. As previously highlighted by Waterman (1993) it is purposeful and repeated engagement that is required to enter a flow state and to experience positive outcomes, such as improved well-being. This can lead to increased discovery of one’s interests, abilities, and potentials in the future (Waterman, 1993). When English language learners engage in repeated, regular reading sessions, they report flow (Kirchhoff et al., 2013), which has increased their interest and understanding of the material (Zare-ee, 2013), although flow may be particularly likely when reading for pleasure and intrinsic interest (McQuillan & Conde, 1996).

7.5 A Few Statistical Addendum

Competence was not directly related to well-being after controlling for all other variables in the model; this is most likely because achievement striving was strongly correlated with competence. That is, it did not predict unique variance above and beyond achievement striving. The weak, negative effect from achievement striving to flow (i.e., a flipped sign from the bivariate correlations) is most likely due to strong relationship between competence and flow. As an endogenous variable in the serial mediation model, flow is the residual after partialing out competence. Considering the strong effect of competence on flow, the residual variation in flow may represent a very different construct (i.e., flow for reasons other than competence). Finally, achievement strivers do experience high levels of competence and well-being, evident through direct effects and bivariate correlations, but not through indirect effects. Though many of our effects were non-significant the confidence intervals remain informative. For example, the serial mediation models’ confidence interval of standardized effects of competence on life satisfaction show that plausible values for the population slope (i.e., the true effect) is between − 0.20 and 0.26. This suggests that if a relationship does exist between competence and well-being, it is likely not larger than these values. The power simulations enabled us to rule out indirect effects derived from bivariate correlations of r =.25 cross-sectionally, but if smaller effects exist our study cannot detect them. Taken together, these findings suggest that there are not medium to large effects in the population, but our data are not informative for smaller effect sizes.

7.6 Limitations

7.6.1 Sample Homogeneity

As this study is part of a larger study on personality and well-being in university students, we recruited participants from two university participant pools and campus flyers, which provides a relatively homogenous sample in terms of gender and ethnic background. As a result, our sample is comprised of mostly White young women in statistics classes, which limits the degree to which our results apply to the general population.

7.6.2 Too Few Activities

The original Personally Expressive Activities Questionnaire asks a participant to identify five activities of personal importance that they would use to describe themselves to another person which is likely burdensome for participants. Accordingly, we only asked for one personally expressive activity, which required participants to prioritize one activity. By reducing the number of personally expressive activities, participants may have chosen one general activity that oversimplifies their idiosyncrasies.

7.6.3 Missing Data

Due to a survey software piping error and attrition, our sample size for those who have data for each variable in the tested models shrunk to 346 (cross-sectional) and 225 (longitudinal). Thus, the inconclusive results may also be attributable to the small sample size that rendered our dataset underpowered for detecting any longitudinal effects less than r =.3–0.325 (N = 244 and 225, respectively), and cross-sectional effects less than r =.25.

7.6.4 Time Lag

Finally, our survey design included surveys 4 months apart and this lag might have been too long to assess change in well-being longitudinally. Previous studies measuring positive psychological processes (e.g., savoring, need satisfaction, eudaimonic motives for activities) have found evidence of well-being boosts with lags of one day (Jose et al., 2012; Sheldon & Niemiec, 2006; Steger et al., 2008). While hedonic happiness is more malleable on a day-to-day level (as it is akin to positive emotional states), sustainable changes in eudaimonic well-being (such as life worthwhileness) are more challenging to induce and to detect.

7.7 Future Directions

Foundationally, researchers should assess if these findings hold in non-student samples. Conceptually, as we focus on the basic psychological need satisfaction of competence in keeping with our interest in ambitious types of people, assessing how different types of personality traits lead to well-being through the two remaining basic psychological needs would provide a fuller picture of how flow fits into basic psychological need satisfaction. For example, people high in openness to experience may be more prone to engaging in personally expressive activities that encourage flow states, particularly when autonomously pursued. Likewise, people high in extraversion may have a higher need for relatedness and may pursue personally expressive activities that include social connections. Methodologically, efforts to identify alternative mechanistic pathways from achievement striving to well-being would help elucidate how, when, and why achievement strivers feel happy, beyond simple competence need satisfaction. A daily diary study would help illuminate the daily nuances of these relationships, such as experiencing basic psychological need satisfaction and well-being. Painting a picture of the daily motives, activities, and experiences of achievement strivers could help identify characteristics of personally expressive activities (e.g., frequency, duration, intensity, absorption) that help produce competence-promoting experiences. Finally, a new direction in the positive psychology of everyday life is the idea of a psychologically rich life (Oishi et al., 2019). Exploring if people who are flow prone tend to pursue rich life experiences, or if those who report a psychologically rich life tend to engage in more personally expressive activities, would be an insightful avenue for building knowledge on what the nuances of well-being look like in everyday life.

8 Conclusions

Though achievement strivers in our data tended to feel happy and competent, the mechanistic pathway to well-being is not yet clear. Nonetheless, through cross-sectional and longitudinal analyses on closed-ended and open-ended data, as well as a series of power simulations, we were able to rule out medium to large serial indirect effects of achievement striving on well-being through competence and flow, but smaller effects might exist.