Contribution
The first longitudinal test of goal contents theory’s claim about attainment, not pursuit: over one year of post-college life, attaining intrinsic aspirations (growth, relationships, community, health) tracked psychological health and attaining extrinsic ones (money, fame, image) did not, and change in basic need satisfaction mediated the intrinsic link. It separates intrinsic from extrinsic attainment and well-being from ill-being, where earlier studies used a relative index and a health composite (“Attainment of Aspirations”, “Present Research”).
Key insight
Getting what you aimed for is not uniformly good: what matters is what the goal was. In the authors’ model, intrinsic attainment pays off because it satisfies the needs, and extrinsic attainment does not satisfy them, so its success brings no well-being and some ill-being. The need-satisfaction mediator is what turns “goal contents matter” into a claim about needs.
Assumptions
- Intrinsic and extrinsic as two factors, from Kasser and Ryan 1996’s factor analysis of seven aspirations; the intrinsic label is given “because its goals are expressive of humans’ inherent growth tendency and are conducive to satisfaction of the basic psychological needs” (Intro). The classification already carries the hypothesis.
- Residualized change. “Change” from Time 1 to Time 2 is the residual of each Time 2 latent construct on its Time 1 value (fn. 6); mediation is by Baron and Kenny’s steps and MacKinnon et al.’s z′ test.
- Self-report throughout: the 21-item Basic Psychological Need Satisfaction Scale–General, the 35-item Aspiration Index (importance at Time 1, attainment at both), SWLS, six MSEI self-esteem items, PANAS, the Pennebaker symptom checklist, six STAI items (Measures).
- Well-being includes self-esteem: the latent well-being construct is life satisfaction, self-esteem and positive affect (Results).
Key results
Sample (Method)
- 246 seniors at two colleges gave permission; 164 answered at one year (Time 1), 147 at two years (Time 2); 79.9% Caucasian; unpaid.
Importance to attainment (Fig. 1)
- Structural model fit χ²(155) = 262.52, CFI = .91, NNFI = .88, RMSEA = .05. Time 1 intrinsic importance → Time 2 intrinsic attainment β = .51; extrinsic importance → extrinsic attainment β = .67 (both p < .01).
Attainment and psychological health at Time 2 (Fig. 1)
- Intrinsic attainment → well-being β = .77, → ill-being β = −.66.
- Extrinsic attainment → well-being β = .00 (ns), → ill-being β = .38 (p < .01).
Change over the year (Figs. 3, 4)
- Well-being: Time 1 → Time 2 β = .67. Change in intrinsic attainment → change in well-being β = .63; change in extrinsic attainment β = −.02 (ns). Change in intrinsic attainment → change in need satisfaction β = .44; extrinsic β = .03 (ns). Need satisfaction change → well-being change β = .83. With the mediator in, intrinsic attainment’s path falls from .63 to .16 (ns); z′ = 2.69, p < .01.
- Ill-being: Time 1 → Time 2 β = .65. Intrinsic → ill-being change β = −.59; extrinsic β = .27 (p < .07, “marginal”). Intrinsic → need satisfaction change β = .45; need satisfaction → ill-being change β = −.39; intrinsic’s path falls to −.42 (still p < .01), partial mediation; z′ = −2.26.
- Time 1 need satisfaction was unrelated to Time 2 need satisfaction (β = −.09, ns), in both change models, while well-being (.67), ill-being (.65) and attainment (.85, .70) were stable. The paper does not comment.
What the zero-order correlations show (Discussion)
- Extrinsic attainment correlates positively with well-being (mean r = .19) and with need satisfaction (mean r = .18; competence .21, autonomy .17, relatedness .15), and not with ill-being (mean r = .03). Intrinsic attainment correlates more strongly with need satisfaction (.44, .38, .41). The authors offer a statistical explanation, that the extrinsic relations “may have been nullified when both aspiration attainments were allowed to compete for variance”, alongside the SDT one.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Valuing either kind of aspiration predicts attaining it a year later | moderate | β = .51, .67; one sample, self-report |
| C2 | Attaining intrinsic aspirations goes with more well-being and less ill-being | moderate | cross-sectional and residualized-change SEM, N = 147 |
| C3 | Attaining extrinsic aspirations does not add to well-being and goes with ill-being | weak | unique-variance paths (.00, .38); the change-model ill-being path is marginal (.27, p < .07); zero-order correlation with well-being is positive |
| C4 | Change in need satisfaction mediates intrinsic attainment’s link to change in well-being (fully) and ill-being (partly) | weak–moderate | Baron–Kenny steps and z′; correlational; mediator measured at the same times as the outcome; mediator unstable across the year (β = −.09) |
| C5 | Goal content, not only goal regulation, matters for psychological health | moderate, as an addition to the self-concordance work | Discussion; relies on C2–C4 |
Concepts
- Aspiration — a life goal that organizes behaviour over long periods; “not isomorphic” with intrinsic or extrinsic motivation, which concerns the reason for a given behaviour (fn. 1).
- Intrinsic / extrinsic aspirations — personal growth, close relationships, community involvement and physical health / money, fame and image (Intro; Table 1).
- Absolute continuity — consistency of an attribute within an individual, modelled autoregressively (Analytic strategy).
Connections
- Kasser and Ryan 1996 (LIT-1977) supplies the two factors; Kasser and Ryan 2001 and Ryan et al. 1999 (Russia) the cross-sectional attainment findings this extends.
- Sheldon and Elliot 1999 (LIT-589, NOTE-448), the self-concordance model, is the regulation half; this paper claims the content half and says “the content, in addition to the regulation, of goals matters” (Broader Implications). Some data were also used in Study 3 of Sheldon et al. 2004 (fn. 4).
Bearing on the record
- NOTE-442’s summary of Ryan and Deci 2000 says extrinsic aspirations’ attainment “does not help”. This paper is the later direct test, and it qualifies that: the null is a unique-variance result with intrinsic attainment in the same model. Alone, extrinsic attainment correlates positively with well-being (mean r = .19) and with need satisfaction (.18). What the paper shows is that extrinsic attainment adds nothing once intrinsic attainment is known, and goes with more ill-being at the cross-section.
- QUESTION-369. The paper uses needs as the mediator that explains why some goal successes do not pay, which is one of the “functional” uses of need-hood the question asks about. But it inherits the question’s circularity: intrinsic aspirations are defined as those “conducive to satisfaction of the basic psychological needs” (Intro), the well-being outcome includes self-esteem, and mediator and outcome are both self-reported at the same occasions. It does not supply a criterion for need-hood that avoids the well-being measure.
- THEORY-049 and THEORY-054, which hold goal contents and regulation second-hand: this is primary evidence for the content side, at N = 147.
- THEORY-052 is not addressed: no frustration measure, no clinical outcome.
- Filed from it: THEORY-1319 (Proposed).
- No instruction for machine-learning practice.
Limitations
- Self-report only; “rather homogenous” sample of recent graduates; and “our findings were correlational, so they do not provide evidence for the direction of effects” (Limitations, the authors’).
- The mediation is weak as causal evidence: residualized change in the mediator and in the outcome are measured over the same interval, so the order between them is assumed, not observed.
- The need-satisfaction measure did not hold its rank order over the year (β = −.09), unlike every other construct. Either need satisfaction in post-college life changes almost completely in a year or the measure is noisy; either way “change in need satisfaction” here is close to “Time 2 need satisfaction”.
- Fit indices are marginal by current standards (CFI .90–.91, NNFI .86–.88).
- The extrinsic null rests on competition for variance with intrinsic attainment, which the authors acknowledge as a possible statistical explanation.
Open questions
- Does the extrinsic pattern hold with outcome measures that are not self-reports, and in older or non-student samples (the authors’ own call)?
- Would a design that changes extrinsic attainment exogenously (a windfall, a promotion) show the null on well-being, or the zero-order positive relation?
- Why was need satisfaction unstable across the year?