Source
Chen et al. (2015), LIT-2000, pp. 220–233; read in NOTE-1527.
What was actually shown
Study 1 (Belgium 359, China 324): autonomy and competence satisfaction predicted a latent well-being of self-esteem and depressive symptoms (β = .36, .41), relatedness did not (.04); paths constrained equal across the two countries fit as well (Δχ²(3) = 3.63); valuation × satisfaction interactions were null (β = .01, .05, −.02).
Study 2 (USA 298, China 309, Belgium 200, Peru 244): with health, financial satisfaction, income, age and gender controlled, composite satisfaction related to life satisfaction and vitality and composite frustration to depressive symptoms (reconstructed paths .51, .73 and .60). Per need, satisfaction of all three predicted vitality and autonomy and relatedness satisfaction predicted life satisfaction; frustration of all three predicted depressive symptoms. Constraining paths equal across four countries did not worsen fit (Δχ²(18) = 27.74; 20.29; Δχ²(9) = 3.26). Desire × satisfaction and desire × frustration interactions were null overall and in each country.
Either test could have come out otherwise: a significant interaction, or a worse fit under constraint, would have shown moderation.
What this does not say
- That the relations are the same size everywhere. Non-significant difference tests with groups under 150 do not bound the difference.
- That satisfying the needs causes well-being: the data are cross-sectional self-reports.
- That need strength never moderates. Only explicit valuation and desire were tested; implicit motives are set aside (pp. 231–232), and desire is itself correlated with frustration.
- That the needs are needs and not universal enhancers. Non-moderation by preference is consistent with both (QUESTION-369).
- That the frustration dimension is valid. See THEORY-1338.