Parenting, Identity Formation, and College Adjustment: A Mediation Model with Longitudinal Data

Abstract

Past research showed the importance of parenting towards adolescent identity formation, which, in turn, has been related to numerous indicators of adjustment. However, no study to date has tested an integrated mediation model that traces how parenting affects adjustment through processes of identity formation. The present study tested such a model using three-wave longitudinal data in a sample of college students and focused specifically on the domain of education. Consistent with expectations, the influence of parenting (i.e., the autonomy support vs. psychological control dimension) on academic adjustment was mediated by the level of identification with a chosen college major, the latter being an important aspect of emerging adults' identity formation. Practical implications are discussed and suggestions for future research are provided.

Notes

¹This may seem to be in contrast with Silk, Morris, Kanaya, and Steinberg (2003) who found that autonomy granting and psychological control are distinct and nearly orthogonal constructs. Importantly, however, in their study, autonomy granting was operationalized as parents' support of independence, that is, the extent to which parents want their children to act free from guidance by parents. In contrast, the conceptualization of autonomy support that is used in the present study has its roots in Self-Determination Theory (SDT) and defined autonomy support as the promotion of volitional functioning, that is, the extent to which parents are empathic to their children's needs and encourage their children to act upon personally endorsed values and needs (Soenens et al., 2007). Defined in this way, parental autonomy support is highly incompatible with and, consequently, correlates strongly negative with psychologically controlling parenting, which characterizes parents who lack any sense of attunement to their children's needs and, instead manipulate and pressure their children to act in accordance with the parents' own wishes and norms (Vansteenkiste et al., 2005).

²Autonomy support versus psychological control was divided among 4 parcels (i.e., two parcels for psychological control and two parcels for autonomy support), two of which contained 4 items, one containing 3 items, and one containing two items. Identification with commitment was divided among 3 parcels, two of which contained 3 items and one containing 2 items. Exploration in depth was divided among 3 parcels, two of which contained 2 items and one containing 1 item. Both academic adjustment and social adjustment were each divided among 3 parcels, two of which contained 3 items and one containing 4 items.

At Time 1, the correlations among the different items parceled together (across all parcels) ranged between .11 and .62 (all ps < .05). At Times 2 and 3, these correlations ranged between .13 and .63 (all ps < .05) and between .11 and .60 (all ps < .05), respectively. So, in line with Little, Cunningham, Shahar, and Widaman (2002), items not correlated with one another were not parceled together.

*p < .05.

**p < .01.

***p < .001.

*p < .05.

**p < .01.

***p < .001.