Skip to main content

How does School Experience Relate to Adolescent Identity Formation Over Time? Cross-Lagged Associations between School Engagement, School Burnout and Identity Processing Styles


The existing research findings still do not provide a clear understanding of the links between adolescent school experience and their identity formation. To address this gap, we analyzed the dynamic links between adolescent school experiences and identity formation by exploring the cross-lagged associations between school engagement, school burnout and identity processing styles (information-oriented, normative and diffuse-avoidant) over a 2-year period during middle-to-late adolescence. The sample of this school-based study included 916 adolescents (51.4% females) in the 9th to 12th grades from diverse socio-economic and family backgrounds. The results from the cross-lagged analyses with three time points revealed that (a) school engagement positively predicted information-oriented identity processing over a 2-year period; (b) school burnout positively predicted the reliance on normative and diffuse-avoidant identity styles across the three measurements; (c) the effects were stable over the three time points and across different gender, grade, and socio-economic status groups. The unidirectional effects identified in our study support the general prediction that active engagement in learning at school can serve as a resource for adolescent identity formation, while school burnout, in contrast, can hinder the formation of adolescent identity. This points to the importance of taking developmental identity-related needs of adolescents into account when planning the school curriculum.

This is a preview of subscription content, access via your institution.

Fig. 1


  1. In general, we focused on running analyses with latent variables. However, to avoid our models being overly complex in terms of free parameters, we used item parcels instead of items as indicators of latent variables. Three parcels were formed to represent each latent variable using item-to-construct balance approach, based on corrected item-total score correlations (Little 2013). Latent variables were scaled using effects coding method (Little et al. 2006).

  2. Specifically, ICCs for school engagement were .03 for T1, T2, and T3. ICCs for low for school burnout were .03 for T1, .04 for T2, and .06 for T3. ICCs for information-oriented identity style were .04 for T1, .06 for T2 and .05 for T3. ICCs for normative identity style were .07 for T1 and T2, and .08 for T3. ICCs for diffuse-avoidant identity style were .08 for T1, .04 for T2 and .05 for T3. Small clustering effects also indicated that multilevel analyses were not appropriate in this study.

  3. Following Little (2013), we tested for configural, weak, strong and strict longitudinal invariance. Although strong (loadings and intercepts constrained to be equal across waves) and strict (loadings, intercepts, and residuals constrained to be equal across waves) invariance was not necessary for our research questions (we did not address mean level or reliability change across waves), we tested it anyway, to see if more parsimonious models could fit to the data. In addition, we tested if variances of our latent variables were equal across time, as a preliminary step to testing the equivalence of cross-lagged paths between different waves (Newsom 2015).

  4. We checked if exclusion of covariates in this model had any effect on reciprocal relationships. Specifically, we ran the same model, but without inclusion of covariates. Exclusion of covariates had a small effect on the size of the relationships, thus indicating the need to include the covariates, however, the paths that were significant and those that were not, with covariates included, remained to be the same. In addition, we checked if the links we found were moderated by groups based on gender, grade and SES. Specifically, we tested if these variables moderate the structural associations between the study variables, using multiple-group (MG) analyses. We conducted these analyses with factor scores (calculated using a measurement model with strict invariance constraints). To test for the moderation effects we compared the constrained model (in which the cross-lag associations were set to be equivalent across groups) with an unconstrained model (in which the cross-lag associations were allowed to vary across groups). The differences between the unconstrained and constrained models were not significant in the case of gender (Δχ2(20) = 21.31, p = .379; ΔCFI = .000), grade (Δχ2(20) = 20.35, p = .436; ΔCFI = .000) and SES (Δχ2(20) = 29.91, p = .071; ΔCFI = −.004). Thus, the results indicated that gender, grade, and SES did not moderate the relationships between the study constructs.


Download references


This research was funded by the Erasmus+ Programme of the European Union, strategic partnership project “Innovative Curriculum for Strong Identities in Diverse Europe (INSIDE)”, No. 2016-1-LT01-KA203-023220 and the European Social Fund under the Global Grant measure, No. VP1-3.1-ŠMM-07-K-02-008.

Authors’ Contributions

R.E. conceived of the study, participated in its design and implementation and drafted the manuscript; R.V. participated in the design and implementation of the study, performed the statistical analysis and contributed to drafting of the manuscript; I.G. participated in the design and implementation of the study and contributed to drafting of the manuscript; S.R. participated in the design and coordination of the study and contributed to drafting of the manuscript. All authors read and approved the final manuscript.

Data Sharing Declaration

The data analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Rasa Erentaitė.

Ethics declarations

Conflict of Interest

The authors declare that they have no conflict of interest.

Ethical Approval

Ethical approval was received from the Institute of Psychology, Mykolas Romeris University.

Informed Consent

Informed consent was obtained from all individual participants included in the study.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Erentaitė, R., Vosylis, R., Gabrialavičiūtė, I. et al. How does School Experience Relate to Adolescent Identity Formation Over Time? Cross-Lagged Associations between School Engagement, School Burnout and Identity Processing Styles. J Youth Adolescence 47, 760–774 (2018).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:


  • Identity styles
  • School engagement
  • School burnout
  • Adolescence