Authors:
Evelien Dietvorst | Tilburg University | Netherlands
PhD Marieke Hiemstra | Utrecht University
Dr. M. H. J. Hillegers | University Medical Center Utrecht | Netherlands
Dr. Loes Keijsers | Tilburg University
Parenting processes take place at the level of the family unit. Existing longitudinal studies are increasingly critiqued for producing results that are unrelated to the actual causal mechanisms at the level of individuals or individual families (Hamaker, Kuiper, & Grasman, 2015). In fact, some studies on parenting have addressed this concern (e.g., Keijsers, 2015; Smetana, Villalobos, Rogge, & Tasopoulos-Chan, 2010) which provides the intriguing suggestion that inferences based on research focusing at the group level (e.g., regression model), may be different from, or even contradictory to, the actual parenting processes at the level of individual families.
This symposium contribution applies novel Random-Intercept Cross-lagged panel models (Hamaker, Kuiper, & Grasman, 2015) on the link of privacy invasive parenting with adolescent secretive behaviours, to come to better estimates of how parenting operates within families. By extending a standard cross-lagged panel model, to distinguish the between-person from the within-person effects, this novel method allows to critically evaluate whether (H1) there are indeed positive effects of privacy invasive parenting on secrecy at the level of a family unit.
Dutch adolescents (n =244, mean age = 14.07, 40% boys) reported three times on perceived parental privacy invasion and secrecy from parents.
The standard cross-lagged panel model on adolescent-perceived privacy invasion and secrecy confirmed earlier studies (e.g. Hawk et al, 2013) that privacy invasive parenting predicts increased secretive behaviours over time. However, RI-CLPM, designed to improve causal inferences show a different pattern. The positive between-person correlation confirmed earlier work that in families with more secrets, more privacy invasive behaviours occur. Within-persons, however, children were more secretive in periods with lower levels privacy invasion (significant correlated change).
In contrast to inferences in earlier work, secrecy was related to decreased privacy invasion within families. Findings of standard cross-lagged panel models may thus be opposing to the processes operating within families.