Poster #124 - Measurement Invariance Across Mother/Child and Father/Child Attachment Relationships.

• In Event: 1-169 - Poster Session 04
  In Poster Session: PS 04 Section - Methods, History, Theory

Integrative Statement

We examine the factorial structure of the new 21-item Security Scale Questionnaire (Kerns et al, 2015), exploring measurement invariance across mother/child and father/child relationships taking in consideration the dependency of the observations and the ordinal nature of the data. R packages lavaan (Rosseel, 2012) and overlapping (Pastore, 2016) were used for CFAs and the evaluation of measurement invariance, respectively. Originally, Security Scale (Kerns, Klepac & Cole, 1996; Kerns, Aspelmeier, Gentzler, & Grabill, 2001) was a 15-items self-questionnaire, for children between 8 to 14 years old, that measured attachment security behavior towards mother/father, specifically focused on safe haven behavior. Yet, fathers’ role could have been underestimated, as a reflection of the measure, since fathers may have a major role supporting their child’s secure exploration of the environment (Bögels & Phares, 2008; Bretherton, 2010; Grossman et al, 2002). We used the new 21-item version (Kerns et al., 2015) that integrates both safe haven and safe base behaviors in two factors that assess both Safe Haven Support- SHS (14 items from the original SS), and Secure Base Support- SBS (6 new items plus 1 item from the original SS). Participants were 457 children (224 girls and 233 boys) ranged from 9 to 14 years old (M=10.84, SD=1.02) from Portuguese and US samples. Taking the ordinal nature of the data, as well as the dependency of observations (same child reporting for relationship with mother and father) into account, we confirmed the two-factor structure of SSQ, although four items were unrelated to the latent structure and for that reason excluded of the final model. Regarding measurement invariance between mother and father, based on the bootstrap-based procedure, we concluded that SSQ can be used to study both child/mother and child/father attachment relationships since the factorial structure was substantially invariant. Yet, to make mother/father item comparisons we should have in consideration that the overlapping procedure highlighted two items as presenting significantly different factor loadings. Those items seem to work better for describe fathers’ SHS then for mothers’. A multilevel analysis comparing the Portuguese and the US sample showed that there was no statistic difference between countries. Although we wouldn’t be expecting for major effect of sex or age on the perceptions of parental availability, girls reported higher values for SBS when referring to their mothers. Also, when we examined age, younger children describe their fathers as more supportive (SBS and SHS) then older children, it may be that the feeling of dependence on the parental figures diminishes with age. Our findings suggest that the 17-item version of the SSQ can be considered a valid and cost-effective tool to measure perceived attachment security in middle childhood for both mother and father relationships.

Authors

• Marília Fernandes, ISPA - Instituto Universitário
  Presenting Author

• Manuela Veríssimo, ISPA - Instituto Universitário
  Non-Presenting Author

• Leandra Coelho, ISPA- WJRC
  Non-Presenting Author

• Antonio Santos, William James Center for Research, ISPA-Instituto Universitário
  Non-Presenting Author

• Marissa Gastelle, Kent State University
  Non-Presenting Author

• Kathryn A. Kerns, Kent State University
  Non-Presenting Author