Abstract
Research suggests that identity and memory are deeply interconnected, but little is known about the identity processes and the memory characteristics involved in this interaction. We employed a longitudinal design to examine relations between the satisfaction of autonomy, relatedness, and competence needs in self-defining memories and identity processing styles within the domain of friendship. We also assessed satisfaction with friends to evaluate the relation of identity processes and memory characteristics to well-being in friendships. Participants were 166 students who responded twice (at about a two-year interval) to the Identity processing style Inventory-3 and the Satisfaction with Life Scale. We adapted these measures to tap into the domain of friendships, which is an important aspect of young adults’ identity. Participants also described a friend-related self-defining memory and rated the degree of need satisfaction they experienced in that event. A cross-lagged panel analysis revealed that memory need satisfaction predicted increases in informational identity processing style and friend satisfaction over time. Implications for the directive function of memory in identity are discussed.
Notes
1. This study is the follow-up of a previous study (Bouizegarene & Philippe, Citation2016) in which we assessed identity styles, self-defining memories, and life satisfaction in two domains: friendships and school. The domain of friendship accommodates a longitudinal study very well, because people tend to have friends regardless of the period of their lives. In contrast, the school domain could become irrelevant at Time 2 for some participants if they have abandoned or completed their academic program. For this reason, only the friendship domain was tested.
2. The Cronbach’s alpha coefficient formula is strongly affected by the number of items of a scale (Cronbach, Citation1951) such that scales with fewer items have lower alphas. Research suggests that the alpha systematically underestimates internal consistency, and that Mcdonald’s Omega represents a more precise indicator of internal consistency (e.g., Dunn, Baguley, & Brunsden, Citation2014; Zinbarg, Revelle, Yovel, & Li, Citation2005). Considering that all the scales we use in the present study are composed of a low number of items we chose to use the Mcdonald’s Omega indices.
3. The Omega coefficient is interpreted in the same way the alpha coefficient is.