Abstract

Pace the ancient Philosopher Plato’s believe, to know good, means to do good. Current empirical research has found a gap between moral judgment and behavior, though (Blake, 2018; Smith et al., 2013). The moral self-concept (MSC) is seen as a bridge between this moral judgment-behavior gap (Blasi, 1983; Hardy & Carlo, 2011). Yet, little is known about its early development and internal dimensions or stability. The present thesis investigates the development of the MSC in early to middle childhood. It contains explicit and implicit interviews and behavioral observations that assess the MSC and prosocial behaviors in child appropriate manners. In the first chapter, I will introduce relevant concepts and theories, and give an overview of how the self-concept and prosocial behaviors develop. I will turn to the question of why it is that children act prosocially, and what psychological mechanisms contribute to development of prosocial behaviors. To address these questions, I introduce broader theories of moral development and relevant conceptualizations. In particular, a rationalist, nativist and constructivist view are explained and discussed. As studies on the early MSC are few and far between, studies from other domains of the self-concept can help to predict the structure and development of the MSC. Previous research on the early MSC does not allow conclusions about the independence of the MSC from other self-concept domains, as previous research has not assessed the MSC in relation to other self-concept domains. Furthermore, research does not provide sufficient data about the invariance of the MSC. In relation to prosocial behaviors, I discuss different trajectories: According to the self-perception approach (Bem, 1972), the self-concept forms as a consequence of own behavior. According to the self-consistency approach (Blasi, 1983), the self-concept impacts own behavior. According to the reciprocal model (Marsh & Craven, 2006; Marsh, Trautwein, et al., 2005), the self-concept and behavior impact each other. In this thesis, I aim to investigate four questions: How does the MSC relate to other self-concept domains? Research on this questions is important in order to gain knowledge about the (in-)dependence of the MSC. What is the internal structure of the MSC? As prosocial behaviors are suggested to be three-fold (Dunfield, 2014; Paulus, 2018) and the self-concept should relate to behavior, the MSC should be three-fold as well. How stable is the MSC in early childhood? Research on other self-concept domains suggests moderate to low stability of self-concepts in early childhood. This stability and invariance need to be investigated for the MSC as well. Lastly, I will address the question, of how the MSC relates to prosocial behavior. I aim to investigate the early MSC as a piece of the puzzle that can meaningfully relate to prosocial behavior. Thus, we get closer to bridging the gap between judgment and behavior. Study one investigated the structure of the MSC and its relation to prosocial behavior in a cross-sectional sample. It aimed to answer the question if 1) the MSC differs from other self-concept domains, 2) it resembles the structure of prosocial behavior and 3) it relates to prosocial behavior meaningfully. To test these questions, I invited 127 four- to six-year-old children (study 1.1). The children responded to a puppet-interview covering domains including the physical, verbal and moral self-concept. The items concerning the MSC contained helping, sharing and comforting. Besides the explicit measure, children participated in an implicit association task (IAT), which assessed their association between good vs. bad and self vs. other. Furthermore, children were observed in situations in which they could help, share and comfort. An exploratory factor analyses revealed that 1) the MSC is separate from the other self-concept domains; 2) the MSC was divided into the three areas of sharing, comforting and helping. A regression analysis revealed that 3) the explicit and not implicit MSC related meaningfully to sharing and comforting behavior, but not helping behavior. Study 1.2 confirmed the result from question 2) with a confirmatory factor analysis with a separate sample (N = 314). The results are in line with the hierarchical, multidimensional model of self-concept. Open questions remain: how stable is the MSC in early childhood? I aimed to answer this question in study two. Furthermore, I showed meaningful relations between prosocial behaviors and MSC, yet the direction of its relation remained unclear. I aimed to address this question in study three. In study two, the focus was the development of the MSC. I investigated the structure of the MSC by conducting a confirmatory factor analysis for a three-dimensional model, testing its invariance, reliability and correlational structure over time. To do so, I conducted a longitudinal study with measurements 18 and 21 months after the first measurement. The 133 children of the sample were aged between four and six years at the first measurement. I assessed the MSC via a puppet-interview. The results showed a stable three-dimensional MSC, containing the dimensions helping, sharing and comforting. This underscores the stability of the early MSC and the measure. In study three, I investigated the developmental relations between the early MSC and prosocial behaviors. In particular, I aimed to show that the constructs are stable across time and that they related to each other. I expected that prosocial behaviors would predict subsequent MSC from the first to the second measurement (self-perception). Then, once the MSC has formed, the MSC would predict subsequent behavior (self-consistency). A cross-lagged panel analysis was conducted on the same sample as the previous studies, as the participants also partook in prosocial situations during each measurement. Results showed moderate stability of the MSC dimensions and prosocial behaviors. Contrary to our hypothesis, self-perception effects became only marginally significant, but sharing MSC at the second measurement predicted sharing behavior at the third measurement. The study shows how stable individual differences in prosocial behavior and MSC are, already in early childhood. Furthermore, these results support the self-consistency approach. I discuss the results in relation to the rationalist, nativist and constructivist approach. The results support the understanding of children as active agents, which is in line with the social-cognitive constructivist approach. Moreover, the results support the direction of self-consistency in relation to MSC and prosocial behaviors, which is a further sign that the MSC develops and stabilizes in early childhood.