Sticker, R.M., Gniewosz, G. & Paulus, M. Moral Self-Concept Instability

Published: 29 June 2021| Version 1 | DOI: 10.17632/vxrx3jx397.1
Regina Maria Sticker, Gabriela Gniewosz, Markus Paulus


The moral self-concept (MSC) is an important aspect of human morality. Developmental studies have explored the MSC in early childhood and have demonstrated its relevance for children’s prosocial behavior. Yet, the structure of the MSC in early childhood remains unknown. Based on a multidimensional concept of MSC, three alternative structural patterns that look at the development of the MSC dimension over time are possible: First, a “stability pattern”, representing a stable dimensional structure early on; second, a “changing dimensions pattern”, showing a change in the number of dimensions and third, a “differentiation pattern”, representing a changing correlational relations between dimensions. This study aimed to fill a gap by investigating the stability of young children’s MSC across three measurement points. More precisely, we explored 1) the dimensional structure of MSC to investigate the stability of it; 2) the existence of a hierarchical structure; and 3) the invariance on a measurement and structural level. Using a three-wave longitudinal sample (N = 133) of children aged between four to six years at T1, we found a highly stable three-dimensional MSC model, including the dimensions helping, sharing and comforting, over three measurement points. The results show a highly stable MSC from an early point in childhood. The results do not support the differential distinctiveness hypothesis and the hierarchical structure model, but evidence invariance of three distinct moral dimensions for each measurement point. The data contain responses from a puppet-interview of the domains helping (he), sharing (sh) and comforting (co) of three measurement points (e.g., co1_2, for the second comforting item at the first measurement). Due to skewed distribution (in terms of a difference from the normal distribution) of the underlying data, the items were categorized into three instead of five responses. This was necessary, as we needed robust scales for the kind of analysis we aimed to conduct in this study. Thus, the response categories 1 and 2 (a lot and bit like the negating puppet) as well as the categories 3 and 4 (like neither/both, a bit like the affirming puppet) have been combined; response category 5 (a lot like the affirming puppet) remained unchanged.



Developmental Psychology, Self-Concept