The different development trajectory of noncognitive abilities of urban and rural children: evidence from China family panel studies

Purpose The purpose of this paper is to analyze the temporal development of noncognitive abilities of children and the development trajectory of rural and urban children's noncognitive abilities in China. Design/methodology/approach Lexis diagram is used as the research framework to depict the development trajectory of rural and urban children's noncognitive abilities in China. By employing the nationally representative longitudinal survey data, China Family Panel Studies (2010–2016), the difference of rural and urban children's noncognitive abilities is disentangled into temporal, age and cohort effects. Findings There is a significant temporal rural–urban difference in children's noncognitive abilities, but the rural–urban gap would expand, narrow or show more complex development trends under different measurements. The results of age and cohort comparison are similar to those of temporal comparison, that is, there are divergent trajectories of rural–urban gap due to the different measurements and different ages and/or cohorts. Specifically, urban children perform better in self-esteem, but rural children always have a higher social responsibility, such a contrast between urban children's weak social responsibility under the advantageous condition and rural children's strong social responsibility in the relatively disadvantageous environment. Originality/value Children's noncognitive ability is not innate but is a gradually acquired characteristic through training and cultivation. The rural–urban difference of children's noncognitive abilities implies educational issues concerning educational principles in the urban environment and the educational resources' allocation in the rural society in China. Besides, as the unidimensional measurement of children's noncognitive ability is insufficient, the systematic measurement composed of multidimensional indicators utilizing cohort data or longitudinal data would be needed.