Abstract:

Researchers have emphasized the importance of early social-emotional and self-regulation skills. However, teachers might not accurately assess all areas, leaving room for standardized assessments. Previous studies have shown that e-assessment can be successfully used to assess young children’s skills.

In Estonia, early childhood education teachers and specialists have access to an e-assessment instrument that includes a standardized, norm-referenced, computer-assessed test for assessing five-year-old children’s social skills. The test, called Playground, comprises seven constructs based on the social information-processing model of children's social adjustment and follows a continuous narrative featuring two main characters in a playground setting. Children are presented with social situations in which they have to choose a prosocial, neutral or negative response. All items are age-appropriate, child-friendly and include audio instructions, illustrations or animations.

The aim of this research was to confirm the reliability and structure of the Playground test and to validate its results against teachers’ ratings. Two studies were conducted. In the first study, 375 children completed the test on a tablet with individual guidance from teachers. In the second study, 45 children completed the test and their teachers filled out the Social Skills Questionnaire. In both studies, teachers received training and a manual before guiding each child individually. Permissions were obtained from kindergartens, teachers, parents, and children. Special attention was given to involving children in research on their own skills, ensuring ethical, respectful, and age-appropriate participation.

In the first study, the test demonstrated acceptable reliability based on McDonald’s omega (ω = .84). The test structure was confirmed using Confirmatory Factor Analysis, and the final model showed an acceptable fit according to common indices: CFI = .92, TLI = .90, RMSEA = .04, SRMR = .04. In the second study, the factors confirmed in the first study were compared with those of the Social Skills Questionnaire using Pearson correlation to examine relationships between children’s test results and teachers’ ratings. The findings indicated that teacher-rated self-centeredness was moderately related with children’s empathic behaviour (r = .41, p < .01) and response construction (r = .40, p < .05). No significant relationships were found between other child-assessed and teacher-rated constructs.

Although previous studies have also reported discrepancies between children’s actual skills and teachers’ ratings, differences in assessment methods should also be considered. Children were required to make a prosocial, neutral, or negative choice in each situation, whereas teachers rated the frequency of observed behaviours, which may explain the disparity. The reliability and structure of the Playground test was confirmed, while the results regarding the validity were inconclusive and need further investigation.

Keywords:

E-assessment, social skills, early childhood education, tablet-based assessment, reliability and validity.