This study investigated whether two educational strategies, providing material and active adult support during play, promote pretend play quality in a group of preschoolers. The sample consisted of 101 preschoolers (50% females; mean age at t1= 43.02 months, SD= 5.94) from 14 Swiss educational play groups. These were randomly allocated to the experimental conditions: material (n=4 groups, n=35 children), play tutoring (n=5 groups, n=32 children), and treatment as usual (control; n=5 groups, n=34 children). Children’s and adults’ behavior during six play sessions, each lasting 30 minutes, was videotaped. Children’s play activities, social pretend play quality (SPPQ) and adults’ level of active pretend play support, were assessed using standardized observation tools. Research Findings: Material and play tutoring groups showed significantly higher frequency of pretend play than the control group. Regarding SPPQ, the material group scored significantly lower than the play tutoring group, but significantly higher than the control group. Additional analyses showed that differences in SPPQ between the control and material condition can be fully explained by adults’ active play support. Practice or Policy: The results suggest that providing roleplay material stimulates children to engage in social pretend play, but adults’ play support increases the quality of social pretend play.

The quality of social pretend play may have a positive impact on children's development. This study investigated to what degree this quality is a characteristic of a child versus a function of the play partner or the specific pairing of two children. For this purpose, preschool children's general pretend play quality (actor effect), their general influence on their play partner's pretend play quality (partner effect) and the role of the specific constellation between two children (relationship effect) were examined in a dyadic setting. Potential associations of the effects with children's age, gender, social cognitive and social competencies, as well as the dyadic composition of these variables and their friendship were examined. Children completed two to four dyadic pretend play situations with different peers. They were tested for their language ability, theory of mind and emotion understanding. Educators rated children's social competence (cooperation and sociability) and evaluated their friendship with one another. A social relations model analysis was conducted. The results indicated that children's pretend play quality was determined to the same degree by the child and by the specific pairing of two children. Positive associations were found between children's pretend play quality and age, emotion understanding and sociability on the individual level. Further, the quality of children's social pretend play benefited from having older and more advanced play partners.

The alternate uses task (AUT) is the most popular measure when it comes to the assessment of creative potential. Since their implementation, AUT responses have been rated by humans, which is a laborious task and requires considerable resources. Large language models (LLMs) have shown promising performance in automatically scoring AUT responses in English as well as in other languages, but it is not clear which method works best for German data. Therefore, we investigated the performance of different LLMs for the automated scoring of German AUT responses. We compiled German data across five research groups including ~50,000 responses for 15 different alternate uses objects from eight lab and online survey studies (including ~2300 participants) to examine generalizability across datasets and assessment conditions. Following a pre-registered analysis plan, we compared the performance of two fine-tuned, multilingual LLM-based approaches [Cross-Lingual Alternate Uses Scoring (CLAUS) and the Open Creativity Scoring with Artificial Intelligence (OCSAI)] with the Generative Pre-trained Transformer (GPT-4) in scoring (a) the original German AUT responses and (b) the responses translated to English. We found that the LLM-based scorings were substantially correlated with human ratings, with higher relationships for OCSAI followed by GPT-4 and CLAUS. Response translation, however, had no consistent positive effect. We discuss the generalizability of the results across different items and studies and derive recommendations and future directions.

This article describes how a series of randomized controlled trials (RCTs) were implemented in the Hector Children's Academy Talent Development Program with the goal of generating mutual benefits for research and practice. The Hector Children's Academy Program, founded in 2010 and located in one of the largest states in Germany, Baden‐Württemberg, is a statewide enrichment program for talented primary school students, with a focus on STEMM topics. The program is financed by a private foundation, supervised by the Ministry of Education, and offered by 69 local sites that are hosted by local schools. About 15,000 primary school students (Grades 1–4) attend enrichment courses (more than 23,000 course participations) offered by the Hector Children's Academy Program every year. A unique element of the Hector Children's Academy Program is the role of empirical research in course development. The Hector Core Courses—offered by all local sites—undergo a strict quality assurance process in which RCTs are used to test their effectiveness with regard to central outcomes. After describing the Hector Children's Academy Program, we explain how the Hector Core Courses were developed and incorporated into the program, summarize key findings from the RCTs, and discuss mutual benefits for research and practice.