The Development of Computational Thinking among Middle School Students Creating Computer Games

Principal Investigator: 
Project Overview
Background & Purpose: 

The focus of this research project is on computational thinking, an approach to problem solving that is widely considered to be an essential part of life in a digital age. Despite the intuitive appeal of computational thinking, its utility is limited because little research has been done on how to define, measure, or promote it in the middle school years. Three research questions will be addressed:

  1. What is a developmentally appropriate definition of computational thinking in middle school?
  2. Does creating a computer game promote computational thinking?
  3. Does pair programming promote more computational thinking than solo programming?

The study is being conducted in extended learning settings on school campuses in rural and suburban settings. Students are voluntary participants in an after school computer game design classes or in-school computer classes. We plan to recruit 240 students in grades six through eight. If representative of the schools, they will be 50% female, 70% Latino/a, and 25% white.

Research Design: 

This is a comparative project designed to generate causal [experimental] evidence. The study contains an intervention [computer game designing with a partner] and a comparison condition [computer game designing alone]. Original data are collected through assessments of learning or achievement tests, observation [personal, audio] and survey research [on-line survey, structured interview administered questionnaires] and computer games created by the students. The online survey draws on our own items, as well as those from Barron (2008) and Bukowski, Hoza, & Boivin (1994).

Analysis of qualitative data will follow Greene’s (2007) seven phases of data analysis: data cleaning, data reduction, data transformation, data correlation and comparison (e.g., typology development and extreme case analysis), and higher order analyses that lead to inquiry conclusions and inferences. Analysis of quantitative data will include multiple group longitudinal APIM analyses for indistinguishable dyads. These analyses represent a modification of the Actor-Partner-Interdependent Model (APIM), a data analytic technique designed for nonindependent data, which simultaneously estimates the effect than an individual’s predictor variable has on his or her own outcome variable and on his or her partner’s outcome variable, partialling out variance shared across them in the predictor and in the outcome variable (Kenny & Cook, 1999). Structural equation modeling will be used to enact a multiple groups APIM procedure that contrasts paths within and between subgroups that contain indistinguishable dyads (Neyer, 2002) in order to test for moderator effects.


This project has not yet generated findings.

Publications & Presentations: 

There are no project publications at this time.