Measuring children's social skills using microcomputer-based videodisc assessment.

Chandler et al. (1992) built a computer program that shows short video clips to kids. The kids watch real-life social scenes on a TV hooked to a laserdisc player. After each clip the child picks what to do next by touching the screen. The machine records every choice. The goal was to measure social skills in a way that feels like a game, not a test.

The paper only describes how the tool works. No children were tested and no scores were reported. The authors explain the hardware, the clip library, and the scoring logic. They say the setup is ready for future studies.

Fyfe et al. (2007) took the same video-assessment idea and added real data. They filmed 97 girls with Rett syndrome and showed the videos to parents and experts. Inter-rater reliability came out strong, giving the 1992 concept its first solid proof of validity.

Snyder et al. (2012) swapped social-skills clips for toy clips and ran a video-based MSWO. Four kids with autism picked toys on screen; three matched later real choices. The 2012 study turned the 1992 prototype into a quick preference screener.

Machado et al. (2021) kept the computer angle but trained adults to score behavior videos at 5× speed. Observer error dropped below 11 % and review time fell by two-thirds. They extended the 1992 hardware focus into faster data handling.

You now have a roadmap. Start with short video clips on a tablet instead of paper checklists. Copy Snyder et al. (2012) and run a two-minute video MSWO to find reinforcers. If you need detailed social skills data, follow Fyfe et al. (2007) and collect reliability stats from two raters. The 1992 paper gives you the blueprint; later studies show it works when you add validation steps.