PsyScale Studio: A Guided, Browser-Based Application for Developing and Validating Measurement Scales

Validating a measurement scale requires a long sequence of analyses, from content validity and item analysis to factor analysis, item response theory, fairness and relations to other variables. These analyses are spread across several programs, and many applied researchers find them difficult to choose, run and report. PsyScale Studio is a free web application that guides users through this sequence. Users describe their study purpose and response format (ten formats, from dichotomous tests to forced-choice and formative indices); the application then builds a purpose-specific roadmap, runs each analysis and explains every result in English or Malay with criteria, references and reporting text. Classical, exploratory and Rasch analyses run in JavaScript. Confirmatory factor analysis, measurement invariance, polytomous item response theory, differential item functioning and cognitive diagnostic models run in R compiled to WebAssembly (webR), using lavaan, semTools, mirt and CDM, while the data stay on the user’s computer. Outputs include an APA-style Word report with references and a reproducibility package with the exact R code and input of each analysis. All 12 comparisons of the JavaScript engine with R (psych, eRm) were within tolerance, and the application’s confirmatory factor analysis reproduced lavaan to four decimals. A simulated 21-item mental health scale with known properties illustrates the workflow; the application recovered its factor structure, weak item and differential item functioning. A second example applies cognitive diagnostic models to Tatsuoka’s fraction-subtraction data. PsyScale Studio is available at https://psyscale-studio.web.app.