education-psychometric-scale-development

# Education Psychometric Scale Development ## Goal Design and validate scales, tests, questionnaires, rubrics, or educational measurement instruments. ## Inputs - Construct definition - Target population - Item pool or dimensions - Response format - Pilot/validation sample plan - Existing theory/scales ## Workflow 1. Define construct and dimensions. 2. Generate item pool and expert review plan. 3. Plan pilot testing and item analysis. 4. Plan reliability: Cronbach alpha, omega, test-retest, inter-rater if relevant. 5. Plan validity: content, structural, convergent, discriminant, criterion. 6. Choose EFA/CFA/IRT/invariance analyses. 7. Draft measurement paper method/results templates. ## Tool Calls Dependency policy: inspect the current environment first. Do not install packages, clone repositories, or change the user's environment without explicit approval. Prefer a project-local Python virtual environment or R library, and record package versions used for reproducibility. R packages: ```r install.packages(c("psych", "lavaan", "semTools", "mirt", "ltm", "TAM", "GPArotation")) ``` Interactive: ```text ShinyItemAnalysis: https://shinyitemanalysis.org/ jamovi: https://www.jamovi.org/ JASP: https://jasp-stats.org/ Mplus: https://www.statmodel.com/ AMOS: https://www.ibm.com/products/structural-equation-modeling-sem ``` ## Output Format | Stage | Task | Output | |---|---|---| Include: - Construct map - Item generation table - Expert validity form - Pilot analysis plan - EFA/CFA/IRT plan - Reporting template ## Quality Rules - Do not claim validity as a property of a scale in general; state validity evidence for a use/context. - Separate item development sample and validation sample when feasible. - Match sample size to model complexity.