Educational Assessment Psychometric Reliability and Factor Analysis Script

Role: Senior Psychometrician and Quantitative Educational Measurement Specialist.

Context

• Survey instrument: {{survey_instrument}}
• Evaluated respondents: {{target_respondents}}
• Scale measurement format: {{likert_scale_levels}}
• Theoretical construct count: {{dimension_count}}
• Data file location: {{data_source_path}}
• Imputation protocol: {{missing_data_strategy}}

Task

Develop an R script that ingests response data from {{data_source_path}}, executes comprehensive item-level psychometric diagnostics, evaluates internal consistency, and fits an exploratory factor model for {{survey_instrument}} administered to {{target_respondents}}.

Method

1. Load core psychometric packages (psych, lavaan, GPArotation, tidyverse) with version checks.
2. Construct data intake pipeline reading {{data_source_path}} with automated data validation.
3. Execute preprocessing and handle non-responses using {{missing_data_strategy}} appropriate for {{likert_scale_levels}} data.
4. Compute item-level descriptive statistics, skewness, kurtosis, and item-rest correlations.
5. Calculate internal consistency coefficients including Cronbach's alpha and McDonald's hierarchical omega.
6. Perform Bartlett's test of sphericity and Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy.
7. Execute Exploratory Factor Analysis (EFA) targeting {{dimension_count}} factors with polychoric correlation matrices.
8. Apply oblimin or varimax factor rotation, outputting standardized loadings, communalities, and model fit indices.

Constraints

• Code MUST execute successfully in base R / R 4.2+ using standard tidyverse and psych ecosystems.
• Script MUST handle ordinal Likert-scale properties via polychoric correlation matrix calculations.
• MUST NOT suppress or hide item loadings below standard retention thresholds without explicit log warnings.
• Provide exhaustive commentary explaining psychometric interpretative thresholds (e.g., KMO > 0.70, Alpha > 0.80).

Output format

Provide a standalone, fully runnable R script (.R) organized into:

1. Global Constants & Parameter Setup utilizing {{dimension_count}} and {{missing_data_strategy}}.
2. Data Ingestion & Sanitization Pipeline (min 20 lines).
3. Scale Reliability & Internal Consistency Function Suite (min 30 lines).
4. Factor Analytic Engine & Model Evaluation Routine (min 35 lines).
5. Summary Output Dispatcher printing clear diagnostic tables.

Self-review

• Confirm polychoric matrix calculation is utilized when handling {{likert_scale_levels}}.
• Check that {{dimension_count}} correctly drives the factor extraction routine.
• Ensure clear error messaging if data rows contain complete column missingness.