Psychology Statistical Analysis Assistant

Problem

Psychology experiments usually require more than reporting a single p value. In gaze-cueing, social-cognition, and implicit-learning studies, researchers must first check assumptions such as normality and homogeneity of variance, then choose among t tests, ANOVA, correlation, or chi-square analyses, and report effect sizes, confidence intervals, and sample sizes. When this workflow is ad hoc, researchers can miss prerequisite checks, misuse parametric tests, or treat statistical significance as evidence of large effects.

How It Works

The skill organizes statistical analysis into a repeatable chain:
- Descriptive statistics: means, standard deviations, standard errors, medians, quartiles, interquartile ranges, and distribution summaries.
- Hypothesis testing: one-sample, independent-sample, and paired t tests; one-way and multi-factor ANOVA; Pearson correlation; linear regression; chi-square tests.
- Effect size and interpretation: Cohen's d, eta squared, correlation strength, and effect-size categories.
- Power analysis: sample-size estimation, statistical power calculation, and study-design planning.

Results are returned as dictionaries containing test statistics, p values, significance thresholds, effect sizes, sample sizes, degrees of freedom, and readable interpretations. Use the assumption checks before inferential tests, and report raw statistics alongside effect sizes. It is useful for quantitative psychology design and result interpretation, but it does not replace expert judgment on experimental design, measurement, or data preprocessing.