The Secret Weapons of Research: Statistical Tools Revealed

Parametric vs. Non-parametric Tests00:00:00

Explains the two families of statistical tests. Parametric tests offer more reliable results but require strict assumptions (normality, homogeneity of variance). Non-parametric tests are used as alternatives when these assumptions are not met, though they are based on data ranks rather than values.

Testing Assumptions: Normality and Homogeneity00:03:10

Details critical prerequisites for parametric tests, specifically the Shapiro-Wilk and Kolmogorov-Smirnov tests for normality, and Levene's test for the homogeneity of variances.

Comparative Statistical Tools00:05:32

Covers essential tools for finding significant differences, including independent and paired t-tests, ANOVA for comparing multiple groups, repeated measures ANOVA, and ANCOVA for controlling covariates like pre-test scores.

Correlation and Categorical Analysis00:12:20

Discusses Pearson’s R for measuring relationships between continuous variables and Chi-square tests for categorical variables. Also distinguishes between simple correlation and regression analysis, where the latter establishes predictive, causal directionality.

Reliability, Validity, and Sample Size00:18:07

Explains tools for instrument reliability (Cronbach's alpha, KR-20/21) and inter-rater reliability/validity (Krippendorff's alpha). Concludes with sample size determination using Cochrane’s and Yamane’s formulas.