Correct answer: (B) Spearman’s rho; (C) Kruskal-Wallis H test
Explanation
Spearman's rho works with ranks, and the Kruskal–Wallis H test compares group rank sums. They do not require a normal population distribution for their usual distribution-free null tests. These are the intended choices.
Pearson's correlation and ordinary ANOVA are the parametric alternatives in this question. Their conventional exact significance tests use distributional assumptions, unlike the rank-based procedures. This concerns inference: the Pearson coefficient itself can still be calculated for non-normal data.
More Statistics in Psychology: Central tendency and dispersion, normal probability curve, parametric and non-parametric tests, effect size and power analysis questions