EFFECT SIZE
COMPARING DATA
RELATING DATA
Interpretation sample format
Test of relationship
Based on the Shapiro-Wilk test, [Variable 1] (W = [value], p = [value]) and [Variable 2] (W = [value], p = [value]) [are/are not] normally
distributed. Given these results, [Pearson correlation/Spearman rank correlation] was selected as the appropriate test to assess the
relationship. The analysis revealed a [strength: weak/moderate/strong] [direction: positive/negative] correlation between [variable 1]
and [variable 2], with r = [r-value] and p = [p-value]. This indicates that [brief practical implication].
Test of difference (2 groups)
The Shapiro-Wilk test indicates that the data for the [first group] (W = [value], p = [value]) and [second group] (W = [value], p = [value])
[are/are not] normally distributed. Levene's test shows that the variances are [equal/unequal] (F = [value], p = [value]). Based on the
results, the appropriate test is the [Student t-test/Welch t-test/Mann-Whitney U test].
For Student t-test/Welch t-test
An independent samples t-test was performed to compare [response variable of interest] in [group 1] and [group 2]. There [was/was not]
a significant difference in [response variable of interest] between [group 1] (M = [Mean], SD = [Standard deviation]) and [group 2] (M =
[Mean], SD = [Standard deviation]); t(df) = [t-value], p = [p-value]. The effect size (Cohen’s d) is [effect size value], indicating a
[small/moderate/large] effect. This suggests that [brief practical implication].
For Mann-Whitney U test
A Mann-Whitney U test was performed to compare [response variable of interest] in [group 1] and [group 2]. There [was/was not] a
significant difference in [response variable of interest] between [group 1] (Mdn = [Median]) and [group 2] (Mdn = [Median]); U = [U-value],
p = [p-value]. The effect size (rank biserial correlation) is [effect size value], indicating a [small/moderate/large] effect. This suggests that
[brief practical implication].
Test of difference (>2 groups)
Based on the output from the Shapiro-Wilk test, the data for all groups [are/are not] normally distributed (Group name: W = [value], p =
[value]; Group name: W = [value], p = [value]; Group name: W = [value], p = [value]). Levene's test shows that the variances among the
three groups are [equal/unequal] (F = [value], p = [value]). Based on the results of the normality and variance tests, the appropriate
statistical test is [One-way ANOVA/Welch ANOVA/Kruskal-Wallis test].
For One-Way ANOVA or Welch ANOVA
A one-way ANOVA was conducted to compare [response variable of interest] across three groups: [group name] (M = [mean], SD =
[SD]), [group name] (M = [mean], SD = [SD]), and [group name] (M = [mean], SD = [SD]). The analysis revealed [was/was not] a significant
difference in [response variable] among the groups (F(df1, df2) = [F-value], p = [p-value], η² = [effect size]).
Post-hoc comparisons using the [Tukey's/HSD/Bonferroni] test indicated the following results: For significant comparisons, there was a
significant difference between [group name] and [group name] (p = [p-value], Cohen's d = [effect size]), indicating a
[small/moderate/large] effect. For non-significant comparisons, there was no significant difference between [group name] and [group
name] (p = [p-value]).
The results suggest that [brief practical implication].
For Kruskal-Wallis Test
A Kruskal-Wallis test was conducted to assess differences in [response variable of interest] among the three groups. The median scores
were as follows: group 1 (Mdn = [median]), group 2 (Mdn = [median]), and group 3 (Mdn = [median]). The results indicated [was/was not]
a significant difference (H(df) = [H-value], p = [p-value]).
Dunn's post-hoc comparisons revealed the following: For significant comparisons, there was a significant difference between [group
name] and [group name] (z = [z-value], p = [p-value], rank biserial correlation = [effect size]), suggesting a [small/moderate/large] effect.
For non-significant comparisons, there was no significant difference between [group name] and [group name] (z = [z-value], p = [p-
value]).
The results suggest that [brief practical implication].
