FAQ/preport - CBU statistics Wiki

Revision 2 as of 2016-03-16 10:01:50

Clear message
location: FAQ / preport

APA guidelines for reporting post-hoc tests

The guidelines suggest only reporting statistically significant results. These can even be reported without a p-value with only an explanation of how the pairs of group means differed required. Full details are here with an excerpt reproduced below.

[SNIP]

There was a significant main effect for treatment, F(1, 145) = 5.43, p = .02, and a significant interaction, F(2, 145) = 3.24, p = .04.

Note from Shen: since ANOVA only report if there is a significant effect without revealing the details of the effect, it is often useful to further explain the effect with additional details.

“Post hoc comparisons using the Tukey HSD test (or you can replace this with t Test or t Test with Bonferroni correction) indicated that the mean score for the sugar condition (M = 4.20, SD = 1.30) was significantly different than the no sugar condition (M = 2.20, SD = 0.84). However, the a little sugar condition (M = 3.60, SD = 0.89) did not significantly differ from the sugar and no sugar conditions.”

You also want to report your results in words that people can understand, as follows.

“Taken together, these results suggest that high levels of sugar really do have an effect on memory for words. Specifically, our results suggest that when humans consume high levels of sugar, they remember more words. However, it should be noted that sugar level must be high in order to see an effect. Medium sugar levels do not appear to significantly increase word memory.”

[SNIP]