Step 6: Hypothesis Testing
Correlations
Use a Pearson correlation for data where both the independent and dependent variables are continuous.
Reporting in APA format:
There was [not] a significant [positive/negative] correlation between [independent variable] and [dependent variable] (r = .###, p < .##), indicating that [participants/specific population] who [experienced/showed/demonstrated] [more/less of] [independent variable] tended to have [lower/higher] [dependent variable].
T-tests
Use a t-test for data where you have one categorical independent variable that has only two data levels (i.e. is only manual vs automatic, rather than green vs yellow vs orange) and one continuous dependent variable.
Reporting in APA format:
[Independent variable category 1] (M = ###; SD = ###) [do not] [report/have/show/feature] significantly [more/less] [dependent variable] between than [independent variable category 2] (M = ###; SD = ###), t = ###, p < .##.
ANOVAs
Use a one-way ANOVA for data where you have one categorical independent variable that can have two or more data levels and one continuous dependent variable.
Reporting in APA format:
A One-way ANOVA showed there was [not] a statistically significant difference in means on [dependent variable] based on [independent variable], F([between groups df],[within groups df]) = ####, p= .###, such that [independent variable category 1] (M= ###; SD = ###) [had/ did not have] significantly different average [scores/rates/etc] than [independent variable category 2] (M= ###; SD= ###) [and/or] independent variable category 3] (M= ###; SD= ###).
If you got a significant result, see how to report your Tukey test at the end of the Factorial ANOVA section:
Factorial ANOVA
Use a factorial ANOVA for data where you have more than one independent variable in your analysis.
Reporting in APA format:
A two-way analysis of variance was conducted on the influence of two independent variables ([independent variable 1], [independent variable 2]) on the [level / accuracy / number / performance] of [describe dependent variable here]. [Independent variable 1] included [3] levels ([list the levels/conditions here in words]) and [independent variable 2] consisted of [#] levels ([list the levels/conditions here in words]). [All/No/Some] effects were statistically significant at the .05 significance level [except forlist non-significant independent variables here]. The main effect for [independent variable 1] yielded an F ratio of F(between-groups df, within-groups df) - ####, p = ###, indicating there was [not] a significant difference betweeen [condition/level 1] (M = ####, SD = ####),[condition/level 2] (M = ####, SD = ####), and [condition/level 3] (M = ####, SD = ####). The main effect for [independent variable 2] yielded an F ratio of F(between-groups df, within-groups df) - ####, p = ###, indicating there was [not] a significant difference betweeen [condition/level 1] (M = ####, SD = ####),[condition/level 2] (M = ####, SD = ####), and [condition/level 3] (M = ####, SD = ####). The interaction effect was [not] significant, yielded an F ratio of F(between-groups df, within-groups df) - ####, p = ###.
Reporting A Tukey test
APA report recipe: Post hoc comparisons using the Tukey HSD test indicated that the mean score for the [condition/level 1] (M = condition DV mean, SD = condition DV SD) was significantly different than the [condition/level 2] (M = condition DV mean, SD = condition DV SD). However, the [condition/level 3] (M = condition DV mean, SD = condition DV SD) did not significantly differ from the[condition/level 1] and [condition/level 2].
OR
Post hoc comparisons using the Tukey HSD test indicated that the mean scores of the [condition/level 1] (M = condition DV mean, SD = condition DV SD), the _____ condition (M = condition DV mean, SD = condition DV SD), and the _____ condition (M = condition DV mean, SD = condition DV SD) were all significantly different from each of the other two conditions.