Here, we’ll test whether there’s a statistically significant difference between the conditions in social dominance orientation:

Condition_x_SocialDominance <- aov(SocialDominance ~ Condition, na.action=na.exclude, data= analysis_data)
summary(Condition_x_SocialDominance)

##             Df Sum Sq Mean Sq F value Pr(>F)
## Condition    2    5.3   2.650   1.225  0.299
## Residuals   86  186.1   2.164

To report an ANOVA properly, you’ll need the p-value (Pr(>F)), the F-value, the between-groups degrees of freedom (Df x Independent Variable – here, it’s Condition), and the within-groups degrees of freedom (Df x Residuals) You’ll also need the grouped means and the grouped standard deviations (which you’ll need to calculate yourself; see descriptive statistics page).

If you get a statistically significant result, you should also run a Tukey post-hoc analysis:

Condition_x_SocialDominance_Tukey <- TukeyHSD(Condition_x_SocialDominance)
Condition_x_SocialDominance_Tukey

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = SocialDominance ~ Condition, data = analysis_data, na.action = na.exclude)
## 
## $Condition
##           diff        lwr       upr     p adj
## 1-0 -0.0237069 -0.9372924 0.8898786 0.9978906
## 2-0  0.5041667 -0.4016435 1.4099768 0.3839524
## 2-1  0.5278736 -0.3857120 1.4414591 0.3568535

You won’t need to report any of these numbers, but you will need to report which condition comparisons get statistically significant results.