Should my effect size be positive or negative when the control group gains more weight than the experimental group (when it is hypothesised that the experimental group would lose more weight i.e., the difference between groups is due to more weight gain in the control rather than weight loss int the experimental group)?
As long as your wording matches the sign of the effect size, it doesn't matter.
"Women weigh 15 pounds less than men" is exactly the same as "Men weigh 15 pounds more than women"
- Solved – Confidence of Effect Size Differences
- Solved – R: Test significance of difference in slopes of same correlations in different groups
- Solved – Analyzing repeated measures ANOVA with two groups
- Solved – Is it right to use a 2-way ANOVA for 3 experimental and 2 control groups? And should the controls be included in the ANOVA
- Solved – Using R’s prop.test function to compare subsets of a population