I want to compare Student confidence observed during intervention(5 days) between control and treatment groups. This variable was recorded during 5 days of the intervention. Which test should I employ to analyse student confidence during day1, 2, 3, 4 and 5 and compare it with control group?
participants are measured here multiple times to see changes to an intervention. for this purpose, I ran Friedman and Wilcoxon signed tests. But if I want to compare two groups which non-parametric test should I run?
Hi Amna,
.How did you measure the dependent variable? how it is distributed ?
Best,
Ruben.
Since you ask for “urgent help”, a simple approach would be to compare Day 5 to Control with a Mann–Whitney test or t-test.
Friedman or Quade could be used to look at the differences over days, for the treatment group only. Treatments are Days, and Blocks are Students. Post-hoc tests to determine differences among days would be pairwise paired-sample Wilcox tests, with a p-value adjustment for multiple tests.
If I understand Your design, there are two groups.
The first group is control, the second is under intervention during five days.
1 When You analyze the second group - the most appropriate method is repeated measures(RM) ANOVA ( parametric or non-paremetric). Post-hoc comparison is executed taken into account RM and multiple testing problem.
You can use Dunnett's test for RM (or non-parametric analogue) to compare day1 with day 2,3,4,5
2 When You analyze the control group comparing with day1,2,3,4,5 of the second group - it is not RM analysis, but multiple testing problem should be taken into account too.
You can use Dunnett's test (unpaired sample) to compare control group with day 1,2,3,4,5 of the second group.
======
If You use R - look at package nparcomp - for nonparametric comparison.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS