+ Reply to Thread
Results 1 to 2 of 2

Thread: Which statistical test to use?

  1. #1
    Points: 4, Level: 1
    Level completed: 7%, Points required for next Level: 46

    Thanked 0 Times in 0 Posts

    Which statistical test to use?

    Hi everyone,

    we ran a within-subjects user study with three different independent variables (A, B, and C). We measured e.g. task completion times and video coded all sessions. For task completion times, we used a Friedman test and Wilcoxon signed-rank tests for post-hoc comparison (our data was not normally distributed).

    Our video analysis revealed three different types of strategies (1, 2, and 3), leading to unequal sample sizes (e.g. 10 participants worked with strategy 1, 20 with strategy 2, and so on).

    Now we wanted to run statistical tests on our dependent variables (e.g. task completion times) based on the different types of strategies...Any idea which statistical test could fit?

    Thanks in advance!

  2. #2
    Points: 1,903, Level: 26
    Level completed: 3%, Points required for next Level: 97

    New Zealand
    Thanked 47 Times in 46 Posts

    Re: Which statistical test to use?

    If you have already used Friedman's test, you are looking at repeated measures so I imagine that A, B, and C are three categories inside one independent variable. When you have the subjects put into three categories you can use Kruskal-Wallis to test strategies against A, B, and C separately or all together using the average. One thing you are probably aware of is the problem of multiple p values so you should consider making the individual tests more stringent.

+ Reply to Thread


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Advertise on Talk Stats