In an experiment, 25 subjects (Group A) are assigned to play an adventure game (with audio and non-audio) while another 25 subjects (Group B) are assigned to play a strategy game (with audio and non-audio). Each game has two levels (audio and non-audio).

The independent var are: (i) groups (A and B) (ii) game types (with two levels: adventure and strategy) and (iii) game versions (with 2 levels: audio and non-audio). Group A and Group B are independent of each other.

  • One within-subjects factor is the game version (audio and non-audio)

  • One between-subjects factor is the game types (adventure and strategy)

  • One between-subjects factor is the group (A and B)

Can this research design be considered a 3-Way mixed-ANOVA, with one within-subjects factors and two in-between subjects factors?
Kindly advise

  • 1
    $\begingroup$ Group and game type are identical so you can drop one. $\endgroup$ Commented Jun 23, 2014 at 21:48
  • $\begingroup$ So we can say it is a 2-Way Mixed ANOVA, with One within-subjects factor is the game version (audio and non-audio) and One between-subjects factor is the game types (adventure and strategy). Thanks $\endgroup$
    – user39531
    Commented Jun 23, 2014 at 21:50
  • 1
    $\begingroup$ Yes, I would concur. If the scale is continuous the analysis can be simplified by computing the difference between audio and non-audio and boil it down to an independent sample t-test, using score difference as dependent, game type as independent. Your proposed way of analysis is also fine. $\endgroup$ Commented Jun 23, 2014 at 21:52
  • $\begingroup$ The independent variable "Game Types" consists of a strategy game and an adventure game. Do we not refer "audio and non-audio" as levels instead of a second set of independent variable? If in case audio|non-audio are considered independent var, then in that case it will be a 2x2 mixed ANOVA, rather than 2-Way Mixed ANOVA. Otherwise,if audio| non-audio are considered levels, then what will be the research design? Please confirm – $\endgroup$
    – user39531
    Commented Jul 16, 2014 at 22:46
  • 1
    $\begingroup$ Two subtleties, maybe important, maybe not. (1) There are 50 subjects and each of them is randomly chosen to play one kind of game? Or there are two groups of 25, one of them chosen to play one kind? GROUP only makes sense in the second, but is confounded with TYPE [because there are no replicated groups within TYPE]. (2) Each of the subjects plays both game versions (audio AND non-audio) or eah subject plays only one game version (audio OR non-audio)? SUBJECT (and, with that, the within-between design) only makes sense if the first option; else, is just a two-factor ANOVA. $\endgroup$
    – FairMiles
    Commented Jul 20, 2014 at 22:51

1 Answer 1


Based on the information in your question (and assuming there are no other variables differentiating A from B), I would say:

It is a 2x2 mixed ANOVA.

  1. Type of game

    • Adventure
    • Strategy
  2. Music

    • On (or audio)
    • Off (or non-audio)

With Type and Music as independent variables (each with two levels).

You can then look into the effects of Type, Music ; and also the Type-Music interaction effect on the dependent variable (which I imagine is score or performance or something similar).


If it helps, you can find more information and research examples (including 2x2 mixed ANOVA):

Edit: PS: @FairMiles makes a very good point in the comments to your original post.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.