1
$\begingroup$

I have the following data:
2 conditions - experimental & control.
20 participants, each having 3 results for each condition.

Like in the following table:

| Condition    | PARTICIPANT #1 | PARTICIPANT #2 | PARTICIPANT #3 |
|--------------|----------------|----------------|----------------|
| EXPERIMENTAL | Result #1      | Result #1      | Result #1      |
|              | Result #2      | Result #2      | Result #2      |
|              | Result #3      | Result #3      | Result #3      |
|--------------|----------------|----------------|----------------|
| CONTROL      | Result #1      | Result #1      | Result #1      |
|              | Result #2      | Result #2      | Result #2      |
|              | Result #3      | Result #3      | Result #3      |
|--------------|----------------|----------------|----------------|  

I need to compare between the 2 conditions in a way that takes into account that the data is paired regarding participants (take into account that the same participant took part in both conditions).

It seems to me like the best approach is a 2 factor repeated measures anova, but i would like to be sure and to know if there are other ways to analyze this type of data.

Also, i would like to know which test should i choose if the order of the results (1-2-3) matters.

I'll appreciate any input! Thank you!

$\endgroup$
1
$\begingroup$

If you enter your title directly into goolge, you already obtain a hint on what else to do instead of a repeated-measures ANOVA (which is only the fourth hit):

enter image description here

A repeated measures t test. It is essentially equivalent to a repeated measures ANOVA (where $t^2 = F$). Furthermore, both tests in the end just test the difference between the scores against zero (so you could also do this yourself, calculate the difference and test them with a one sample t test against zero).

$\endgroup$
  • 1
    $\begingroup$ Since Google search results are individually tailored, different people generally won't get the same results in the same order - what's fourth for you won't necessarily be fourth for them - or even on the first page. Further, such results vary over time, making the image you show unlikely to be repeatable by you over time; this would seem to be particularly liable to linkrot. Could you make your answer less reliant on the particular outcome of a Google search. $\endgroup$ – Glen_b -Reinstate Monica Jul 3 '14 at 14:33
  • $\begingroup$ @Glen_b Come on. The title of OP will most likely lead to a hit involving the repeated measures t test, given that t is the only word missing in the title to hint towards the given response. The question basically answers itself given the title. Hence my short but in no way snarky response. $\endgroup$ – Henrik Jul 3 '14 at 14:38
  • 2
    $\begingroup$ I understand the impulse to answer in this form. Nevertheless I'd encourage you to base the main part of your answer around the information such a search reveals instead of making the search itself central to the answer. I see little harm in suggesting a search as a source of additional information, however. [If the answer was primarily 'try this search', it's really more like a comment.] $\endgroup$ – Glen_b -Reinstate Monica Jul 3 '14 at 14:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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