I really suck at this so hope you could help me

ve a group of people attending a training course, in order to measure whether the training course is effective in improving attendees' knowledge. We did the following:

attendees will complete a pre-test (TRUE or FALSE only) before the training
attendees attending the training
attendees will complete a post-test (TRUE or FALSE only) after training

Both pre and post tests are having exactly the same questions, However the test is NOT tagged with barcode so i have NO way to know if a certain pre and post questionnaire belong to the same individual

Each paper has been scored out of 20 ,, and the scores i have entered in SPSS for every student for both the pre and post questionnaire

So for this can i just use the paired t test and will the interpretation still be valid or is there an alternative i can use to find the significance of the intervention and whether it was effective or not? or should i do the t-test to the sum of the scores for pre and post?

Thanks alot

  • $\begingroup$ You can't use the paired t-test in this case, standard t-test will have to be it. $\endgroup$
    – Affine
    Mar 26, 2014 at 18:18
  • $\begingroup$ Also, since presumably the goal is to measure if there is an impact of training on test scores - having the post-test have the same questions as the pre-test will cloud the interpretation (is it really the training that improved scores or familiarity with the test?) $\endgroup$
    – Affine
    Mar 26, 2014 at 18:19
  • $\begingroup$ Thanks ,, but how can the familiarity with the test affect the scores ? because the only way someone can get a better score is if they have improved their knowledge via the intervention which was done or through luck by chosing randomly a different answer - so that way i can know if my intervention was generally effective or not $\endgroup$
    – Ahmad
    Mar 26, 2014 at 18:25
  • $\begingroup$ Familiarity with the test format can help. For example, there's a well-known "bump" in SAT scores when students re-take the exam, even if they haven't studied. The actual questions are different, but students now know a bit more about the structure of the test. $\endgroup$ Mar 26, 2014 at 18:28

1 Answer 1


How would you go about pairing them? A paired t-test asks whether the mean difference between paired observations is different from zero. For example, you might weigh people, put them on a diet, and then weigh them again some time later. An unpaired t-test would ask whether the average weight of the "before" group differs from the average weight of the "after" group. A paired t-test, however, compares each person's weight loss/gain.

It doesn't sound like you have any sensible way of pairing your questionnaires, so you cannot use a paired t-test. This isn't necessarily the end of the world--you can use an unpaired t-test instead: this would tell you whether there was a significant different in the two total scores.


Your Answer

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

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