I am currently doing a research on effects of writing styles on the perception of reader.

Currently I have the following 4 groups

Group 1: Pre Test - Treatment 1a - Post Test
Group 2: Pre Test - Treatment 1b - Post Test
Group 3: Pre Test - Treatment 2a - Post Test
Group 4: Pre Test - Treatment 2b - Post Test

Treatment 1 refers to a normal writing style and Treatment 2 refers to the enhanced writing style. 

I understand the most researchers will assume that pretest for all 4 groups are the same due to randomly chosen participants. However, what it can show is only that the treatments influence the participants.

Is there any statistical tests I can use to test that treatment 1 influences the participants more than treatment 2?

Thank you! any help is greatly appreciated.

  • 1
    $\begingroup$ Do you have a control group? $\endgroup$ Jan 20, 2013 at 16:19
  • $\begingroup$ Treatment 1 is the control group. $\endgroup$
    – Wilson
    Jan 21, 2013 at 15:56
  • $\begingroup$ What do the small a and b refer to? Any reason why the methods discussed in stats.stackexchange.com/questions/3466/… do not apply? $\endgroup$
    – Gala
    May 21, 2013 at 9:56

1 Answer 1


One option is to use ANOVA with orthogonal contrasts. In this case, this would be an example of "planned multiple comparisons". See http://web.uct.ac.za/depts/psychology/psy400w/stats/anova/multcomp.pdf

  • $\begingroup$ Hmm. a question. Do i take the difference between pre test and post test as the "sum" in the example given in the link? thank you. $\endgroup$
    – Wilson
    Jan 21, 2013 at 15:57
  • $\begingroup$ Will you do calculations by hand or are you using any particular software (SPSS, Excel, R)? It would make things easier to explain. $\endgroup$
    – noumenal
    Jan 21, 2013 at 19:04
  • $\begingroup$ You need to calculate one sum for each treatment. You then apply the contrast on these sums. $\endgroup$
    – noumenal
    Jan 21, 2013 at 19:06
  • $\begingroup$ I am using SPSS. $\endgroup$
    – Wilson
    Jan 22, 2013 at 5:40
  • $\begingroup$ I hope this helps: statisticshell.com/docs/contrasts.pdf $\endgroup$
    – noumenal
    Jan 22, 2013 at 7:53

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.