For my thesis I am conducting my analysis, and am having problems with the second part of my analysis. For my second effect, I am looking at the effect of tone of voice (informal vs formal) on purchase intention, moderated by the Schwartz culture value of power.
Can I use a two way anova here? As the construct of power values has continuous and therefore not grouping variables.
Please help!!
Note: ANCOVA assumes a linear relationship between contrinuous DV and IV (for each tone-group). You may check that and, if required, linearize the relationship by transforming the variables. Another assumption is that the conditional distribution of the DV is normal. Check that this is at least approximately the case for "income intention".
Dear Isabel
Yes you can use ANOVA on continuous IV but the selection of type of ANOVA (one way, two way and so on...........) depends upon the number of IVs. In your case it seems that you have IV (tone voice with two levels) and DV (purchase intention) and schwartz culture value of power as moderator. In your case one way ANCOVA is suitable (Schwartz culture value of power can be taken as covariance).
Hi Isabel
If you use SPSS you can input your variables in the "General Linear Model" - "Univariate" menu. You put 'Purchase intention" as DV (supposing it is a continuous variable) , "Tone voice" as IV (fixed factor, meaning dichotomous variable) and "Schwartz culture value" (SCV) in the covariance box. In the "option" button, you can ask for "estimated marginal means" for the two groups (Tone voice") which gives you means for the groups "controlled" for "SCV"
Best
John-Kåre
Ps. That procedure is based on ANOVA, see for ex. Pallant SPSS Survival manual, "Two-way between-groups ANOVA" (p 274 in the 5th ed.)
jk
The power value however is not a control variable, but something I am wanting to measure the interaction of. Therefore it seems I should not use ANCOVA as its not covariance I am trying to uncover!
thanks all for your help!
Ok. You can create an interaction term in the procedure described above.
Jk
Hi, how to run two way anova with continous independant variable?
@john-kare vederhus...Dr..can i run three way anova using continous independent variable?
Thanks. This clears my doubt too. Ultimately, when you have datapoints in place of an equation, your continuity is already discretized. Needless to say that the range should be sufficiently large, otherwise there will be no statistical significance at all!
If I were you, I would install PROCESS and run model 1 to test for moderation with a continuous variable (choosing the Johnson-N. option for significance per each level of the moderator)
- Badges
- Science topic