Hello Everyone,

I am pondering upon which statistical test to use for my project.
I would appreciate any input/advice from anyone.

I am using a between subjects design, i.e.: different participants would be assigned to different conditions in the experiment. I have for example, four scenarios and I am looking at collecting 30 Males and 30 Female participants for each of the scenarios.
For example, the Dependent Variables for the study are the following:

- Attitudes to perpetrator dangerousness
- Attitudes to punishment
- Attitudes to treatment or rehabilitation.
- Fear
- External blame attributions for perpetrator behaviour

And I have three Independent Variables each with two levels. For example:

Gender of the perpetrator : Male/Female
Gender of the participant: Male /Female
Mental Health status of the perpetrator with two variables.

I believe that female perpetrators would elicit more negative perceptions and feelings than male perpetrators. I am also looking at overall gender differences with regards to participants perceptions/blame attribution, etc.
For the other hypotheses, there is an absence of research in the area, therefore, those would be only exploratory in nature.

Now the question is whether I should use a 2x2x2 ANOVA or a Regression analysis?
I know that the ANOVA would tell me if there is a difference but there are more tests to carry out in order to find out where that difference lies. . . While the regression estimates the relationships between variables. . .

Which statistical test is better or more appropriate for the following example?
I would appreciate any input or comment on this issue.

Many thanks in advance.

Methys