I agree with David. The main question is whether you had 3 groups of about 90 subjects or whether you only had 1 group of 280 subjects. In other words, did all 280 subjects 3 times (repeated measures) answered the DV, or were there 3 groups who only answered 1 DV. In the latter case you had a between subjects/factorial design, in the former case you had a within subjects/repeated measures design.

Did you make repeat measurements on any of your subjects? If you did you need to take that relation in to account. If the answer is yes you need a repeated meas. design of some sort. Otherwise you need a factorial structure of some sort. Best, David Booth.

I agree with David. The main question is whether you had 3 groups of about 90 subjects or whether you only had 1 group of 280 subjects. In other words, did all 280 subjects 3 times (repeated measures) answered the DV, or were there 3 groups who only answered 1 DV. In the latter case you had a between subjects/factorial design, in the former case you had a within subjects/repeated measures design.

Classification of the data obtained is the first step of analysis.

Whether Factorial ANOVA or Repeated-Measures ANOVA or other Methods of Analysis will be applicable depends on the type of classification of data and the characteristics of the data.

Jos Feys , thanks for your response. I only had 1 group of 280 respondents . The respondents were given questionnaire to answer where the questionnaire measures 3 IVs. In this case, can I use Factorial ANOVA to indentify main and interaction effect?

Dhritikesh Chakrabarty , thanks for your response. Could you please explain on the type of data classification?

David Eugene Booth , thanks for your response. I don't make any repeated measurement. In my survey, I had a group of 280 respondents where the respondents were given a questionnaire to response. The questionnaire consists of 90 items to measure 3 IVs. My focus is to identify main and interaction effect between 3 IVs and 1 DV. In this case, can I use Factorial ANOVA?

Dear Kavitha Selvaraja,

I have understood that you have collected data from 280 respondents.

How do you classify these data: whether in one way table or in two way table or multi way table. This classification is to be clearly mentioned. Also, whether there is single observation per cell or whether there are more than one observation per cell, these are to be cleared. Only then it may be possible to search for the method of analysis of data.

I believe that factorial ANOVA should work in the situation you described. Please continue to ask if you have more questions. Best, David Booth

If you have only one group of respondents, factorial ANOVA can give you the main effect .

David Eugene Booth has mentioned that factorial ANOVA should work in the situation you described. However, I have failed to understand how factorial ANOVA can give the main effect and the interaction effect in this case.

An interesting scientific question.

Well we don't know how the 90 items will be dealt with. As long as there are a positive number of degrees of freedom for variance the interactions should be of the form IVa*IVb. I probably would not consider the 3 way interaction unless that is an important hypothesis. I assume there is a plan for some how getting the DV from the items. Hope I understand it. Best, David Booth

David Eugene Booth . Thanks a lot for your explanation. Based on my understanding, factorial anova works best to identify the main and interaction effect. Hence, I proceed to run the analysis. Please find the output table (as attached) for your reference. However, there is no reading for the interaction between 3 variables. What does it mean? Kindly advise.