my first empirical study - which test?

[Nich95]

Hi everyone, I’m preparing my first empirical research and I have some troubles with my study design because I’m not very familiar with data analysis and statistic.

I want to study the effect of the country of origin of a product on the willingness to buy that product. In a survey I show 3 times the same product but each time it comes from a different country and each time willingness to buy is measured on a 7-point scale. What test could I use to measure the effect of the country on the willingness to buy? Could I use analysis of variance even if the 3 groups are always composed by the same respondents?

Then I have to integrate some moderating variables, like product knowledge. How do I study the effect of the moderating variables on the relation between country of origin and willingness to buy? For example: scholars say consumers with low product knowledge are more influenced by country of origin. How do I proof/disproof it? I thought about a regression analysis between product knowledge and willingness to buy, but I don’t see how this could tell me about the influence of country of origin.

If I sound confused it’s because I am. I’m trying to understand which analysis I should run in order to do a better study design.

Thank you in advance to everyone will be so patient to try to help me.

 

[Karabiner]

In a survey I show 3 times the same product but each time it comes from a different country and each time willingness to buy is measured on a 7-point scale.

Won't your subjects not find it strange that the same product is
presented with 3 different origins?

At least, the sequence of the 3 origins should be balanced as
to decrease order effects.

Could I use analysis of variance even if the 3 groups are always composed by the same respondents?

If you can justify to treat your dependent variable as
interval scaled, then repeated-measures analysis of
variance could be used. But you did not state how many
subjects will be included. For small sample sizes,
or if you consider your dependent variable as oridnal
scaled, the Friedman test could be used instead.

Then I have to integrate some moderating variables, like product knowledge. How do I study the effect of the moderating variables on the relation between country of origin and willingness to buy?

Well, this will indeed only work if you consider
your dependent variable als interval scaled.
Moderator effects can be represented as interactions.
In the present case, as the interaction between the
repeated-measures factor and the moderator.
This will show you whether the extent of the "origin"
effect is different between levels of your moderator.

With kind regards

Karabiner