Close Home Forum Sign up / Log in

Methodology driving me mad - need help

M

Hi,

I am struggling to figure out what to do with all the data my survey is collecting as part of my MBA dissertation and really need some advice as I have been going round in circles the last couple of weeks. The overall research objective is to examine whether there are opportunities for local food producers and retailers to embrace Web 2.0 (including social media) marketing techniques to market their key points of difference and overcome barriers to satisfying consumers increasing desires to purchase local foods.

I therefore thought my first research objective should be to check whether consumers with desire to purchase local foods have online channel preference as there would be no point local food producers and retailers using the online channel for marketing if it won’t reach those with desire for local foods.

But how do I test this? The two variables (desire for local food and online channel preference) are unrelated so I figure that rules out all the tests of difference and association. Is it enough to just use descriptive statistics and say x% of consumers with a desire to purchase local food would be willing to use the online channel, or is that not "academic enough"? Or can anyone think of a better way of analysing the data in this instance?

Any guidance or advice would be much appreciated.

S

It sounds like you're trying to predict something so I'd say regression but even though you're writing clear English I can't understand it, sorry, this isn't my area at all, just trying to help so you might need to change info into stats talk? I'm guessing desire for local food is categorical as either they have or they dont. How did you measure online channel preference?

S

I could be totally misunderstanding, but could you use crosstabs in SPSS this instance? For example, crosstabulating to find the number of people who had a desire for local food and also have online channel preference. You can then also perform post-hoc tests of significance (selection of which depends on how many categories you have etc).

So that way you're identifying your 'core group' of consumers, if you like, who have both the the desire for local food and online channel preference.

This is part of what I am doing in my thesis (another topic)...it seemed to be ok for my supervisor.

Does that make sense?

Would be great if someone else could help out and chime in as to whether my idea would be right or not in this case :)

M

I am measuring desire for local food using a multiple-item likert construct question which will produce continuous data, although I might group into people with low, medium and high desire to form categorical data (still undecided). I am measuring online channel preference in a similar way using a construct of items that I found in the literature which also produces a continuous variable.

I’ll need to read up on crosstabs tonight as I have never used it before, but from what you describe it sounds like it could work for me. You describe my research objective better than I do as that’s exactly what I am trying to do here, identify the consumers who have both the desire for local food and online channel preference before I then test which of the different online marketing techniques those consumers will engage with. In essence, I am eliminating those respondents who have no desire and won’t use the online channel from further analysis.

Sparkles – out of interest, what is your topic?

F

======= Date Modified 28 Nov 2011 09:40:56 =======
============= Edited by a Moderator =============
--SPAM--

20428