I have a revised theoretical model adapted for qualitative research and at this stage i would like to use it for a quantitative approach.
A grading mechanism to quantify the measurement of each factor being assessed.
And - A weighting factor to describe the inter-relationships between factors, and show the relative
contribution: the weighting factor should be applied to the stages of the adoption process, as
the relative contribution may be different from one stage to another.
Any guidance on how i can grade the different factors on the theoretical model?
I am not sure what you mean by a "grading mechanism" for each factor. Are you familiar with factor analysis as a statistical technique? It is designed to assign individual items to underlying factors (i.e., it groups together items that are highly correlated and thus measure the same thing).
To do this kind of analysis you would need to survey a large sample, as a follow-up to your original qualitative study.
Hi David, thanks for your reply. The previous researcher did a qualitative study and now i am intending to do a Survey (quantitative). Let me rephrase my question.
The original study followed a qualitative study. The researcher identified a number of factors which had some influence on a phenomena. So now i would like to somehow do a grading or weight on each of the factors to quantify the measurement of each factor being assessed. Let me know if i am making sense here.
The usual approach for the kind of design you are describing is to develop a set of items for each of your separate factors. For example, if you have four factors, you should have a minimum of three items per factor, and thus twelve items total. Personally, I would aim for four items per factor, in case one of them does not work, then you would still have three available.
These items would typically be constructed using a LIkert-scoring format, so that they can later be summed into a scale. Before doing that, you would consider statistical analyses such as Cronbach's alpha and factor analysis.
If you are not familiar with creating survey instruments, I would suggest looking at Fowler's book, Survey Research Methods.
David once again thanks for replying. Just for further clarity ( What do you mean by saying developing a set of items for each of the separate factors?) can this be questions in a survey related to a Factor?
The linkert-scoring is well understood (thanks) _ I have began reading through both Cronbach and Factor Analysis -
Any scale (or factor) that you create will need to have a number of items associated with it in order for it function effectively. So, the recommended process is to identify the core topics that you need to measure, and then attempt to create a multi-item scale for each. When you have a prior theoretical framework, that could provide the content for each scale/factor. Alternatively, you could do preliminary qualitative research to learn how your respondents think about and talk about the issues that your want to measure.
Thanks a lot David, this is quite clear. I will certainly contact you if i get stuck? may i contact you further? [email protected]
David, now i would like to find out if i wanted to design a conceptual model? based on set of constructs identified in literature to have a positive influence on a certain IT adoption ( this was done qualitatively) and now i intend to do it quantitatively - and i intend on using the same questions used in the qualitative interviews but in a more quantitative approach (likert scale) ... i mean have only a year to do this research. Would i need to do a validity check on the constructs first?
I would recommend checking the effectiveness items of the items beforehand, and one common method for doing so is called "cognitive interviewing." For a review of those procedures, see Beatty & Willis (2007) in Public Opinion Quarterly.
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Biochemical Techniques
- Analytical Chemistry
- pH