I am working to find the reasons behind formation of industrial clusters In a city where there were no any industry before. I design a questionnaire to know the perception of workers and owners of the firms to know the reasons of this clustering. I ask the questions about, tax holiday, incentives for new investment, special industrial park in the city.
Tthe problem for me is, how to know the actual reason behind this clustering?
I ask the question on a Likert Scale from 1-7
If I understand you correctly, in this case you are not able to use regression, because your sample consists only of companies that actually are in that cluster. Perhaps the factor analysis will be more suitable for you if you have more variables and you need to focus on certain factors “behind” them. Such as for example taxes, labour costs, permission cost..., could be very likely aggregated into cost factor. However, you still don’t know much about the effect of certain factors on their decision. If you have the sample with firms in the cluster as well as those out of the cluster you can use logit or probit regression to test potential causalities and quantify the effect of certain factor
dear Jan,
Thanks for your reply, which mean a lot to me
I have asked question about following variables
Law and Order Situation
Herding Effect
Land
Infrastructure
Labor
Capital
Geographical Characteristic of City
Business Culture
Taxation
Incentives By the Government
it is an emerging cluster, and many automobile firms gather here in few years. i want to know the reasons why they are gathering here. what is the most influential factor(s) which are attracting the investors to come in this city.
No, you cannot do regression without a dependent variable.
It sounds like all the questions you asked would be independent variables to predict the "perception of workers and owners of the firms" with regard to something (clustering?). Except it also sounds like you didn't remeasure anything related to this dependent variable, so you don't have thing to predict.
In that case, you will have use a different model and try to explain differences in the list of perceptions that you provided. This assumes that you have other "background" and demographic variables that could serve as your predictors (e.g., whether someone is a worker rather than an owner, etc.)
Are you committed to doing quantitative research, and do you have a way to generate a sufficiently large sample?
The reason I ask is because your research question seems like it would be a better fit to an exploratory, qualitative study.
@ Kelvyn Jones. I have the start date of every firm. all are startups not relocations.
@ David L. can you please guide me how to make it exploratory in qualitative study. I am not stick with questioner
A qualitative study with involve in-depth interviews with perhaps 20 or so key informants to learn more about the nature of the "cluster" -- how idd it form why did it happen here, what were the key factors that made a difference, etc. Rather than testing hypotheses, an exploratory study would be discovery oriented, and it might generate hypotheses that others could assess in later studies.
I agree with David it is probably necessary to be qualitative to begin with, but bear in mind that you want your clusters eventually to be measurable, so pick up whatever quantitative facts you can. Ask the firms (and others):
Groundings: what resources (human, natural, technology and other) and infrastructure are they using and where does it/ did it come from?
Enterprise: what are the firms, how do they interact (compete/cooperate/ignore and how do they compete - eg design, price, speed of service ....) and how do their suppliers and related and supporting businesses interact?
Markets: what's the size and nature of the local market (and how much of that is met from within the cluster) and what kind of access does the cluster have to markets further afield?
If you can get answers to questions like these, covering more than one person in each organisation if you can, to pick up diversity of views, then you can begin to form an opinion of why there 'had' to be a conglomeration or cluster at this place, and what kind of things can sustain it and make it competitive.
Good fortune!
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics
- Survey Design
- Questionnaire