How to empirically study critical success factors?
Critical Success Factors (CSF) are contributing factors for achieving certain objective(s). E.g. CSF can be antecedents for certain event(s) to happen or can be mediator or moderator factors contributing to certain outcome(s).
CSF can be studied depending on how a researcher try to operationalize the CSF e.g. adopt / adapt from literature review or develop a new survey questionnaire in a correlational research, or conducting an experimental research to demonstrate the CSF is taking effect in an experimental group vs control group etc.
My research interest is also on CSF, particularly in the IS/IT project environment. Specifically, the paper below proposed an alternative method to determine CSF based on the review of the previous literature on the topic. Although, it is a simple method, it may give you a preview.
Article An Alternative Method for Determining Critical Success Facto...
I respectfully disagree with Dr. Kettunen - CSFs definitely *are not* strategic objectives. However, they are (or might be) *closely related* to strategic objectives. Perhaps, Dr. Kettunen wanted to emphasize this connection in his answer, in which case the wording could be clearer.
As for your question, I would suggest you to look at structural equation modeling (SEM) techniques. For example, I have used SEM methods to explore success factors (not CSFs specifically) in my Ph.D. dissertation, which you can easily find online.
CSF have been defined in several different ways in the literature and there is some disagreement in these definitions, starting with the designations. For instance, Porter (1991) calls them drivers of competitive advantage, or sources of competitive advantage. These are defined as 'Structural determinants of differences in the cost or buyer value of activities or groups of activities'. Examples include economies of scale, cumulative learning, linkages between activities of the value chain, capacity utilization, location, timing of investment, institutional factors and others. Some say these CSF or drivers are external factors, others that they constitute requirements from customers, others that they are structural caracteristics of an industry, and some argue that they are in the intersection of external and internal factors. Some argue that CSF can be introduced (created) or changed by a company, others that they cannot be changed. There are several different points of views in the literature and in my opinion little agreement.
Edit: Several statistical methods can be used to identify CSF (eg regression, factor analysis, discriminant analysis...) and non statistical methods (eg brainstorming). The problem with statistical methods is that they rely on past data, and assume the future will be a continuation of the past.
Ref.:
Porter (1991) Towards a dynamic theory of Strategy, SMJ, vol 12, 95-117.
My research interest is also on CSF, particularly in the IS/IT in bussiness and public organizations. You can read about the methodology in my papers, e.g.
I hope my research will be interesting for you :). In my new book CSFs for ICT adoption in public organizations, enterprises, and households wer examined. You ca view: