Sutton, R. I. and Staw, B. M. (1995) What theory is not. ASQ 40:371-384.

---

Authors routinely use references, data, variables, diagrams, and hypotheses in lieu of good theory. Journals should be more receptive to papers that thest part than all of a theory and use illustrative than definitive data.

There is lack of agreement whether a model and a theory can be distinguished, whether a typology is a theory, and the value of "interestingness" on theory, and whether falsyfiability is a pre-requisite. Scholars are forced to make tradeoffs etween generality, simplicity, and accuracy.

Parts of an Article That are not theory

1. References are not theory.

Lots of references to existing theories does not create a new theory. Sometimes references are a smoke screen or are "throw-away" references. Authors need to explain which concepts and arguments are adopted from cited sources and how they are linked to the developed theory.

2. Data are not theory
Data describe which empirical patterns were observed and theory explains why empirical patterns were observed. Data provides support but does not constitute a theory. Those who use qualitative data must develop causal arguments to explain why findings are observed if they want to include theory.

3. Lists of Variables or Constructs are Not Theory
Theory is not conceptual definitions. Lists of variables that cover all possible determinants help explain but is not theory by themselves. Comparative tests of variables is not a comparative test of theory. They key issue is again WHY certain variables are more important.

4. Diagrams are Not Theory

Diagrams can help explain how a phenomenon is created, but they again don't explain why. Good theory is often representational and verbal. With a strong theory one can discern when a major hypothesis is most or least likely to hold.

5. Hypothesis (or predictions) are not theory
A theoretical model is not simply a statement of hypotheses. Hypotheses are statements about what is expected to occur, not why it is expected to occur. Predictions without logic are not theory. Stong theory papers have both simplicity and connectedness.

Identifying Strong Theory
Theory answers why. It's about connections among phenomenon. It delves into underlying processes. It is laced with t set of convincing and logically interconnected arguments. Weick says a good theory explains, predicts, and delights.

The Case Against Theory
Van Maanen believes there is too much mediocre theory and there should be a 10 year moratorium while we understand organizations better through observation. Others think we should focus more on accumulating empirical findings for meta-analysis.

Most researchers are trained in experimentation, not in theory building. It's difficult to successfully do both and still get published. Often theory is crafted around the data. Thus the craft of manuscript writing becomes an art of fitting concepts and arguments around what has been measured and discovered.

Some Recommendations
Right now it's easier to agree around a strong empirical paper with no theory than one with a weak test of a new theoretical idea. Unfortunately, papers chosen for revision seem to be those with acceptable methods and undeveloped theory.

The author's recommendation is to rebalance the selection process between theory and method. Journals should publish papers that are stronger in theory than method. It's sort of been happening for qualitative research but not quantitative research. Seldom are ethnographic descriptions published when they are not also a source of new concepts or ideas. In qualitative studies often theory is emphasized too much.