Links and Levels: Minimizing Problems in Cross-Functional Research Using the Principles of Data Modeling

S ustainable competitive advantage in an organization arises from distinctive competencies that reside deep within the organization. Capabilities, or the way things work together, are intangible sources of distinctive competencies that are not easily imitated. Cross-functional capabilities within an organization are perfect candidates for such distinctive competen-cies because they are both hard to observe and difficult to imitate. Operations is a function that interfaces with most other areas within the organization. Therefore, it is easy to conceptualize much of the interest in cross-functional research having operations as one side of a dyad or triad. For example, the manufacturing marketing interface has been an area of academic research for over 30 years (Lawrence & Lorsch, 1967). Although much of the work on the manufacturing and marketing interface was precipitated by marketing (e. has opened the door for significantly more cross-functional research centered on POM. However, while the importance of cross-functional capabilities is recognized in industry, cross-functional design is less common in POM research. In fact, cross-functional research carries such unique problems and challenges that we may wonder if these complications have stunted the advancement of research in this area. One of the challenges is that of the multiple informant design of research projects. In this design, there may be a dyad, triad, or even a network of respondents of interest as well as multiple constructs of interest. Multiple informant methodology has been used in a number of different Marketers and organizational behavioralists have used this research design, and, while patterns of research also emerge in psychology, sociology , and anthropology, in particular, the use of multiple informant techniques in most fields is limited. Methodologically, 1981) are topics commanding past interest. However, a search of the literature doesn't give us any tips as to how to manage multiple informant data collection or analysis. What is the difference between the more commonly used single and the less-used multiple informant design? The answer may lie in the difference between a " flat file " and a " database. " A flat file is a term used in information systems that refers to a one-dimensional set of data elements that all describe a single unit. A database is a shared collection of files and associations between the files. A more complex and robust answer is demanded by a cross-functional design. An example of one way to approach the challenges of multiple informant issues is data …