Analysis of the Incidents
Once the data in a research project are gathered, they must be analyzed. (As we saw earlier, both the
collection and analysis phases of such a project are potentially problematic.) Herzberg and his
colleagues broke each interview down into almost 500 separate thought items and then classified
these items into categories based on the degree of similarity that appeared among them. Notice that
the researchers allowed the nature of the items to determine the categories to emerge from this
analysis rather than sorting them into predefined categories. Care was taken to assure that independent
researchers could agree on the classification of particular items into the various clusters that
resulted. The items were also categorized as having either long- or short-term effects on the
employees at the time.