The nature of performance effects deserves some special mention. It is clear that we have no quantitative measure of changes in output, such as are available in many studies in industry. When improvement in performance was reported, productivity was not merely at standard levels but considerably above and beyond them. The effects reported were very real and in no sense inconsequential. Even with ideal measures of effects, a positive relationship between morale and productivity would come from that part of the morale measure that depended on satisfiers. If the proportion of positive attitude-effect relationships were low, the overall result could easily approach randomness. The kinds of reports that were coded as mental-health effects were of many varieties. Job attitudes are a powerful force and are functionally related to the productivity, stability, and adjustment of the industrial working force.