Upward communication in industrial organizations : a laboratory experiment

The purpose of this study was to test empirically variables contributing to the frequency and accuracy of information transmitted up the chain-of-command by employees in industrial organizations. Hypotheses were generated from organizational communication theory which predicted the effects of manipulated job situations (problem and routine), trust in the immediate superior, and personal ambition on the following dependent variables: (l) projected frequency with which subordinates transmitted messages upward and (2) distortion of those messages, which could be achieved by withholding information altogether, by puffing or exaggerating details, and by sieving or selectively filtering part of the information.

A pretest survey was conducted at the Aluminum Company of America (ALCOA), Tennessee Operations in Alcoa, Tennessee to establish valid experimental and control treatments. Following the survey, a posttest-only laboratory experiment was executed. Data were gathered from 40 randomly selected exempt ALcoA employees from the Badin Works plant in Badin, North Carolina. Subjects were randomly assigned to experimental or control groups. The former was exposed to a problem work situation identified by the pretest, and the latter received a routine, everyday situation revealed by the pretest as being a neutral job occurrence. Subjects were asked how often (on a five-interval scale) they would convey news about the situation to higher echelons. They then communicated to their bosses pertinent situational details via memoranda which were later content analyzed for distortions. Finally, the participants completed scales measuring their ambition and trust in their superiors.

Major findings were that job situation, trust, and ambition did not adequately explain the projected frequency with which employees communicated upward; however, they did account for distortion. Secondly, in neutral job situations, trust and ambition were negatively related to projected frequency of communication. Thirdly, of all distortion types, withholdings were most prevalent. Next, in problem job situations, trust and ambition were negatively related to withhold distortions. Lastly, in neutral job situations, ambition was negatively related to puff distortions. Multiple regression analyses were conducted to obtain results.

Industrial managers need accurate data if they are to make intelligent, rational decisions. Much of their information is supplied by underlings who are closer to job situations. This research demonstrated that distortion is a predictable part of a subordinate's behavior. Future investigations could test alternative populations, work situations, and communication networks.