Managerial selection decision models: examination of configural cue processing

Michael A. Hitt Department of Management Texas AM Wright, 1969) but have re- ceived much criticism (Arvey & Campion, 1982; Schmitt, 1976). Although the use of job-irrelevant variables in the selec- tion of employees may be illegal if it causes adverse impact, little research has systematically evaluated the relative influence of such characteristics on actual managerial selection decisions. Typically, studies have used student subjects and paper appli- cants for an entry-level position, have examined only one or a few variables, and have provided little or no job information, thereby raising serious questions about their external validity (Dougherty, Ebert, & Callender, 1986; Gorman, Clover, & Doh- erty, 1978; Wexley & Pulakos, 1982). The purpose of this research was to examine determinants of applicant favorability and recommended starting salary in a more realistic, socially rich environment by using managers experienced in making selection decisions. This study used more cues and evaluated more applicants than used in previous research. The task was therefore more realistic and allowed for examination of configural cue processing by managers in the selection decision process. In this study we considered both job-relevant and job-irrele- vant variables. We also provided both written information (an application form) and visual data (video presentation by the applicant), allowing the managers to see and hear the applicants We gratefully acknowledge Gerald Ferris, Eugene Stone, William Warde, Charles Williams, and three anonymous reviewers for com- ments on previous drafts of this article. Furthermore, we are indebted to Tim Keaveny for his suggestions and input regarding the examination of recommended starting salaries. Correspondence concerning this article should be addressed to Mi- chael A. Hitt, Department of Management, Texas AM O'Neill, 1985). Some authors have argued that these differences are largely re- lated to non-job-related variables (e.g., age, race, and sex), whereas others have argued that they are based primarily on differences in education and experience. As a result, questions remain regarding the reasons for differentials in compensation among groups. 53 Information Processing, Cognitive Bias, and Judgment Management and cognitive psychology literatures suggest that an underlying cognitive model governs the manner by which persons integrate items of information into a single judg- ment in making decisions (Hitt & Middlemist, 1979; MacCrim- mon & Taylor, 1976; March Simon, 1958). However, Slovic Lichtenstein (1971), Tversky Kahneman (1974), Slovic, Fischhoff, and Lichtenstein (1977) noted that decision makers have a difficult time weighting and combining informa- tion viewed as relevant to their decisions. When dealing with complex tasks involving uncertainty, complexity, or ill-struc- tured problems, such as selection decisions, people try to sim- plify the decision process by developing and relying on heuris- tics (Payne, 1976; Prahalad & Bettis, 1983). This introduces cognitive bias and can lead to systematic errors (Duhaime & Schwenk, 1985; Tversky & Kahneman, 1974). Nisbett and Ross (1980) suggested that few decision makers search for or analyti-