Evaluation As a Two-Way Process

Evaluation tends to be seen as a one-way process; the evaluator evaluates the evaluated by its features. Asymmetry between the evaluator and the evaluated is often stressed, such as in research on processes of signaling/screening or stigma. What is overlooked within this framework is that the evaluated counter-evaluates the evaluator by the way (s)he evaluates him/her. This tendency brings shortcomings such as the underestimation of the agency of the evaluated. Thus, a theory of two-way evaluation is necessary in both descriptive and normative terms.

This presentation aims to explain the actual relationship between actors who were formerly called the evaluator and the evaluated, by drawing on a case study of job-matching in Japan.

As for job matching, economic and sociological theories were based on the framework of one-way evaluation. Human capital theory and signaling/screening theory both seek to explain how firms evaluate applicants, and discuss the effectiveness of the methods of evaluation, while the choice of firms by applicants are neglected. Even research focusing on the lived reality of job searchers share this framework. For example, Sharone (2014) describes how job searchers in the United States (where it is assumed one’s personality is evaluated) who are rejected internalize the evaluation and experience self-blame.

Contrary to these descriptions, data taken from semi-structured interviews on job searchers’ choice of firms revealed that job searchers are evaluating firms by the way the firms evaluate applicants. The job searchers’ preference of firms were not fixed. They were judging from the questions the interviewer asked and the verbal/nonverbal responses to the applicants’ answers the firms’ competitiveness and culture. Also, despite the commonality of the evaluation processes, self-blame was not the main response to rejection.

Drawing on these examples, this presentation seeks to construct the framework of two-way evaluation, and discuss the advantages it brings.