Encouraging Consensus‐Challenging Research in Universities*

abstract Drawing from self‐efficacy theory and transcriptions of in‐depth interviews, we construct a conjoint experiment that we then administer to 54 tenure‐tracked assistant professors from two Research‐I universities in the United States. Findings from their 1728 nested decisions show that the administrative effectiveness of outcome expectations and time constraints in encouraging highly uncertain, consensus‐challenging research depends on the research self‐efficacy of scholars. As expected, we find that increases in anticipated credit are more effective at encouraging consensus‐challenging research when scholars perceive themselves to be highly competent in the line of research being pursued. Surprisingly, however, we also find that increases in both blame and time pressures are more discouraging of consensus‐challenging research when scholars perceive themselves to be highly competent in a research area. We conclude by discussing the findings and their implications for research and practice.