RESUMEN
Advice stemming from sources with errors that are dependent on each other is usually less accurate than advice provided by sources with independent errors, while simultaneously exhibiting greater consensus. We investigate whether or not individuals express a preference for advice with dependent errors by choosing it over advice with independent errors and by weighting it more strongly. We test for this preference both in a situation where error interdependence does not negatively affect advice accuracy as well as in a situation where advice with dependent errors is less accurate than advice with independent errors. In a series of six studies, we show that, when being given the opportunity to choose between the two types of advice, participants only prefer advice with dependent errors if this is not detrimental for accuracy. However, when being sequentially provided with both types of advice, they generally weight advice with dependent errors more than advice with independent errors, even if the latter is more accurate. This effect is mainly driven by the fact that advice with dependent errors exhibits greater consensus, leading participants' initial estimates to lie outside the range of the advisors' judgments more frequently. (PsycInfo Database Record (c) 2022 APA, all rights reserved).
Asunto(s)
Juicio , HumanosRESUMEN
Escalating commitment describes the phenomenon that decision makers may become stuck in losing courses of action, throwing good money after bad. In a seminal study, testing interventions against escalating commitment, Simonson and Staw (1992) found that holding decision makers accountable for the decision process (i.e., the decision strategies they use) decreases escalating commitment, whereas accountability for the decision outcomes tends to increase it. The initial aim of our study was to extend the original findings by testing for interactive effects of both types of accountability. However, as we did not replicate the original effects in a first experiment, in spite of the fact that our materials and our procedure resembled the original study as closely as possible, we conducted a second experiment with an even stronger accountability manipulation as compared to the original study, and with an increased sample size. Once again, no effects of accountability were found. Taken together, the results of these two experiments question the robustness of the original findings. (PsycInfo Database Record (c) 2021 APA, all rights reserved).