Selecting-the-Best vs. Eliminating-the-Worst: An Experimental Investigation of Idea Evaluation Processes Under Cognitive Bias Conditions

The conventional wisdom in idea selection literature typically assumes that selecting-the-best ideas and eliminating-the-worst ideas represent the two sides of the same coin. In other words, selecting-the-best ideas from a pool of ideas should be equivalent to eliminating-the-worst ones until only the best remain. However, our explorative experimental investigation regarding the accuracy of these two idea evaluation processes indicates major differences. Specifically, our results suggest that the elimination process outperforms the selection process in terms of the probability of selecting the highest quality innovation ideas. Our text analysis further reveals that when participants are asked to do the selection or elimination tasks, their cognitive perception of each idea tends to focus on different aspects of the ideas, namely, the positive (pros) vs. negative (cons) sides of the same idea. We use a 2 × 2 experimental design by priming the participants with pros and cons information in selecting-the-best and eliminating-the-worst scenarios. Surprisingly, we find that with pros, the selection process outperforms the elimination process, whereas with cons, the efficacies of the two idea evaluation processes are equivalent. Additionally, we find that the efficacy of the selection process does not change whether the participant has pros or cons, yet the efficacy of the elimination process is significantly improved with cons compared to with pros. Based on analysis of the experimental data, we present and test an explanatory model in which the evaluation accuracy, measured in terms of the percentage of matches, is influenced by factors, such as the evaluation process, response duration, and the moderating effect of cognitive biases.