|M.Sc Student||Livneh Iris|
|Subject||Strategic Help-Seeking and "Don't Know" Responses in|
|Department||Department of Industrial Engineering and Management||Supervisor||Professor Rakefet Ackerman|
When answering a challenging knowledge question, one can apply various strategies to improve reliability of the provided information. According to the metacognitive approach, people choose how to answer based on their confidence in the answers they consider. One strategy, previously studied, is the option to withhold an answer when one is not confident enough that the available candidate answer is correct. Another strategy, rarely studied, is help-seeking: the decision to ask for help when one is not certain about the correctness of an answer. The aim of the present study was to compare these two strategies. The guiding hypothesis was that there are differences in confidence levels that initiate the decision to use each of them. I hypothesized that when having one option or the other, the criterion for choosing to report “don’t know” is guided by lower confidence compared to the one adapted for help-seeking. In addition, I hypothesized that this criterion is not a clear confidence “cut-off point” for providing an answer, and compared the extent of inconsistencies between the two strategies. Inconsistency was operationalized by confidence variability and control sensitivity?the association between confidence and decisions to apply the available strategy. To examine these hypotheses two experiments were conducted by using general-knowledge questions in a large difficulty variability. Experiment 1 examined the levels of confidence that lead to utilizing each strategy (“don’t know”/help-seeking). For guiding the participants to a strategic behavior, they had a limited number of times for utilizing the strategy. As expected, participants set a higher confidence criterion for help-seeking than for reporting "don't know" with more inconsistencies. That is, the confidence range was wider and control sensitivity was weaker for help-seeking than for “don’t know”. Experiment 2 was designed to examine whether limiting strategy utilization created a list-level consideration that affected the pattern of strategy utilization throughout the course of answering the set of questions in the first experiment. In this experiment the option to report “don’t know” or to request for help was subjected to an incentive structure at the item level. Furthermore, two different incentive structures were examined - higher punishment than reward and a balanced structure, where punishment and reward were the same. Experiment 2 replicated the results of Experiment 1 and no differences were found between the two incentive structures. This study provides a first sight hole into the metacognitive nature of help-seeking when comparing it to withholding an answer. It appears that when people seek help they are less decisive, but more confident in their answers compared to those who acknowledge that they do not know.