Skill-based differences in option generation in a complex task: a verbal protocol analysis

These three exemplar video scenarios were the three most discriminating trials.

To provide some indication of whether the decision-making process resulted in a physical response (i.e., from reaching for their weapon to aiming/shooting) that was ballistic in nature and/or whether this response was cognitively mediated (i.e., participants spent additional time that might suggest they used new information as it unfolded while assessing and intervening), we also measured the mean time taken to place their hand on their weapon, unholster, and aim it at the perpetrator in the three discriminating trials. These actions were highly correlated (r = .49–.74, P = .008 to .000). Consequently, we analyzed this data using a one-way MANOVA, with skill as a between participant variable. The multivariate skill effect was significant, λ = .612, F (3, 24) = 5.08, P = .007, η _p ²  = .39. The univariate analysis indicated that, although no group differences were observed in time taken to place their hand on their weapon (M diff = 0.87 s), F (1, 26) = 3.23, P = .084, η _p ²  = .11, skilled participants unholstered (M diff = 1.50 s), F (1, 26) = 10.74, P = .003, η _p ²  = .29, and aimed (M diff = 2.07 s), F (1, 26) = 12.57, P = .002, η _p ²  = .33, their weapon earlier than the less-skilled participants. Although the skilled participants completed this action sequence (incl. shoot), on the average, more quickly than less-skilled participants (M = 5.88 s and 7.72 s, respectively), the time taken to complete this sequence was considerably greater than might be expected if participants were executing it in a ballistic manner. The data suggest that all participants spent time, over and above that required to perform the psychomotor skill of shooting “from the hip”, to consciously mediate their response by engaging in an option generation process during both assessment and intervention.