Friendly fire and the Sustained Attention to Response Task: Using basic laboratory research to investigate a real-world problem

The primary aim of this thesis was to investigate whether the Sustained Attention to Response Task (SART; Robertson, Manly, Andrade, Baddeley, & Yiend, 1997), a high Go, low No-Go response task, may provide an empirical model of friendly fire accidents in some battle scenarios. A growing body of evidence suggests that rather than sustained attention failures, errors of commission in the SART are due to failures of response inhibition or response strategies. Some friendly fire accidents may also be due to failures to inhibit a motor response rather than commonly-cited factors such as mistaken identity. In recent times, soldiers often share battle spaces with foes but also allied soldiers and non-combatants (e.g., civilians). Engagements are at close range and extremely fast-paced, and therefore it is imperative to fire weapons not only accurately, but quickly too. These conditions share important characteristics with the SART and as such some modern battle environments may give rise to the same sort of behaviour seen in the SART. That is, people have difficulty withholding responses to rare No-Go (or friendly) stimuli in situations where there are a high proportion of Go (or foe) stimuli. Six experiments were conducted to explore whether failures of response inhibition may contribute to some friendly fire accidents. Firstly, relatively applied paradigms that incorporated characteristics of the SART into simulated battle scenarios were conducted. The same behaviours that are commonly seen in the SART, namely speeded responses to Go stimuli (foes) and frequent failures to withhold to No-Go stimuli (friends), were evident in the battle scenarios. Following this a more basic experimental approach was taken to examine the underlying mechanisms of performance, using several modified computer-based SARTs. This also presented the opportunity to explore how additional factors that might be relevant to combat might affect response inhibition. Firstly, as the proportion of foes in a battle scenario surpasses two-thirds, the likelihood of response inhibition failures (e.g., friendly fire) appears to increase markedly. Secondly, decreasing the physical or manual requirements to execute responses—such as automated weapons systems often do—may exacerbate failures of inhibition. Thirdly, stress or anxiety, a common response during combat, appears to further impair an individual’s ability to withhold responses. The current thesis suggests that the likelihood of response inhibition failures (e.g., friendly fire) may increase in battle scenarios where soldiers confront a small proportion of friends or non-combatants amongst a high proportion of foes. Additionally the findings provide further evidence that commission errors in SART-like tasks are primarily due to response inhibition failures or response strategies, as opposed to perceptual decoupling. Despite the SART’s artificial appearance, it has been used here to uncover new contributing factors to a deeply applied problem. This thesis also further reinforces the relevance of basic research for understanding complex real-world problems