The negative feedback dysregulation effect: losses of motor control in response to negative feedback

ABSTRACT

Negative feedback has paradoxical features to it. This form of feedback can have informational value under some circumstances, but it can also threaten the ego, potentially upsetting behaviour as a result. To investigate possible consequences of the latter type, two experiments (total N = 159) presented positive or negative feedback within a sequence-prediction task that could not be solved. Following feedback, participants had to control their behaviours as effectively as possible in a motor control task. Relative to positive feedback, negative feedback undermined control in a manner suggesting emotional upset (Experiment 1). These reactions lasted for at least three seconds and were especially pronounced among people reporting that they typically lose control in the context of their negative emotions (Experiment 2). The findings document a novel form of behavioural dysregulation that occurs in response to negative feedback while also highlighting the utility of motor control perspectives on self-control.

KEYWORDS:

• Feedback
• behaviour
• self-control
• emotion
• reactivity

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 That the motor control task was somewhat difficult will be supported with appropriate sources of data. For now, it is worth mentioning that only 0.28% of the position samples were precisely on target in Experiment 1. In Experiment 2, this figure was 0.54%. In both experiments, then, perfect control was exceedingly rare.

2 As another way of examining feedback effects across time, we used the trend analysis procedures of SAS PROC MIXED, with difference scores to reflect the relative magnitude of negative feedback disruption for time intervals 1 (−0.39), 2 (3.44), 3 (14.94), and 4 (18.27). This analysis indicated a significant trend across time, such that negative feedback effects increased across the time intervals, t = 3.88, p < .001. The linear effects of time were also significant in a second analysis that dropped interval 1, t = 2.93, p = .004.

3 These observations about time course were reinforced by trend analyses. The disruptive effects of negative feedback, relative to positive feedback, did not systematically increase across time intervals 1 (0.01), 2 (3.44), 3 (6.89), 4 (5.76), 5 (5.17), and 6 (4.44), t = 1.27, p = .207. They also did not systematically increase, nor decrease, in an analysis that contrasted intervals 2–6, t = 0.07, p = .946.