Prediction and Control as Determinants of Behavioural Uncertainty: Effects on Task Performance and Heart Rate Reactivity.

SARAH R. BAKER [1]

DAVID STEPHENSON [2]

Abstract--Control or control-belief is often viewed as being directly instrumental in facilitating coping mechanisms in aversive situations, and yet the empirical evidence for the beneficial effects of control is inconclusive. In this study we investigated the role of predictability in determining the effects of perceived control during an aversive reaction time task. Fifty-six subjects were allocated to one of four groups; predictable-control, predictable-no control, unpredictable-control, unpredictable-no control. In the predictable conditions, subjects could temporally predict the occurrence of an aversive noise. In the perceived control conditions, duration of the aversive tone was contingent on subject's performance. All subjects were matched in terms of the nature of the task and in the number and time of receipt of both the warning signal and noise. Heart rate reactivity and two performance parameters were measured, reaction time and performance increase. Both predictability and control-belief led to a reduction in heart rate reactivity, although they appeared to function independently and at different points in the sequence of events. That is, predictability or perceived control was sufficient to mitigate the effects of an aversive situation. Neither perception of control or predictability led to better task performance. These results are discussed in terms of behavioural uncertainty explanations.

Introduction

THE EXPERIENCE OF control or the belief in control over one's environment has been implicated as a key factor in determining cardiovascular responses to aversive situations (Steptoe and Appels, 1989). However, experimental support for the role of control remains inconclusive. Several studies have shown that situations involving instrumental control lead to lowered cardiovascular responses (e.g. Breier et at., 1987; Carr and Wilde, 1988; Corah and Boffa, 1970; Geer and Maisel, 1972; Peters et al., 1998), whilst others report higher responsivity with such control (e.g. Bongard, 1995; Bongard and Hodapp, 1997; Bongard et at., 1994; Hodapp et al., 1990. 1992).

It has been suggested that the disparity in the effects of control is due to the (mental) effort required to exercise control (Bongard and Hodapp, 1997; Light, 1981). Yet, of these studies, few have included any assessment of effort such as task performance. Where performance parameters have been included, some studies have reported better task performance with control (Contrada et al., 1982; Lovallo et al., 1985), whilst others obtained no performance differences between control and no-control situations (Bongard and Hodapp, 1997; Bongard et al., 1994; Hodapp et al., 1990, 1992). From these data alone, it cannot be concluded that the differential effects of control are mediated solely by effort.

One factor, which may be of importance in understanding the effects of control, is predictability. It has been suggested that predictability is a building block in the perception of control (Fisher, 1984). That is, in order to exercise effective control over an event, to attenuate, reverse, or remove the event, one must be able to predict the occurrence of that event. As such, both predictability and control are directly instrumental in determining cardiovascular responses in aversive situations. Consequently, in order to study the effects of control, the role of predictability in determining instrumental action needs to be examined. Yet, most studies investigating instrumental control have either ignored the effects of predictability, or have been confounded by changes in predictability (see Baker, 1994; Miller, 1979a; Nickels et al., 1992). This is because predictability facilitates control by providing information concerning the timing of coping responses. As such, both control and predictability may be d irectly instrumental in determining cardiovascular responses in aversive situations, and any effects cannot be attributed to control per se.