Motivation and emotion/Book/2010/Learned helplessness

Overview
Learned helplessness theory relates to the psychological phenomenon whereby individuals exposed to uncontrollable events learn that their responses and event outcomes are independent of each other. Such learning can lead to an expectation within the individual that responses are futile which can then generalize to new situations and interfere with future learning (Alloy, Peterson, Abramson, & Seligman, 1984).

The early dog experiments


Overmier and Seligman (1967) demonstrated a surprising finding. In a series of experiments investigating escape-avoidance behaviour in dogs, Overmier and Seligman (1967) discovered that dogs who had received prior inescapable electric shocks in a Pavlovian harness were later helpless when shocked in an unharnessed situation where shock could easily be escaped. In describing the unharnessed shock conditions, Seligman writes;

However, in startling contrast to this, it was discovered that dogs pretreated with inescapable shock would initially react like the naïve dog, running about and howling, but would soon cease this behaviour and begin to passively accept the shock and its shocking fate. In this way, the dog appears to give up, helplessly accepting its fate (Hiroto, 1974). What’s more, on each successive trial, this pattern of behaviour only worsened, with the animal becoming more ingrained in its passivity. If by chance any of the dogs accidentally jumped the barrier or fell outside the shuttlebox, the animal still failed on successive trials to replicate this behaviour. So while the naïve dog’s shocks were relatively minor because it could escape them, the pretreated inescapable dog’s shocks would continue for as long as the trials did (Seligman & Maier, 1968).

Overmier & Seligman (1968) dubbed this phenomenon "learned helplessness”. Citing that learned helplessness results from receiving aversive stimuli in a situation in which any contingent responses or attempts to respond occur in the presence of the aversive stimuli, and where such responses hold no significance to eliminating or reducing the severity of the traumatic stimulus.

It was shown, however, that if pretreatment and testing were 48 hours apart then the pretreated dog did not show interference on its escape/avoidance response. But if the dog responded helplessly in the shuttlebox 24 hours after initial pretreatment, then the learning of helplessness could persist for up to another 168 hours (Seligman & Maier, 1968; Overmier & Seligman, 1968).

In an attempt to verify the study, Seligman & Maier (1967) demonstrated that the passivity of the dogs was not created by incompatible skeletal responses due to the pretreatment with the electric shocks. To demonstrate this, experimenters sedated muscular movement of the dogs during pretreatment using curare. This was to ensure during testing that the results were a cognitive function of passivity rather than a muscular-skeletal reflexive response pattern. As such, dogs treated with curare replicated the response of the initial study.

Seligman and Maier’s (1967) studies added depth to the Learning Theory of the time. Where it was once proposed that explicit contiguity between events (acquisition) and explicit non-contiguity between events (extinction) produce learning, they proposed that a third variable, independence between events also produces learning. It was believed that such learning could have effects upon behaviour different from those of the two original pairing and non-pairing dimensions, in such a way that individuals could also learn that outcomes are not dependent on contingency.

Early helplessness experiments with humans
Thornton & Jacob (1971) were the first to test the learned helplessness phenomenon on humans. The first problem they encountered, however, was the reality of the subject stressor levels. All of the experiments thus far had used near traumatic shock in the acquisition phase of their experiments, and using such a shock on humans had certain ethical complications. As such, the experimenters had to moderate the stress-set levels of the experiment to a level that was unpleasant for the individual but not painful. This, however, also had impractical implications; once a shock level for an individual has been set at an unpleasant but not painful level, the individual will become accustomed to the level, and as such the conditions become predictable, and thus manageable. By making the experience seem manageable, it makes conditions unfavourable to instil a sense of helplessness.

To combat the predictability dilemma in their experiment, Thornton & Jacob (1971) resolved to use a variable level of shock, set around the individual’s unpleasant but not painful set-point, a method which effectively induced and maintained a stressed state for the individual.

While results in these first experiments were promising, and did show that learned helplessness could be replicated in humans, they weren’t amazing statistically. The experimenters attributed the lack of highly significant results to the insufficient shock level, attesting that without traumatic shock it was hard to determine what extent learned helpless could be produced in man (Thornton & Jacob, 1971).

This research was significant however, as it represented for the first time that there could be different degrees of learned helplessness; with 72% of those individuals who received a constant level of shock trying to make an escape response; while only 35% of those individuals who received a variable amount of shock trying to make an escape response. This gives evidence that there individuals can learn variable amounts of helplessness (Thornton & Jacob, 1971).

The second experiment of learned helplessness in humans was conducted by Dweck & Reppucci (1973) using problem solving tasks in children. In these experiments, the children were first presented with unsolvable problems, and then, once they’d learned helplessness, presented with solvable problems. This study was highly interesting because not only did it replicate the learned helplessness experiment in humans, in particular children, it also demonstrated incidents of resiliency in the face of uncontrollable outcomes. What it showed more than this, however, was that humans have habitual explanatory styles, which predispose them to specific explanatory styles that they can use when they are faced with either a controllable or uncontrollable event. This experiment was also the first to use predictive measures to determine which children were less persistent in their explanatory style, and thus more likely to take less personal control for outcomes. Similarly, it was the first research to apply cross-situational paradigms to a human helplessness study, proving the sometimes situational specificity of helpless.

Following the Dweck & Repucci (1973) study, another highly significant study was published a year later by Hiroto (1974). Hiroto’s study was the first on humans to use a loud tone as an aversive stimuli rather than an electric shock, and it was done in an effort to extend the generality of learned helplessness to other stimuli.

Hiroto’s (1974) experiment came off the back of trying to amend some of the methodological flaws of the Jacobs & Thornton (1971) experiments, which had major problems with subjects knowing too much about the experiment for the non-contingency effect to take hold, and with the pretreatment having a shock that was avoidable and thus not as effective as Hiroto liked.

Hiroto’s experiments more truly established in humans the learned helplessness deficits which had been so clearly demonstrated in previous animal experiments. The details of the experiment are as follows; in pretreatment, one group received an aversive loud noise which it could escape by pressing a button, another group received inescapable noise, and a third received no pretreatment. Then, in the testing procedure, all groups received controllable noise which they could all turn off by pressing a button. As in the animal experiments, the inescapable pretreatment group sat and took the noise without responding, while the pretreatment escape, and no-pretreatment groups, readily escaped the aversive tone by pressing the button (Hiroto & Seligman, 1975).

Hiroto’s experiments were the first to accurately reflect the strong research findings of the earlier experiments with animals. Hiroto suggested that the helplessness deficits demonstrated in his study may be due to a perceived external locus of control that had been learned by the pre-treatment no-escape human subjects (Nolen-Hoeksema, Girgus, & Seligman, 1986).

The fourth helplessness study on humans saw Hiroto & Seligman (1975) team up to attempt further helplessness investigation, especially to ascertain whether helplessness could be learned across modalities. In these experiments, individuals were pretreated with either unsolvable problems or aversive noise, and then tested afterwards on solvable problems. Although the modalities differed, one being a cognitive task and the other being an aversive stimulus, they both related in that neither could be overcome.

The results were similar to those of Hiroto (1975) and Dweck & Repucci (1974); individuals pretreated with uncontrollable conditions failed to take control in test conditions once in the shuttlebox, while those pretreated with controllable conditions and no pretreatment readily escaped the aversive stimuli or, as required, solved the problems in test conditions. But the interesting thing to come from these experiments was the outcomes in the cross-modality testing. Individuals who were pretreated in one modality (either aversive tone or unsolvable problems) showed as much helplessness in testing on the opposite modality as individuals who were pretreated to be without control in the same modality (Hiroto & Seligman, 1975). While these results were good, it is also to be noted that the subjects may have perceived the different modalities as existing within the same experiment, and as such it is hard to determine how accurate the cross-modality findings were.

The reformulated helplessness theory
The original learned helplessness theory was created before helplessness experiments were performed on human subjects. Originally, human versions of the helplessness studies were designed to replicate the findings in animals, however, greater outcomes such as depressive symptoms and self-esteem loss from depression, called for a revised edition of the helplessness theory to uncover the causality of such findings (Peterson & Seligman, 1984). The new helplessness model stated that people’s explanatory style influences the causal attributions they will make for a particular uncontrollable event, and that it is this causal attribution for a particular uncontrollable event which influences their subsequent expectations of helplessness, which then, as a result, determines how generalized learned helplessness will become for the individual (Alloy, Peterson, Abramson, & Seligman, 1984).

The role of contingency
Contingency is noted as the mediating factor in learned helplessness theory. Contingency stands as the individual’s ability to act upon their environment and ultimately change the consequences of outcomes. In a learned helplessness scenario, the individual has learned that they have no contingency over events (either specifically or globally), whether this is true or not. As such, the individual has learned that their behaviour and the situational outcomes are independent. The result of such impaired contingency also somehow affects the individual’s ability to learn new response behaviour, and as such the individual needs to be forced to make the appropriate contingency time and time again, as this is the only way it can learn to become contingent again (Seligman & Maier, 1967; Maier & Seligman, 1976; & Seligman & Garber, 1980).

The role of explanatory style
Explanatory style relates to the attribution individuals place on the causes of the events if a significant degree of ambiguity can be attributed to the outcome (Alloy, Peterson, Abramson, & Seligman, 1984). If the situation is ambiguous enough, it likely that an individual will explain the outcome of the event in a way that is consistent with how they often explain ambiguous events; projecting and imposing their own habitual beliefs for the causes of outcomes (Metalsky, Abramson, Seligman, Semmel, & Peterson, 1982). If the event is not ambiguous, it often speaks for itself, but individuals may also overshadow the true nature of those causes, in order to reflect their already established explanatory style (Peterson, & Seligman, 1984).

When individuals face a specific outcome or event, they use an explanatory style to explain to themselves why the specific of causes of an outcome have occurred. In doing this, there are three attributional dichotomies which individuals can use to explain the causes of outcomes; internal-external factors, stable-unstable factors, and global-specific factors (Peterson, & Seligman, 1984). In the internal-external dichotomy, the individual is attributing the cause of the outcome to the contingency they applied to it (an internal factor), or attributing the cause of the outcome to something situational or contextual (an external factor). With the stable-unstable factor, the individual can either attribute the cause of the outcome to something pervasive and long lasting (a stable factor), or to something transient (an unstable factor). With respect to the global-specific factor, the individual can relate the cause of an outcome to something that either affects all outcomes (a global factor), or to something that affects only this outcome (a specific factor). As such, the explanations people give for specific outcomes influences their expectations regarding future outcomes, which in turn ends up influencing their reactions to outcomes, and can result in an underpinning explanatory style which perpetuates experienced helplessness and low self-esteem (Nolen-Hoeksema, Girgus, & Seligman, 1986).

According to the learned helplessness theory, the type of explanations individuals make regarding the control they influence over an outcome, determines whether their helplessness in the situation will also be marked by low self-esteem, and whether or not their helplessness in this situations will carry-over to other situations, and will continue across time (Seligman, Abramson, Semmel, & Von Baeter, 1979). If, when an individual faces a situation they cannot control, they attribute their lack of control to their own inability (an internal factor), the ever-present nature of their inability (a stable factor), and the consistency of their inability across all situations (a global factor), then that individual is likely to approach the next situation they face, regardless of how controllable it is, with the established notion that they are helpless to affect the outcome in a favourable way (Alloy, Peterson, Abramson, & Seligman, 1984).

With regards to the difference between ambiguous and non-ambiguous events, it is important to note the distinction between the causal explanation and the expectation of consequences. An individual’s explanatory style will usually predict helplessness deficits, but there can sometimes be inconsistency, such as if the circumstance is particularly non-ambiguous. In such circumstances, an individual’s habitual explanatory style may not accurately reflect how they explain a particular event’s outcomes (Peterson, & Seligman, 1984).

The effects of learned helplessness on the individual
In studies which demonstrated people’s changing explanatory style, from a more internal locus of control to a more external locus of a control (where the individual perceives less control), Dweck & Reppucci (1973) found that reactive frustration became less constructive the less control an individual perceived they had over the outcomes of an event. It was also found that debilitating anxiety reaction scores increase and facilitating anxiety reaction scores decrease as the individual’s perceptual notion of control moved further away from the self.

Many cognitive and behavioural effects are caused by learned helplessness. Hiroto and Seligman (1975) cite that due to the close nature of learned helplessness and depression, many of the same characteristics are present, such as self-devaluation, psychomotor retardation, and feelings of sadness. Many other deficits have been cited, such as passivity; poor problem solving; anxiety; hostility; (and in contrast) a lowering of aggression; a lowering of appetitive drives; a set of neurochemical deficits; and an increase in susceptibility to disease (Peterson, & Seligman, 1984; Peterson & Barrett, 1987). Generally, however, the effects of learned helplessness can be seen to have a negative effect on three broad psychological constructs; motivation, cognition, and also emotion (lowered self-esteem) (Nolen-Hoeksema, Girgus, & Seligman, 1986).

Evidence from studies on American freshmen university students, showed that those students who exhibited the learned helplessness explanatory style of attributing (citing internal, stable, and global causes) were much more at risk for poor grades in their first year of college. Similarly, even when controlling for factors like ability and depression, such attributional styles were also correlated with a lack of academic goals, as well as a failure to use academic advising services, factors which also happen to correlate strongly with good grades (Peterson & Barrett, 1987)

Another study regarding the role of helplessness and rheumatoid arthritis sufferers showed that the levels to which individuals are depressed by their condition (a condition which is renowned to imbue depression), was more reflective of the patient’s explanatory style than it was of the severity of the individual’s condition. Such that, despite the severity of the individual’s condition, it was the belief in their ability to control or cope with their disease that determined their level of depression, rather than the reality of their subjective condition (Smith, Peck, & Ward, 1990).

Learned helplessness and depression
In the reformulated model of learned helplessness, Abramson, Seligman, & Teasdale (1978) proposed that certain attributional styles may predispose individuals to have depressive reactions to certain negative life events. Results show that individuals who have a explanatory style of attributing negative outcomes to global causes show greater potential, compared to people who do not attribute to global causes, of generalizing learned helplessness to new situations (Alloy, Peterson, Abramson, & Seligman, 1984; Nolen-Hoeksema, Girgus, & Seligman, 1992). But attributional style itself does not cause depression, rather, it is the occurrence of a bad outcome, or the expected occurrence, and the individual’s perception of that event using a depressive explanatory style (attributable to internal, stable, and global factors), that causes the depression to take place (Seligman, Abramson, Semmel, & Von Baeter, 1979).

Seligman (1978) considers learned helplessness to be at least some type of subclass of depression, with the reason being that depressed subjects and non-depressed subjects made helpless show the same behavioural deficits in experiments relative to normal controls. Nolen-Hoeksema, Girgus, & Seligman (1986) hypothesise that learned helplessness may be born out of depression, such as where a traumatic life event can induce a depressive state, with the helplessness surrounding the incident (such as the death of a parent) flowing on to other aspects of the individual's life, ensuring a ripple effect and possibly producing a maladaptive explanatory style. So what started out as a genuine uncontrollable event, with an appropriate depressive response, could end up affecting the individuals coping mechanisms in such a way that, although the depression of the original event may subside, the individual may have acquired a resilient negative explanatory approach that could leave them open to depression when faced with a future negative life event.

An interesting finding to come out of learned helplessness and depression studies is that those individuals with depression more accurately judge the amount of control they have over events (Alloy, & Abramson, 1982), as opposed to non-depressed individuals who falsely attribute the amount of control they have over events, particularly in events where winning or skill is involved. This false attribution has been named as an ‘illusion of control’, and it is something that depressed individuals don’t fall privy to as they fail to see themselves and their contingencies in an optimistic light. In general, depressed individuals accurately assess their impact on events and, unlike nondepressed people, do not overestimate the limited amount of control they have over objectively controllable events that were associated with failure. It should be noted, however, that depressed individuals also underestimate the amount on control they could have on certain controllable events (Alloy, & Abramson, 1982).

Learned contingency/optimism
An important finding to come out of all learned helplessness literature is that it has been shown that prior escapable shock serves to protect an individual against any interference on any following helplessness trials, and that any learned helplessness can be alleviated by a therapeutic procedure of forced responding (Thornton, & Jacob, 1971). Similarly, individuals might be somewhat immunized against the debilitating effects of traumatic events by having had prior experience with contingent control over such trauma. If deficits were caused by a traumatic event, then an effective method of therapy is to present the same traumatic event to an individual but in a context where their responses are favourably contingent toward the outcome (Seligman, & Maier, 1968).

In studies on life insurance salesmen (Seligman & Schulman, 1986), and college students (Peterson & Barrett, 1987), it was shown that individuals with an optimistic explanatory style (external, unstable, and specific) were more likely to remain persistent in the face of setbacks and adversity. In effect, such persistence enables the individual to continue, as opposed to giving up in the learned helplessness explanatory style, which in turn leads to more opportunities for success. Such an effect can have reinforcing potentials on an individual’s already established explanatory style, making that style more likely to continue in the event of future causal attributions.