Advances in Personality Psychology

Volume II

Andrzej Eliasz, Sarah E. Hampson, Boele de Raad, Andrzej Eliasz, Sarah E. Hampson, Boele de Raad

• Publisher: Psychology Press
• Year: 2006
• Print ISBN: 9781841695464
• eBook ISBN: 9781135421762
• Edition: 1
• Topic: Psychology
• Subtopic: History & Theory in Psychology
• Index: Psychology

The second volume in the Advances in Personality Psychology series, this book presents an authoritative collection of works by leading experts in the field. It focuses on three of the major issues in personality psychology: personality, affect and arousal; personality and intelligence; and personality structure.

The first part of the book seeks to analyse cognitive biases dependent on anxiety and the biological foundations of thought and action. It also looks at the influence of temperamental traits on reaction to traumatic events. In the second part, contributions consider the mutual relations between personality and intelligence, the similarities and differences between personality and intelligence, and the cognitive mechanisms of human intelligence and personality. The final part analyses personality structure across cultures and presents a model of personality relevant to situational descriptions.

All the authors are experienced and renowned experts in the field of personality psychology. The volume incorporates critical reviews, bringing the reader up-to-date with key issues, and unique data from contemporary empirical research projects, reflecting the diversity and vigour of current work on personality psychology.

Chapter 1
Cognitive approaches to trait anxiety

Michael W. Eysenck

There is increasing evidence that individual differences in the personality dimension of trait anxiety can be understood in part within a cognitive framework. More specifically, individuals high in trait anxiety possess various cognitive biases (e.g. attentional bias; interpretive bias) which lead them to exaggerate the threateningness of external and internal stimuli. These cognitive biases have recently been shown to have causal effects on individuals’ level of experienced anxiety. The original cognitive approach to trait anxiety was limited, because no distinction was drawn between two types of individuals scoring low on trait anxiety: (a) the truly low-anxious, who are non-defensive; (b) repressers, who are defensive. There is accumulating evidence that repressors possess opposite attentional and interpretive biases leading them to minimize the threateningness of external and internal stimuli. In contrast, the truly low-anxious do not possess cognitive biases or opposite cognitive biases. The represser group is of particular theoretical significance, because repressers show large discrepancies across the three major domains in which anxiety is assessed: self-report; behavioural; and physiological. These discrepancies depend on repressers’ opposite cognitive biases. It will be important in future research to integrate the cognitive approach to trait anxiety with a biological approach emphasizing the role of genetic factors in producing individual differences in trait anxiety.

INTRODUCTION

For many years, there was a considerable amount of controversy concerning the issue of the number and nature of major personality factors. However, in more recent years, there has been a growing consensus that there are five main personality factors, often referred to as the ‘Big Five’. The research of Goldberg (e.g. 1981) was influential in establishing five major factors, but the most influential theorists to emphasize the Big Five have probably been McCrae and Costa (e.g. 1985). According to their approach, the five factors are neuroticism, extraversion, agreeableness, conscientiousness and openness to experience.

This chapter is concerned with one of the Big Five personality factors, which has been variously described as neuroticism or trait anxiety. Neuroticism and trait anxiety overlap substantially with each other, as a result of which measures of the two dimensions typically correlate about +0.7 with each other (Eysenck and Eysenck, 1985). The key difference between them is that trait anxiety correlates negatively with extraversion, whereas neuroticism typically does not (Eysenck and Eysenck, 1985). More generally, there is convincing evidence that most measures of trait anxiety and neuroticism (as well as measures of depression) correlate highly with a personality dimension sometimes labelled negative affectivity (Watson and Clark, 1984).

The approach adopted by most advocates of the Big Five factor approach to personality has focused on description rather than explanation. In general, there has been more progress in terms of identifying the structure of human personality than there has in terms of understanding the underlying mechanisms associated with individual differences along each of the dimensions identified. However, some progress has been made in recent years, with various researchers conducting twin studies in order to assess the role of heredity. For example, Loehlin et al. (1998) found that individual differences in all five personality factors were determined to a moderate extent by genetic factors.

Historically, the main attempts to explain individual differences in trait anxiety or neuroticism were provided by H.J.Eysenck (1967) and by Gray (1982), both of whom emphasized the role of genetic influences in determining individual differences in personality. For example, according to Eysenck (1982, p. 28), ‘genetic factors contribute something like two-thirds of the variance in major personality dimensions’. Genetic factors exert their influence by affecting the responsiveness of various parts of the physiological system. More specifically, H.J.Eysenck (1967) argued that those high in neuroticism have greater activity than those low in neuroticism in the visceral brain, which consists of several parts of the brain (hippocampus, amygdala, cingulum, septum and hypothalamus). In similar fashion, Gray (1982) argued that anxiety depends on the workings of a septo-hippocampal system.

The biological approach adopted by H.J.Eysenck (1967) and by Gray (1982) has received inconsistent support from psychological research. So far as the hypothesis that two-thirds of individual differences in neuroticism or trait anxiety are attributable to heredity is concerned, one of the most thorough studies (with many twins brought up apart) was the one reported by Pedersen et al. (1988). They assessed neuroticism in 95 monozygotic twin pairs brought up apart, 150 monozygotic twin pairs brought up together, 220 pairs of dizygotic twins brought up apart, and 204 pairs of dizygotic twins brought up together. They found that monozyotic twins brought up together had a correlation of +0.41, against +0.24 for dizygotic twins brought up together. For twins brought up apart, the correlations were +0.25 for monozygotic twins and +0.28 for dizygotic twins. These figures suggest that about 31 per cent of individual differences in neuroticism depend on genetic influences. However, the mean age of Pedersen et al’s sample (58.6 years) was higher than in most other studies, and a recent review has suggested that about 40 to 50 per cent of individual differences in neuroticism depend on genetic influences (Bouchard and Loehlin, 2001).

The findings from twin studies indicate that genetic influences account for half (or a little less than half) of individual differences in neuroticism or trait anxiety. Thus, it is clearly important to consider environmental factors in order to achieve a good understanding of neuroticism or trait anxiety. What about the second hypothesis of the biological approach, namely, that individual differences in trait anxiety or neuroticism depend on individual differences in the responsiveness of the visceral brain or septo-hippocampal system? The evidence is almost uniformly negative. Fahrenberg (1992, pp. 212–213) carried out a review of all of the available evidence, and came to the following pessimistic conclusion: ‘Over many decades research has failed to substantiate the physiological correlates that are assumed for emotionality and trait anxiety. There is virtually no distinct finding that has been reliably replicated across studies and laboratories.’

The evidence from empirical research demonstrating the limitations of the biological approach produced a situation in which there was no overall theory of trait anxiety and neuroticism which appeared adequate. However, the situation has changed to some extent in recent years. One of the main themes of this chapter is to argue that many of the limitations of the biological approach stem from its failure to consider seriously the role played by the cognitive system. It is increasingly recognized by personality researchers that an understanding of cognitive processes and structures can serve to enrich theories of personality (McCann and Endler, 2000).

As will be seen, there is compelling evidence that there are systematic differences in cognitive functioning between individuals high and low in trait anxiety. More speculatively, these individual differences in cognitive functioning can be regarded as providing a partial explanation for the limitations of the biological approach. The remainder of this chapter is devoted to a consideration of cognitive approaches to trait anxiety, with the ultimate goal being to combine such approaches with the earlier biological approach.

RECENT RESEARCH

During the 1980s and 1990s, several theorists argued that a cognitive approach can play an important part in providing an understanding of individual differences in trait anxiety or neuroticism. Examples of such theories include those of Williams et al. (1988, 1997), Wells and Matthews (1994) and Eysenck (1992). There are many important differences among these theories, but they do share some major assumptions. Of particular importance, it was assumed in all of these theories that individuals high in trait anxiety or neuroticism possess a range of cognitive biases which lead them to exaggerate the threateningness of many stimuli. It was also assumed that individuals low in trait anxiety or neuroticism lack such cognitive biases. We turn now to a consideration of some of the evidence relating to these assumptions.

Cognitive biases

Several reviews of the literature on trait anxiety and cognitive biases have been published (e.g. Eysenck, 1992; Williams et al., 1997), and so only a brief description of some of the main findings will be attempted here. In essence, the main focus has been on three cognitive biases. First, there is attentional bias, which is defined as a tendency to pay attention to threat-related rather than to neutral stimuli. There is convincing evidence that individuals high in trait anxiety have an attentional bias, and this bias tends to be stronger when the situation is stressful. For example, MacLeod and Mathews (1988) found that high-anxious students had an attentional bias for examination-related words shortly before an important examination, but did not do so several weeks beforehand. There was no evidence for an attentional bias among individuals scoring low on trait anxiety.

Second, there is interpretive bias, which can be defined as the tendency to interpret ambiguous stimuli and situations in a threatening fashion. There is convincing evidence that high-anxious individuals have an interpretive bias whereas low-anxious individuals do not. For example, Eysenck, MacLeod and Mathews (1987) and Byrne and Eysenck (1993) studied the interpretations given to auditorily presented homophones (e.g. PANE, PAIN) possessing a threat-related and a neutral interpretation and spelling. In both studies, it was found that individuals high in trait anxiety interpreted more of the homophones in a threatening way than did individuals lower in trait anxiety. Third, there is memory bias, in which memory performance is better for negative or

Third, there is memory bias, in which memory performance is better for negative or threatening information than for positive or neutral information. This memory bias has been found in tests of explicit memory depending on conscious recollection (explicit memory bias) and tests of implicit memory in which conscious recollection is not involved (implicit memory bias). Williams et al. (1997, pp. 285–288) discussed studies on explicit and implicit memory biases in anxious and depressed individuals, and came to the following conclusion:

Out of nine studies using indirect [implicit] tests of memory in anxious subjects or patients, seven have found significant bias towards negative material…no study has yet found word congruent bias in implicit memory in depression…all published studies appear to find explicit memory biases in depression, yet only a third of the studies on trait anxiety or GAD [generalized anxiety disorder] find explicit memory biases.

There are various reasons why these memory biases are discussed much less than attentional and interpretive biases in the remainder of this chapter. First, it is not altogether clear why there are differing effects of anxiety on explicit and implicit memory bias, or why there are systematic differences between anxious and depressed individuals. Second, the findings seem less consistent than was suggested by Williams et al. (1997). For example, Richards et al. (1999) carried out three experiments to study implicit memory bias in high-anxious individuals. They failed to replicate previous findings, concluding: ‘None of the experiments offered any support for the prediction of a threat-related implicit memory bias in high-trait anxiety’ (Richards et al., 1999, p. 67). Third, there is a lack of persuasive theoretical reasons for assuming that memory biases (whether explicit or implicit) play a major role in accounting for individual differences in trait anxiety. As we will see, the situation is very different so far as attentional and interpretive biases are concerned.

Pre-attentive vs. attentional processes

An issue of theoretical importance is whether the attentional and interpretive biases exhibited by individuals high in trait anxiety involve pre-attentive processes. Most of this research has focused on attentional processes, and the majority of the relevant studies have uncovered evidence that pre-attentive processes are important. For example, Mogg, Kentish and Bradley (1993) carried out a study using the emotional Stroop task, in which attentional bias is revealed by slower colour naming in the presence of threat-related words than of neutral words. The words were presented either subliminally or supraliminally. The high-anxious participants showed a selective bias when the threat-related words were presented subliminally, but they failed to do so when the words were presented supraliminally. Van den Hout et al. (1995) also used the emotional Stroop task under subliminal and

presented supraliminally. Van den Hout et al. (1995) also used the emotional Stroop task under subliminal and supraliminal conditions. They found that the high-anxious participants showed a significant selective bias effect when the threat-related words were presented subliminally, and the same was also the case when the words were presented supraliminally.

The available evidence suggests that interpretive biases probably do not involve only automatic or pre-attentive processes. There have been several studies (e.g. Calvo, Eysenck and Castillo, 1997; Calvo and Castillo, 1998; Calvo and Eysenck, 2000) in which the time course of the development of an interpretive bias for ambiguous material has been assessed. The consistent finding has been that it takes of the order of several hundred milliseconds for an interpretive bias to develop. The finding that interpretive bias does not develop rapidly makes it unlikely that the bias depends primarily on automatic or pre-attentive processes.

Causality

One of the major problems with most of the research on cognitive biases is that the evidence obtained is essentially correlational in nature, and thus precludes assignment of causality. More specifically, it has been found repeatedly that individuals who report high levels of anxiety typically have various cognitive biases, but it is not clear whether the cognitive biases play a part in producing the anxiety, whether anxiety produces cognitive biases, or whether the causality is bidirectional. However, some recent research (discussed below) has shed light on this important issue.

Mathews and Mackintosh (2000) carried out a study in which a number of different procedures were used in order to produce an interpretive bias in the participants. In essence, the situation was set up so that ambiguous material would predominantly be interpreted in a negative fashion. The key findings revolved around the discovery that state anxiety was increased when the procedures used necessitated the generation of personally relevant meanings, but this did not happen when personally relevant meanings were not constructed. According to Mathews and Mackintosh (2000, p. 602): These findings provide evidence consistent with a causal link between the deployment of interpretative bias and anxiety.’

There is also evidence that inducing an attentional bias can increase experienced anxiety. For example, MacLeod et al. (2002) compared individuals who received training designed to produce an attentional bias with other individuals who did not receive such training. They found that the individuals with an induced attentional bias had a more negative mood state than the individuals in the control when both groups were given a stressful anagram task.

If inducing an attentional bias can increase individuals’ level of experienced anxiety, then inducing an opposite attentional bias (i.e. avoidance of processing threat-related stimuli) should lead to a reduced level of anxiety. This prediction was supported in a number of experiments reported by Mathews and MacLeod (2002) in which the participants consisted of individuals with initially high levels of trait anxiety. In one of their experiments, one group of participants high in trait anxiety received a total of 7500 training trials designed to i...

Table of contents

• Cover Page
• Front Matter Page
• Title Page
• Copyright Page
• Table of Contents
• List of figures
• List of tables
• List of contributors
• 1 Cognitive approaches to trait anxiety
• 2 The functional significance of temperament empirically tested Data based on hypotheses derived from the regulative theory of temperament
• 3 Personality and information processing Biological foundations of thought and action
• 4 Three superfactors of personality and three aspects of attention
• 5 Personality, trait complexes and adult intelligence
• 6 Phenotypes and genotypes of personality and intelligence Similarities and differences
• 7 Personality structure across cultures Indigenous and cross-cultural perspectives
• 8 Situations that matter to personality
• Index