Advances in Personality Psychology

Volume 1

Sarah E. Hampson

• Publisher: Psychology Press
• Year: 2014
• Print ISBN: 9780415227681
• eBook ISBN: 9781317797746
• Topic: Psychology
• Subtopic: History & Theory in Psychology
• Index: Psychology

In the first volume of this new series, Sarah E. Hampson brings together a unique collection of critical reviews of key areas of personality psychology and integrative accounts of important work by internationally recognised experts in the field.
Advances in Personality Psychology includes chapters on cross-cultural evidence for the Big-Five framework for personality description, type and trait approaches to understanding childhood personality, developments in psychometrics, the relationship between hostility and cardiovascular disease, and the connections between personality and emotions. In further chapters the view that personality cannot change in adulthood is challenged and the importance of environmental factors is revealed by an observational study of twins.
This state-of-the-art volume will provide students, teachers and researchers of contemporary personality psychology with a highly valuable resource on recent developments in this area.

Preface to volume 1

The chapters in this volume are by a selection of invited speakers and presenters at the Ninth European Conference on Personality Psychology held at the University of Surrey, Guildford, UK in July 1998. These authors were invited to produce extended versions of their papers with additional collaborators if they wished. The result is a collection of chapters that take some of the most fundamental questions of our field and address them in depth.

The first chapter, by Saucier, Hampson and Goldberg, reviews the evidence for the lexical Big Five in cross-language studies. This chapter addresses the fundamental question of whether the broad dimensions of personality description are the same across numerous different languages. The authors identify a series of methodological features on which cross-language studies vary and which affect the assessment of similarity between the Big Five in English and the factor structures that emerge in other languages. The Big Five have become widely accepted as a unifying framework in personality psychology and this framework is used by authors of several subsequent chapters. Mervielde and Asendorpf, in Chapter 2, relate the personality and temperament factors found in children to the Big Five in adults. In so doing, they evaluate the similarities among the various theories of temperament, and between theories of temperament and theories of personality in children. They review both variable-centred and person-centred approaches to the study of childhood personality and conclude that both make unique contributions to our understanding of children’s personalities. The issue of personality stability versus change in adulthood is the focus of Chapter 3 by Helson and Kwan. This chapter reviews findings from several cross-sectional and longitudinal studies of adults, and then gives an overview of the many findings from the Mills study. The authors demonstrate that cross-sectional and longitudinal findings are consistent in identifying certain aspects of personality that change over the course of adulthood. In their discussion of longitudinal studies, and of the Mills study in particular, they relate personality change in adulthood to the social influences that have affected different cohorts as they moved along their unique pathways through time.

In Chapter 4, Borkenau, Riemann, Spinath and Angleitner address the behaviour genetics of personality by focusing on observational studies. The genetics of adult personality is based exclusively on self-report measures of personality, which may have distorted the estimates of genetic, shared and non-shared contributions. They provide a comprehensive review of observational studies of young twins, and conclude that the shared environment emerges as a more important source of variance in these studies than in self-report studies. They present findings from the German Observational Study of Adult Twins to support their hypothesis that genes affect behaviour at a more global level, whereas the environment is the only source of variance at more specific levels.

The last three chapters address the interface between personality and other fields in psychology. In Chapter 5, ten Berge bridges the gap between psychometricians and personality psychologists by providing a discussion of some recent developments in factor analysis and the assessment of internal reliability. In Chapter 6, Whiteman, Deary and Fowkes consider the relationship between personality and health by examining in depth the association between Type A and hostility-related traits and cardiovascular diseases. Their review of past evidence indicates a small but consistent association that needs to be studied longitudinally so that causal relations and explanatory mechanisms can be identified. They present data from the Edinburgh Artery Study for both concurrent and prospective associations between various measures of hostility and disease variables, and point to possible underlying biological pathways. In Chapter 7, Matthews, Derryberry and Siegle examine the associations between personality and emotion by studying individual differences in emotional reactivity within a cognitive science framework. They focus on three cognitive science perspectives: cognitive neuroscience, connectionism, and transactional stress processes, each reflecting one of the authors’ research programmes. This chapter is an integrative and collaborative enterprise that has generated a novel approach to a classic issue in personality.

All the contributions to this volume underwent the peer review process. I would like to take this opportunity to express my thanks to the colleagues whose advice I sought: John C. Barefoot, Paul Costa, Alexandra M. Freund, Lewis R. Goldberg, Robert Krueger, Robert R. McCrae, Roy P. Martin, Ivan Mervielde, Todd Q. Miller, Edward Necka, Mary K. Rothbart, Gerard Saucier, Warren W. Tryon and Gerdi Weidner. Finally I would like to thank the Oregon Research Institute for providing a congenial environment for my sabbatical during which time much of the editing of this book was completed.

Sarah E. Hampson
Guildford, UK

Chapter 1

Cross-language studies of lexical personality factors

Gerard Saucier, Sarah E. Hampson and Lewis R. Goldberg

The rationale for lexical studies rests on the assumption that the most meaningful personality attributes tend to become encoded in language as single-word descriptors. Based on this rationale, a number of studies have been conducted examining the factor structure of adjectival descriptors extracted from dictionaries. Using as an initial working hypothesis an Anglo/German version of the Big Five factor structure, we review lexical studies in English and 12 other languages, and examine their fit with the Big Five. We find that the inclusion of highly evaluative terms and physical descriptors may lead to factors beyond the Big Five. With more conventional variable selections emphasizing disposition descriptors, indigenous structures resembling the Big Five seem to emerge most readily in northern European languages. The Big Five have been only partially reproduced in many other studies, although this might reflect the differences in their procedures for selecting variables. Factor structures with fewer factors (one, two or three) may be more generalizable cross-culturally, perhaps because they are less affected by these methodological variations. At this point, more attention should be given to the influence of the major types of procedural differences on the resulting factor structures. Such variations in research design and analysis may account for much of what otherwise might be prematurely interpreted as cultural differences. Suggestions for improving the lexical-study paradigm are offered.

INTRODUCTION

After many years of dispute, personality psychologists have recently shown some signs of agreement on a framework for classifying and organizing personality traits. A consensus among many in the field, which was evident by the early 1990s (e.g. McCrae and John, 1992), posits that five broad dimensions of personality capture the most important aspects of lexicalized personality traits. Table 1.1 indicates subcomponents associated with each of these ‘Big Five’ factors from a recent study of English and German personality descriptors (Saucier and Ostendorf, 1999).¹

In an influential critique, Block (1995) cautioned against settling prematurely on the five-factor approach. However, even those most closely allied with the five-factor framework regard it as a working hypothesis, rather than a final solution (e.g. De Raad, 1998; Goldberg and Saucier, 1995). Like any scientific model, its prime function is to stimulate research and either to be proved wrong or, more likely, to be refined in the light of new evidence. Additional dimensions could emerge in studies of different cultures or studies of different forms of personality language (e.g. typenouns), or different kinds of personality data. Moreover, findings from any of these sources could point to a better structure at a different hierarchical level, one with more factors or fewer factors.

Table 1.1 Big Five subcomponents found in English and German

		Example terms in English
Domain	Subcomponent	Positive pole	Negative Pole
Extraversion (I)	Assertiveness Activity–adventurousness Unrestraint Sociability	assertive daring talkative cheerful	weak unadventurous shy unsociable
Agreeableness (II)	Warmth–affection Generosity Gentleness Modesty–humility	affectionate generous agreeable modest	cold selfish harsh egotistical
Conscientiousness (III)	Orderliness Reliability lndustriousness Decisiveness	organized responsible ambitious decisive	sloppy undependable lazy inconsistent
Emotional Stability (IV)	Insecurity^a Emotionality^a Irritability^a	relaxed unemotional undemanding	insecure excitable irritable
Intellect (V)	Intellect Imagination Perceptiveness	intelligent creative perceptive	unreflective unimaginative shortsighted

Source: Saucier and Ostendorf (1999).
Note
a These subcomponents are labelled by their negative rather than positive pole.

Initial problems in developing a scientifically compelling descriptive classification include (a) defining the universe of personality attributes that will be classified and (b) determining the particular attributes from that universe that are important enough to be represented in the final classification. The lexical approach to personality solves these problems by a major and far-reaching hypothesis: the most important individual differences eventually become encoded as single words in the natural (i.e. non-scientific) language (Allport and Odbert, 1936; Cattell, 1943; Goldberg, 1981; Saucier and Goldberg, 1996b). Therefore, at least for the major modern languages, their dictionaries contain a reasonably comprehensive set of those individual differences that past speakers of those languages have needed to communicate. This ‘lexical hypothesis’ has two reasonable corollaries: (a) frequency of use has a rough correspondence with importance, and (b) the number of words referring to a particular personality attribute will be a rough guide to the importance of that attribute for the speakers of the language² (Saucier and Goldberg, 1996b). These two conceptually independent criteria can both be satisfied if one examines the most frequently used subset of relevant terms in the language, and then searches for structure (i.e. factors, clusters) in that subset.

Although the peculiarities of trait structure derived from any particular language could be of considerable interest, cross-language regularities have even more scientific import. In the search for such regularities, the most important personality dimensions will be those that replicate across samples of participants, targets of descriptions, variations in analytic procedures, and across languages (Saucier and Goldberg, 1996b). If a factor structure can be recovered in a wide variety of the world’s languages, then this would provide support for the possibility that those dimensions might represent cultural ‘universals’ of person description (Goldberg, 1981). This chapter reviews the evidence for the reproducibility of structural representations – especially the Big Five – across those languages that have been studied to date. Some of the major studies in English are summarized first to describe how the five-factor framework came to be established. Next, lexical studies in other languages are examined for the reproducibility of the five-factor framework. Because these studies differ in a number of their methodological details, we consider the potential impact of these differences on the findings obtained. Possible alternative structural frameworks are also considered. Finally, we consider the broader significance of this body of research.

THE LEXICAL APPROACH IN ENGLISH

The lexical assumption has been the guiding rationale for personality taxonomers dating back at least to the 19th-century British scientist, Francis Galton (1884), who used a dictionary to identify terms that described personality attributes. Particularly influential were Allport and Odbert (1936), who extracted from an unabridged English dictionary almost 18,000 words that could be used to describe individual differences. Of these, roughly 4500 were designated by Allport and Odbert as likely personality traits, the remainder having been classified as temporary moods or activities, social effects, physical or medical terms, predominantly evaluations, and the like. This classification scheme proved highly influential, leading to the decision by Norman (1967) and Goldberg (1982) to exclude from their initial analyses terms describing one’s physical appearance, the effects one has on others, sheer evaluations, and temporary activities, moods and states.

The detailed history of the initial extraction and later modification of trait lists has been documented repeatedly (e.g. Digman, 1990; Goldberg, 1993, 1995; John, 1990; McCrae and John, 1992; Wiggins and Trapnell, 1997), and thus will not be repeated here. In this summary, we will focus primarily on the English-language studies that permit comparisons with those in other languages.

At the dawn of the computer age, lexical studies in English found impetus from work by Norman (1963, 1967) and Goldberg (1976, 1982). Norman’s (1963) use of Roman numerals to identify the Big Five factors has become a hallmark of the lexically based taxonomic tradition; the five factors were ordered roughly by the number of reasonably familiar trait-descriptive adjectives in English available to describe that domain: I (Extraversion) and II (Agreeableness) with the most such terms, followed by III (Conscientiousness), and then IV (Emotional Stability) and V (first labelled Culture, here labelled Intellect) with the least adjectives.

Norman (1967) retraced the footsteps of Allport and Odbert (1936) and extracted person-descriptive terms from a new edition of the same unabridged English dictionary that they had used. Norman’s four-person research team refined the resulting set of 18,125 terms by excluding those classified as pure evaluations and as physical and medical terms and categorizing the remainder as (a) stable traits, (b) temporary states and activities or (c) social roles, relationships and effects. Through this process, Norman identified roughly 2800 stable trait terms. Goldberg (1976, 1982) reduced Norman’s stable trait pool to 1710 by eliminating the least commonly used terms and those that seemed the least dispositional in nature. Using these same subjective criteria, Goldberg later developed a 540-term set that was employed in a number of studies in which university students described themselves and others (Goldberg, 1990). After additional samples of university students had provided familiarity ratings for these and other common English adjectives, Saucier and Goldberg (1996a) analysed the most familiar subset of 435 adjectives in a combined sample of nearly 900 (including 507 self-descriptions and 392 descriptions of other), and found a clear five-factor structure.

Peabody (1987) developed 53 bipolar scales to represent a large pool of adjectives that included Goldberg’s 540 terms. Four college students made judgements of semantic similarity between all of the terms from each pole and each of the bipolar scales. Analyses of these judgements of ‘internal structure’ revealed the Big Five factors, plus a small ‘Values’ factor; the first three factors (Assertiveness [Extraversion], Agreeableness, and Conscientiousness) were, however, much larger than the remaining ones. Further analyses of these bipolar scales in self and peer descriptions by college students consistently found the Big Five factors (Peabody and Goldberg, 1989). But again, the first three factors (i.e. Extraversion [I], Agreeableness [II] and Conscientiousness [III]) were substantially larger than the other two (Emotional Stability [IV] and Intellect [V]).

The English-language studies of Goldberg and Peabody...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• List of figures
• List of tables
• List of contributors
• Series preface
• Preface to volume 1
• 1 Cross-language studies of lexical personality factors
• 2 Variable-centred and person-centred approaches to childhood personality
• 3 Personality development in adulthood: the broad picture and processes in one longitudinal sample
• 4 Behaviour genetics of personality: the case of observational studies
• 5 Linking reliability and factor analysis: recent developments in some classical psychometric problems
• 6 Personality and health: cardiovascular disease
• 7 Personality and emotion: cognitive science perspectives
• Index