Individual Differences

An entry for the Encyclopedia of Psychology (in press)
Kazdin, A.E. (Editor).
William Revelle
Department of Psychology
Northwestern University
Evanston, Illinios
Email: revelle@northwestern.edu

Individual Differences

That people differ from each other is obvious. How and why they differ is less clear and is the subject of the study of Individual differences (IDs). Although to study individual differences seems to be to study variance, how are people different, it is also to study central tendency, how well can a person be described in terms of an overall within-person average. Indeed, perhaps the most important question of individual differences is whether people are more similar to themselves over time and across situations than they are to others, and whether the variation within a single person across time and situation is less than the variation between people. A related question is that of similarity, for people differ in their similarities to each other. Questions of whether particular groups (e.g., groupings by sex, culture, age, or ethnicity) are more similar within than between groups are also questions of individual differences.

Personality psychology addresses the questions of shared human nature, dimensions of individual differences and unique patterns of individuals. Research in IDs ranges from analyses of genetic codes to the study of sexual, social, ethnic, and cultural differences and includes research on cognitive abilities, interpersonal styles, and emotional reactivity. Methods range from laboratory experiments to longitudinal field studies and include data reduction techniques such as Factor Analysis and Principal Components Analysis, as well as Structural Modeling and Multi-Level Modeling procedures. Measurement issues of most importance are those of reliability and stability of Individual Differences.

Research in Individual Differences addresses three broad questions: 1) developing an adequate descriptive taxonomy of how people differ; 2) applying differences in one situation to predict differences in other situations; and 3) testing theoretical explanations of the structure and dynamics of individual differences.

Taxonomies of individual differences:

Taxonomic work has focused on categorizing the infinite ways in which individuals differ in terms of a limited number of latent or unobservable constructs. This is a multi-step, cyclical process of intuition, observation, deduction, induction, and verification that has gradually converged on a consensual descriptive organization of broad classes of variables as well as on methods for analyzing them. Most of the measurement and taxonomic techniques used throughout the field have been developed in response to the demand for selection for schooling, training, and business applications.

Test Theory

Consider the case of differences in vocabulary in a particular language (e.g., English). Although it is logically possible to organize people in terms of the specific words they know in English, the more than 2^(500,000) possible response patterns that could be found by quizzing people on each of the more than 500,000 words in English introduces more complexity rather than less. Classical Test Theory (CTT) ignores individual response patterns and estimates an individual's total vocabulary size by measuring performance on small samples of words. Words are seen as random replicates of each other and thus individual differences in total vocabulary size are estimated from observed differences on these smaller samples. The Pearson Product Moment Correlation Coefficient (r) compares the degree of covariance between these samples with the variance within samples. As the number of words sampled increases, the correlation of the individual differences within each sample and with those in the total domain increases accordingly.

Estimates of ability based upon Item Response Theory (IRT) take into account parameters of the words themselves (i.e., the difficulty and discriminability of each word) and estimate a single ability parameter for each individual. Although CTT and IRT estimates are highly correlated, CTT statistics are based on decomposing the sources of variance within and between individuals while IRT statistics focus on the precision of an individual estimate without requiring differences between individuals. CTT estimates of reliability of ability measures are assessed across similar items (internal consistency), across alternate forms, and across different forms of assessment as well as over time (stability). Tests are reliable to the extent that differences within individuals are small compared to those between individuals when generalizing across items, forms, or occasions. CTT reliability thus requires between subject variability. IRT estimates, on the other hand, are concerned with the precision of measurement for a particular person in terms of a metric defined by item difficulty.

The test theory developed to account for sampling differences within domains can be generalized to account for differences between domains. Just as different samples of words will yield somewhat different estimates of vocabulary, different cognitive tasks (e.g., vocabulary and arithmetic performance) will yield different estimates of performance. Using multivariate procedures such as Principal Components Analysis or Factor Analysis, it is possible to decompose the total variation into between domain covariance, within domain covariance, and within domain variance. One of the most replicable observations in the study of individual differences is that almost all tests thought to assess cognitive ability have a general factor (g) that is shared with other tests of ability. That is, although each test has specific variance associated with content (e.g., linguistic, spatial), form of administration (e.g., auditory, visual), or operations involved (e.g., perceptual speed, memory storage, memory retrieval, abstract reasoning), there is general variance that is common to all tests of cognitive ability.

Personality and Ability

Although to some the term personality refers to all aspects of a person's individuality, typical usage divides the field into studies of ability and personality. Tests of ability are viewed as maximal performance measures. Ability is construed as the best one can do on a particular measure in a limited time (speed test) or with unlimited time (power test). Personality measures are estimates of average performance and typically include reports of preferences and estimates of what one normally does and how one perceives oneself and is perceived by others.

The same procedures used to clarify the structure of cognitive abilities have been applied to the question of identifying the domains of personality. Many of the early and current personality inventories use self-descriptive questions (e.g., do you like to go to lively parties; are you sometimes nervous) that are rationally or theoretically relevant to some domain of interest for a particular investigator. Although there is substantial consistency across inventories developed this way, some of this agreement could be due to conceptually overlapping item pools. Other researchers have advocated a lexical approach to the taxonomic problem, following the basic assumption that words in the natural language describe all important individual differences. This shifts the taxonomic question from how are individuals similar and different from each other to how are the words used to describe individuals (e.g., lively, talkative, nervous, anxious) similar and different from each other.

Dimensional analyses of tests developed based on lexical, rational, or theoretical bases suggest that a limited number (between three and seven) of higher order trait domains adequately organize the thousands of words that describe individual differences and the logically infinite way that these words can be combined into self or peer report items. The broadest domains are those of introversion-extraversion and emotional stability-neuroticism, with the domains of agreeableness, conscientiousness and intellectual openness or culture close behind. These domains can be seen as asking the questions that one wants to know about a stranger or a potential mate: are they energetic and dominant (extraverted), emotionally stable (low neurotic), trustworthy (conscientious), loveable (agreeable), and interesting (intelligent and open).

Measures of ability and personality reflect observations aggregated across time and occasion and require inferences about stable latent traits thought to account for the variety of observed behaviors. However there are other individual differences that are readily apparent to outside observers and require little or no inference about latent traits. The most obvious of such variables include sex, age, height, and weight. Differences that require some knowledge and inference are differences in ethnicity and social economic status. These obvious group differences are sometimes analyzed in terms of the more subtle measures of personality and ability or of real life outcomes (e.g, sex differences in neuroticism, mathematics ability, or income).

Predictive Validity

Individual differences are important only to the extent that they make a difference. Does knowing that people differ on a trait X help in predicting the likelihood of their doing behavior Y? For many important outcome variables the answer is a resounding yes. In their review of 85 years of selection in personnel psychology, Frank Schmidt and John Hunter (Psychological Bulletin, 1998, 124, 262-274) show how differences in cognitive ability predict differences in job performance with correlations averaging about .50 for mid complexity jobs. These correlations are moderated by job complexity and are much higher for professional-managerial positions than they are for completely unskilled jobs. In terms of applications to personnel psychology, a superior manager (one standard deviation above the mean ability for managers) produces almost 50% more than an average manager. These relationships diminish as a function of years of experience and degree of training. General mental ability (g) also has substantial predictive powers in predicting non-job related outcomes, such as likelihood of completing college, risk for divorce and even risk for criminality.

The non-cognitive measures of individual differences also predict important real life criteria. Extraversion is highly correlated with total sales in dollars among salespeople. Similarly, impulsivity can be used to predict traffic violations. Conscientiousness, when added to g substantially increases the predictability of job performance. Although the size of the correlation is much lower, conscientiousness measured in adolescence predicts premature mortality over the next fifty years.

Sources of individual differences

The taxonomic and predictive studies of individual differences are descriptive organizations of thoughts, feelings, and behaviors that go together and how they relate to other outcomes. But this categorization is descriptive rather than causal and is analogous to grouping rocks in terms of density and hardness rather than atomic or molecular structure. Causal theories of individual differences are being developed but are in a much earlier stage than are the descriptive taxonomies.

Descriptive taxonomies are used to organize the results of studies that examine genetic bases of individual differences. By applying structural modeling techniques to the variances and covariances associated with various family constellations it is possible to decompose phenotypic trait variance into separate sources of genetic and environmental variance. The most common family configurations that are used are comparisons of identical (monozygotic) with fraternal (dizygotic) twins. Additional designs include twins reared together or apart, and biological versus adoptive parents, children and siblings. Conclusions from behavioral genetics for most personality traits tend to be similar: Across different designs, with different samples from different countries, roughly 40-60% of the phenotypic variance seems to be under genetic control with only a very small part of the remaining environmental variance associated with shared family environmental effects. Additional results suggest that genetic sources of individual differences remain important across the lifespan. However, this should not be taken to mean that people do not change as they mature but rather that the paths one takes through life are similar to those taken by genetically similar individuals.

Genes do not code for thoughts, feelings or behavior but rather code for proteins that regulate and modulate biological systems. Although promising work has been done searching for the biological bases of individual differences it is possible to sketch out these bases only in the broadest of terms. Specific neurotransmitters and brain structures can be associated with a broad class of approach behaviors and positive affects while other neurotransmitters and structures can be associated with a similarly broad class of avoidance behaviors and negative affects. Reports relating specific alleles to specific personality traits emphasize that the broad personality traits are most likely under polygenic influence and are moderated by environmental experience.

Subtle differences in neurotransmitter availability and re-uptake vary the sensitivity of individuals to cues about their environment that predict future resource availability and external rewards and punishments. It is the way these cues are detected, atttended to, stored, and integrated with previous experiences that makes each individual unique. Current work on the bases of individual differences is concerned with understanding this delicate interplay of biological propensities with environmental opportunities and constraints as they are ultimately represented in an individual's information processing system. With time we can expect to increase our taxonomic and predictive power by using these causal bio-social theories of individual differences.


References:
Visit the personality project for more information about individual differences and personality.