Personality questionnaire.

One way in which to assess degree of extroversion in a person might be to observe how talkative and friendly he or she is among a group of people. But there are two problems with this behavior sampling approach to personality assessment. The first is that the method is very resource intensive, requiring substantial time and effort to obtain individual difference scores on the traits of interest. The second problem is that the situations in which the person is observed are very limited and circumscribed, and the behaviors elicited therein might not be typical of the person's behaviors more generally.

Contemporary methods of personality assessment generally rely on questionnaires, inventories, checklists, and so on. A personality questionnaire comprises one or more personality scales, each of which is designed to quantify differences among people on a personality trait or construct. The rationale behind the personality questionnaire is that people can be trusted to describe their characteristic behaviors accurately. Furthermore, to the extent that a person's self-description includes behaviors that are exemplars of a particular trait, that trait is inferred to be part of his or her personality makeup.

The basic unit of the personality scale is the personality item. Typical personality scales have items that represent statements describing specific behaviors, behaviors that are considered to be exemplars of personality traits. To illustrate, an extroversion measure might contain the item “I am usually the life of a party.” A person taking the questionnaire (respondent) might be asked to answer (endorse) the item as true or false. Alternatively, the response could take the form of choosing a number from a 5-point scale, where smaller numbers mean stronger disagreement with the statement and larger numbers mean stronger agreement. (Some personality questionnaires use single traitdefining adjectives as items instead of statements describing behaviors in situations. For example, an extroversion scale might instruct the respondent to rate the degree to which the term outgoing is characteristic of him or her.)

In the example item about being the life of a party, someone who endorses the statement (i.e., exemplar of extroversion) as true, or who agrees with the item, would be classified as being higher on the trait of extroversion than would someone who says false to the item or disagrees with the item. Note that an item measuring a trait might be reversekeyed. This means that a person endorsing the item as false (or disagree) would be measured to be higher on the trait. An example of a reverse-keyed extroversion item would be “I prefer spending time alone rather than in a large group.”

For obvious reasons, one would not generally measure a personality trait with only one behavior statement. Someone high in extroversion, for example, might paradoxically say false to the item “I go out of my way to talk to strangers” for any number of good reasons. A personality scale generally contains several such statements, and a respondent's attribute score is based on the sum or average of his or her endorsements of all the items in the scale. Statistically it can be shown that whereas any one item might grossly overestimate or underestimate the level of trait in a person, the overestimates tend to cancel the underestimates when several such items are averaged, resulting in a more accurate estimate of the person’s true level of trait. In general, a scale with more items measuring a trait has better psychometric properties than does a scale with fewer items.