Where Do the Data Come From?
1. Data Come from a Variety of Sources: Data about the person can come from the person’s life sphere or life space. Data from the person’s life space may include institutional information from school records, biological information from medical tests, and observer ratings. Data can also come from within the person, for example, from personal reports, including self-reports. What kinds of data come from the person him or herself?
What Research Designs Are Used in Personality Psychology?
2. Case Studies: Probably the earliest method used in personality psychology is the case study approach. When describing a single person, one is using a case study method. What can case studies accomplish that other methods can’t do as well? What are some of the drawbacks of case studies?
3. Observationism: How is observationism similar to the case study? How is it different? What historical events led to a de-emphasis on observation?
4. Correlational Research Designs: What does correlational design involve? What does the method tell us? A correlation coefficient indicates the relationship between two variables. It is of central importance to conducting studies using correlational designs. These coefficients have a range from −1 to 1. Can you say what a coefficient of 1.0 means? What about 0.0 and −1.0?
5. Natural Experiments and Full Experimental Designs: In an experiment, two groups are compared and hypothesized to be different in relation to an outcome, called the dependent variable. Typically, the two groups are called the experimental and the control groups. The experimental group has been changed through a treatment. The treatment is a manipulation of some sort that is expected to make the experimental group different from the control group. In a full experimental design, people are randomly assigned to the control and experimental groups, and then the treatment is applied in the experimental group. In a natural experiment, however, two groups that are naturally different are selected, such as two different occupational groups. What is the advantage of a full experimental design relative to a natural experiment? Given that superiority, why are natural experiments considered so useful in the study of personality?
What Does It Mean to Measure Personality?
6. The Psychometric Approach: The field of psychometrics is concerned with creating workable mathematical models for measuring people’s qualities. One doesn’t measure a person, but rather, one measures the attributes or features of a person. The fundamental theorem of psychometrics is that X = T + E, or, in words, that any observed score (X) reflects both a person’s true quality (T) on the attribute, and some degree of error (E). What happens if an error is positive, or if it is negative?
7. Test Reliability: Reliability is a property of all adequate measurement and concerns the degree to which tests measure with consistency. There are several different ways of assessing reliability, including the parallel test, internal consistency, and test-retest methods. Can you define each of these?
8. Test Validity: Validity refers to whether a test measures what it claims to measure. There are a number of kinds of evidence that are relevant to a test’s validity. Can you define each of them?
How Do Personality Psychologists Manage the Study of So Many Variables?
9. Psychologists Study Multiple Variables: Personality is a comprehensive system, including many individual parts, their organization, and their development. Measuring enough aspects of personality to make sense of it (and enough aspects of the life sphere to make sense of that) requires large numbers of individual observations. What is it about how we think of personality that encourages measuring so many different variables?
10. Approaches to Multiple Variables: Statistical techniques that examine many different variables are called multivariate techniques. A widely used method of looking at multiple variables is called factor analysis. When multiple variables are used, one often begins by calculating a simple index of the relationship between each pair of variables. What statistic tells us about how two variables are related?
11. Factor Analysis: Factor analysis concerns studying how many things a test measures. It can be used to reduce the number of variables of a test and also to demonstrate a type of validity. Do you know which sort of validity it helps determine? If you were presented a test that claimed to measure five things, and a factor analysis said it only measured three things, would you be able to conclude anything about the test’s structural validity?
12. Reading a Factor Analysis: Basically, findings from a factor analysis are presented as follows: A list of original variables forms a left hand column. To the right are columns representing factors, each labeled with a Roman numeral. The columns themselves contain a correlation between an original variable and its factor. Do you know what the correlation between a test item and a factor is called?
13. Factor Analysis and Dimensions: Factors with only positive (or negative) loadings are considered unipolar and are designated with one name (e.g., extraversion). Factors with both positive and negative loadings are called bipolar and are designated with two opposing names: (e.g., extraversion-introversion). Seeing a factor table, would you be able to read a factor analysis and name the factors?