Research Designs in Personality

In this chapter, we have examined the types of personality measur es and the means for evaluating the quality of those measures. The next step in personality research is to use these measures in actual research designs. Although the variations are nearly infinite, there are three basic research designs in the field of personality psycholog experimental, correlational, and case study . Each has strengths and weaknesses. Each provides information that complements the information provided by the others.

Experimental Methods

Experimental methods are typically used to determine causality—that is, to find ou whether one variable influence another variable. A variable is simply a quality that differs, or can take dif ferent values, for dif ferent people. Height, for example, is a variable because individuals differ from each other in height. Aggressiveness is a variable because individuals dif fer in their levels of aggressiveness. Personality characteristics, such as extraversion and agreeableness, are other examples of variables. In order to establish the influence of one variable on anothe , several key requirements of good experimental design must be met: (1) manipulation of one or more variables and (2) ensuring that participants in each experimental condition are equivalent to each other at the beginning of the study .

In the first requirement, manipulation, the variable thought to be the influen is manipulated as part of the experiment. For example, if a drug is hypothesized to influence memor , then some participants get the drug and other participants get sugar pills; then all participants have their memories tested. The second requirement, equivalence, is accomplished in one of two ways. If the experiment has manipulation between groups, then the random assignment of participants to experimental groups is a procedure that helps ensure that all groups are equivalent at the beginning of the study. However, in some experiments, manipulation is within each single group. For example, in the memory experiment, participants might get the drug and have their memories tested, then later take the sugar pills and have their memories tested again. In this case, each participant is in both conditions. In this kind of experiment (called a within-participant design), equivalence is obtained by counterbalancing the order of the conditions, with half of the participants getting the drug first and sugar pill sec ond, and the other half getting the sugar pill first and the drug second

The meaning of each of these features will become clear through an example of a personality experiment. Perhaps you are curious about why some people like to study with an iPod or TV on, whereas others demand total silence for studying. A personality theory predicts that extraverts prefer lots of stimulation and introverts

Introverts Prefer Read

Figure 2.

Performance on math and reading problems.

Figure 2.

Performance on math and reading problems.

prefer very little. Imagine being interested in testing the hypothesis that extraverts function best under conditions of high external stimulation, whereas introverts function best under conditions of low stimulation. To test this hypothesis, you could first give a grou of participants a self-report questionnaire that measures extraversion -introversion. Then you could select only those individuals who score at either extreme—as very introverted or very extraverted—to participate in your experiment. Next you would take these participants into the laboratory and have them work on math and sentence comprehension problems under two dif ferent conditions—in one condition, a radio would be blaring in the background and, in the other , there would be total silence. Half of each group (that is, half of the extraverts and half of the introverts) should be randomly placed in the noisy condition first and the quiet condition second. The other half should be placed in the quiet condition first and the noisy condition sec ond. Then, you would measure the number of errors each group makes under each of the two conditions. If the personality theory you are testing is correct, you should get a pattern of results like that in Figure 2.1. The hypothetical results in Figure 2.1 show that the extraverts made few errors in the noisy condition and more errors when it was quiet. The introverts showed the opposite pattern—noise hampered their performance, whereas they functioned best under conditions of silence.

This study, although hypothetical, highlights the key features of good experimental design. The first is manipulation. In this case, the external condition (th independent variable ) was manipulated—whether there was a lot of or a little ambient noise in the laboratory . The second feature is counterbalancing—half of the participants received the noisy condition first, whereas the other half received the quie

condition first. Counterbalancing is critical because there might be order effects as a consequence of being exposed to one condition first. Counterbalancing allows th experimenter to rule out order ef fects as an explanation for the results. The third feature is random assignment. Through random assignment, all persons have an equal chance of being selected for a given condition. Randomization can occur by flippin a coin or, more commonly, by the use of a table of random numbers. Randomization ensures that there are no predetermined patterns linked with condition assignment that could account for the final results

In experimental designs, it is desirable to establish whether or not the groups in the dif ferent conditions are significantl different. In the introversion/extraversion example, we want to know if the performance of introverts and extraverts in the noisy condition is significantly di ferent. Is the performance of the introverts significantl different from that of the extraverts in the quiet condition? To answer these questions, we need to know five things—sample size, the mean, the standard deviation, the t-test and the p-value (significance of the di ferences between the conditions).

The mean refers to the average—in this case, the average number of errors within each condition. The standard deviation is a measure of variability within each condition. Since not all participants make the same average number of errors, we need a way to estimate how much participants within each condition vary; this estimate is the standard deviation. Using these numbers, we can use a statistical formula—called the t-test—to calculate the dif ference between two means.

The next step is to see whether the dif ference is lar ge enough to be called significantly di ferent (the p-value). Although "large enough" is a somewhat arbitrary concept, psychologists have adopted the following convention: if the dif ference between the means would be likely to occur by chance alone (i.e., due to random fluctuations in the data) only 1 time out of 20 or less, then the di ference is statistically significan at the p <.05 level (the .05 refers to 5 percent chance level, or 1 time in 20). A difference between means that is significant at the .05 level implie that the finding would be likely to occur by chance alone only 5 times out of 100 Another way to think about this is to imagine that, if the experiment were repeated 100 times, we would expect to find these results by chance alone only 5 times

In sum, the experimental method is effective at demonstrating relationships among variables. Experiments similar to the one described, for example, have established a link between extraversion-introversion and performance under conditions of high versus low noise. The procedures of manipulating the conditions, counterbalancing the order in which the conditions occur , and randomly assigning participants to conditions help to ensure that extraneous factors are canceled out. Then, after calculating means and standard deviations, t-tests and p-values are used to determine whether the dif ferences between the groups in the two conditions are statistically significant. These procedures determine whether personality influences how people perform

Correlational Studies

A second major type of research design in personality is the correlational study . In the correlational method a statistical procedure is used for determining whether or not there is a relationship between two variables. For example, do people with a high need for achievement in college go on to earn higher salaries in adulthood than persons lower on need for achievement? In correlational research designs, the researcher is attempting to identify directly the relationships between two or more variables, without imposing the sorts of manipulations seen in experimental designs. Correlational designs typically try to determine what goes with what in nature. We might be interested, for example, in the relationship between self-esteem, as assessed through S-data, and the esteem in which a person is held by others, as assessed through O-data. Or we might be interested in how a measure of achievement motivation relates to grade point average. A major advantage of correlational studies is that they allow us to identify relationships among variables as they occur naturally . To continue the extraversion-introversion and performance under noise conditions example, we might measure people's preferences for studying with or without music in real life, then see if there is a correlation with their scores on a measure of introversion-extraversion.

The most common statistical procedure for gauging relationships between variables is the correlation coefficient To understand what correlation coef ficients indi cate, consider examining the relationship between height and weight. We might take a sample of 100 college students and measure their height and weight. If we chart the results on a scatterplot, we see that people who are tall also tend to be relatively heavy and that people who are short tend to be less heavy . But there are exceptions, as you can see in Figure 2.2.

Correlation coefficients can range from +1.00 through 0.00 to —1.00. That is, the variables of interest can be positively related to each other ( +.01 to +1.00), unrelated to each other (0.00), or negatively related to each other ( — .01 to —1.00). Height and weight happen to be strongly positively correlated with each other—with a calculated correlation coef ficient of + .60, for the data shown in Figure 2.2.

Consider a more psychological example. Suppose we are interested in the relationship between people's self-esteem and the amount of time they are unhappy . We might see a scatterplot as depicted in Figure 2.3. This scatterplot was obtained from a sample of college students, using a standard questionnaire measure of self-esteem. As the second variable, a measure of unhappiness, the participants were asked to keep a diary for two months, noting for each day whether that day was generally good (felt happy) or generally bad (felt unhappy). Then the percentage of days for each participant being unhappy was calculated. As you can see in Figure 2.3, as self-esteem goes up, the percentage of time a person is unhappy tends to go down. In contrast, those with low self-esteem tend to be unhappy a lot. In other words, there is a negative correlation between self-esteem and the percentage of time unhappy—in this case, approximately —.60.

As a final example, suppose we are interested in the relationship betwee extraversion and emotional stability (the tendency to be calm and secure). The

Height in feet

Figure 2.2

Fifty-five cases plotted, showing a strong positive correlation between height and weight. Each symbo (•) represents one person who was measured on both height and weight. Heavier persons tend to be taller; lighter persons tend to be shorter.

Height in feet

Figure 2.2

Fifty-five cases plotted, showing a strong positive correlation between height and weight. Each symbo (•) represents one person who was measured on both height and weight. Heavier persons tend to be taller; lighter persons tend to be shorter.

relationship is depicted in Figure 2.4. As you can see, there is no relationship between extraversion and emotional stability; as one variable goes up, the other may go up, down, or stay the same. In this case, the correlation coef ficient is 0.00. This means that you can find people with all the di ferent combinations of extraversion and emotional stability, such as those who are outgoing and sociable but also highly neurotic and unstable. In sum, relationships between variables can be positive, negative, or neither, as signified by positive, negative, or zero correlations

Percentage of time unhappy over 60 days

Figure 2.3

Fifty-eight cases plotted to illustrate the negative correlation between self-esteem and the percentage of time reported as being unhappy over two months. The correlation is —.60, indicating that people with higher self-esteem tend to be less unhappy than people with low self-esteem.

Most researchers are not merely interested in the direction of the relationship; they are also interested in the magnitude of the relationship, or how lar ge or small it is. Although what is considered lar ge or small depends on many factors, social scientists have adopted a general convention. Correlations around .10 are considered small; those around .30 are considered medium; and those around .50 or greater are





• •• M - • •


• •



2 1

1 1 1 1 1 1


5 2.5 3.5 4.5 5.5 6.5 7.5 Emotional stability

Figure 2.4

Fifty-seven cases plotted to show the relationship between emotional stability and extraversion. The correlation between these two variables is essentially 0.00, meaning that there is no relationship. Consequently, in the scatterplot, we see that people fall fairly equally in all sections of the plot, with no clear pattern.

considered large (Cohen & Cohen, 1975). Using the examples in Figures 2.2-2.4, the + .60 correlation between height and weight is considered lar ge, as is the - .60 correlation between self-esteem and percentage of time unhappy . These correlations are equivalent in magnitude but dif ferent in sign.

The concept of statistical significance can also be applied to correlation values. This is basically part of the statistical calculation, and it results in a numerical statement about how likely you are to find a correlation this size b chance, given the variables measured and the size of the sample. Here psychologists also require a probability of .05 or less before referring to a correlation as significant

It is important to keep in mind that one cannot infer causation from correlations. There are at least two reasons why correlations can never prove causality . One is called the directionality problem. If A and B are correlated, we do not know if A is the cause of B or if B is the cause of A. For example, we know there is a correlation between extraversion and happiness. From this fact alone, we do not know if being extraverted causes people to be happy or if being happy causes people to be extraverted.

The second reason that correlations can never prove causality is the third variable problem. It could be that two variables are correlated because a third, unknown variable is causing both. For example, the amount of ice cream sold on any given day may be correlated with the number of people who drown on that particular day . Does this mean eating ice cream causes drowning? Not necessarily, since there is most likely a third variable at work: hot weather . On very hot days, many people eat ice cream. Also, on very hot days, many people go swimming who otherwise don' t swim very much, so more are likely to drown. Drowning has nothing to do with eating ice cream; rather, these two variables are likely to be caused by a third variable: hot weather . With both correlational and experimental methods, it' s important to recognize that not all individuals conform to the generalizations established in the studies that use them.

Case Studies

Sometimes a personality researcher is interested in examining the life of one person in-depth as a case study . There are many advantages to the case study method. Researchers can find out about personality in grea detail, which rarely can be achieved if the study includes a lar ge number of people. Case studies can give researchers insights into personality that can then be used to formulate a more general theory to be tested on a larger population. They can provide in-depth knowledge of particularly outstanding individuals, such as Mahatma Gandhi or Martin Luther King. Case studies can also be useful in studying rare phenomena, such as a person with a photographic memory or a person with multiple personalities— cases for which lar ge samples would be dif ficult or impossible to obtain

One case study occupied an entire issue of the Journal of Personality (Nasby & Read, 1997). This study presents the case of Dodge Mor gan, who, at the age of 54, completed a nonstop solo circumnavigation of the earth by small boat. The case study reported by Nasby and Read is a highly readable account of this interesting man undertaking an almost impossible task. The focus is on how Mr . Morgan's early life experiences formed a particular adult personality , which led him to undertake the extreme act of going around the world alone in a small boat. The psychologists used Mor gan's voyage log book, autobiographical material, interviews, and even standard personality questionnaires in conducting their case study . The report is noteworthy in that the psychologists also discussed the strengths and weaknesses of the case study method for advancing the science of personality psychology . The authors concluded that personality theories provide a language for discussing individual lives; analysis of individual lives, in turn, provides a means for evaluating personality theories on how they help us understand specific individuals

Dodge Morgan was 54 when he completed a nonstop, solo circumnavigation of the earth in his boat American Promise. An extensive case study of this fascinating man was conducted by psychologists William Nasby and Nancy Read and reported in their paper "The Voyage and the Voyager" published in the Journal of Personality, 1997, volume 65, pages 823-852.

Dodge Morgan was 54 when he completed a nonstop, solo circumnavigation of the earth in his boat American Promise. An extensive case study of this fascinating man was conducted by psychologists William Nasby and Nancy Read and reported in their paper "The Voyage and the Voyager" published in the Journal of Personality, 1997, volume 65, pages 823-852.

Case study design can use a wide array of tools. One can develop coding systems to be applied to written texts, such as personal letters and correspondence. One can interview dozens of people who know the individual. One can interview the participant for hours and at great depth. One can follow the person around with a video camera and record, with sound and image, the actions in his or her everyday life. In sum, the assessment techniques used in case study designs are limited only by the imagination of the investigator .

Case Study: An Attention-Seeking Boy

One of the strongest advocates of the case study method was Gordon Allport, one of the founders of the field of modern personality psycholog . Allport firmly believe that important hypotheses about personality could come from examining single individuals in great depth. He also believed that one could test hypotheses about the underlying personality characteristics of a single individual using case study methods. The following example illustrates this sort of hypothesis formation and testing:

A certain boy at school showed exemplary conduct; he was or derly, industrious, and attentive. But at home he was noisy , unruly, and a bully toward the younger childr en . . .

Now the psychologist might make the hypothesis: This boy' s central disposition is a craving for attention. He finds that he gains his end best a school by conforming to the rules; at home, by disobeying them.

Having made this hypothesis, the psychologist could then actually count the boy's acts during the day (being checked by some independent observer) to see how many of them wer e "functionally equivalent," i.e., manifested a clear bid for attention. If the pr oportion is high, we can r egard the hypothesis as confirmed, and the p.d. [personality disposition] as established. (Allport 1961, p. 368)

Case Study: The Serial Killer Ted Bundy

Although Ted Bundy was convicted of killing three women, he was suspected of raping and killing as many as 36 women during his half-decade murder spree in the states of Oregon, Washington, Colorado, and Florida in the 1970s (Rule, 2000). Case studies have been devoted to explaining what drove Bundy to rape and kill. Some traced it back to the fact that he was adopted and felt a burning shame over the fact that he never knew his biological parents. Some tied it to his failed aspirations as a lawyer—where a status-striving motive was frustrated. Some traced it to the fact that he developed a deep-seated hostility toward women after being rejected by his fiancée—a woman wh was considerably higher than he in socioeconomic status and who he felt was impossible to replace. All case studies of Bundy revealed, however, that he shared many traits with other serial killers. He had a "classic" sociopathic personality—characterized by grandiosity, extreme sense of entitlement, preoccupation with unrealistic fantasies of success and power, lack of empathy for other people, a long history of deceitfulness, repeated failures to meet normally expected obligations of school and work, and high levels of interpersonal exploitativeness. Furthermore, Ted Bundy showed early behavior and personality dispositions that are known to be associated with serial killers, the so-called "serial killer triad": (1) torturing animals while young, (2) starting destructive fires, and (3) bedwetting. Case studies such as those of Ted Bundy can reveal unique aspects of his life (e.g., being rejected by a higher status fiancée, failure to achieve statu

Was this article helpful?

+1 0
Living Life

Living Life

Want to be and stay positive? Download This Guide And Discover 50 Tips To Have A Positive Outlook In Life. Finally! Reach your goals, face your fears and have a more fulfilled and happy life.

Get My Free Ebook


  • pauli
    What is aspect of personality in designing a research study?
    2 years ago
  • michael
    Which aspect of your personality is significantly involved in designing a research?
    2 years ago
  • Gaudenzia
    Which aspect of your personality is significantly involved in designing a research study?
    2 years ago
  • Harri
    What aspect of your personality is significantly involved in designing a researcher study why?
    2 years ago
  • nahand
    What aspect of your personality is significant involven in designing a research study . Why?
    2 years ago
  • hollie
    Which aspects of your personality is significantly involved in designing a research study?
    2 years ago
  • pertti
    Which aspect of your personality is significantly involved in designing a reseach study why?
    1 year ago
  • francis
    Which aspect of personality is significant involved in designing?
    1 year ago
    What is correlational designs in personality theory?
    1 year ago
  • nina
    Why evaluate personality psychology research?
    10 months ago
    What can cause the experimental method for studying personality not to work?
    7 months ago
  • muhammed
    Is personality mostly correlational research?
    7 months ago
  • kaisa
    How personality researchers use experimental research?
    7 months ago
  • ornella marchesi
    What is the best design to use in studying personality type?
    7 months ago
  • lisa
    How researchers have designed research to study personality?
    4 months ago
  • Luukas Kalliomäki
    How to design a personality study?
    4 months ago
  • beato
    What methods to psychologists use to study personality?
    4 months ago
  • Stefan
    What is the research design of personality?
    2 months ago
  • Andrew McIntyre
    What case study culture and personaltiy?
    1 month ago

Post a comment