Traits Across the Lifespan: Continuity and Change

c06.qxd 3/30/05 1:20 PM Page 201

CHAPTER

6

Traits Across the Lifespan: Continuity and Change

The Continuity of Traits Two Kinds of Continuity Differential Continuity in the Adult Years Childhood Precursors: From Temperament to Traits

The Origins of Traits: Genes and Environments The Logic of Twin and Adoption Studies Heritability Estimates of Traits Shared Environment Nonshared Environment Feature 6.A. Birth Order: A Nonshared Environmental Effect How Genes Shape Environments

Change and Complexity Different Meanings of Change Trait Change in the Adult Years Patterns of Traits Over Time Feature 6.B. Happiness Over the Human Lifespan What Else Might Change?

Summary

201

c06.qxd 3/30/05 1:20 PM Page 202

202 PART 1I SKETCHING THE OUTLINE: DISPOSITIONAL TRAITS AND THE PREDICTION OF BEHAVIOR

You may not be old enough to have gone to your first high school reunion. I have only been to one. In 1982 I went back to Gary, Indiana, for my 10-year reunion (class of '72, Gary Lew Wallace High School). As I remember it, probably 200 people showed up for the reunion, out of a graduating class of almost 500. So many things change in your life between your 18th and 28th birthdays, especially if you move away from home after graduation. Returning to the scene of so much adolescent anxiety, I was now a young adult, married, employed, well-educated, and more confident in life than I had been 10 years before. I felt I had changed, and I am sure that many of my former classmates in the crowd felt that they had changed significantly, too. After all, we had all grown up now; many of us had gone on to earn college degrees; some had joined the military; many had traveled; most of us were working; many had started families; a few were embarking on a second (or even third) marriage. Certainly, our life circumstances had changed.

But had we changed? Were we really different, as people, than we were before? For some of us, the answer surely seemed to be yes. Take, for example, Mary Anne Cromwell (I've changed the names). I hardly recognized her. I remembered Mary Anne Cromwell as a shy and painfully thin adolescent girl, poorly dressed, not very popular. I know she was the butt of jokes. I am ashamed to admit that as a junior I spread rumors that an ex-friend of mine was in love with Mary Anne; I wanted to get back at him for some insult he had hurled my way (I don't remember what it was). Mary Anne sat next to me in junior-year trigonometry, and I don't believe we ever exchanged words. Well, the 10-year reunion could be labeled "Mary Anne's revenge." Mary Anne had grown up to be a beautiful woman, attracting a crowd of men at the reunion who seemed not even to know that she existed 10 years before. Mary Anne had gone on to get an M.B.A., and she had already enjoyed considerable success in business. What struck me most, though, was her increased poise and her assertiveness. Gone was the timidity and the awkwardness I remembered. Now, you may say that I just didn't know her very well in high school, and that certainly is true. Maybe she always had this poise, and I never saw it. But not too many other people saw it either, it seemed, for everybody I spoke with at the reunion seemed astounded by how much Mary Anne Cromwell had changed.

But let's look at a counter-example--Robert Amundson. Robert is the guy about whom I spread the rumors concerning Mary Anne Cromwell back in our junior year. After my prank, Robert had the good sense to dump me as a friend and move on to more exalted social circles. Actually, Robert was always a social climber, which was one of the reasons our friendship was stormy and short-lived. In high school, I remember him as socially dominant, outgoing, spontaneous, and not especially conscientious. His social dominance was clearly in evidence the night of the reunion. Just like old times, Robert was the center of attention. It seemed to me that he boasted just as much as he always had. He dominated conversations, as he always had. People seemed to accord his opinions an especially high status, just as they always had. The only difference I could see with respect to Robert Amundson was his newfound eagerness to talk with Mary Anne Cromwell.

While Mary Anne Cromwell seemed to have changed significantly, my overall impression from the high school reunion was that Robert Amundson was more the norm. The evidence for continuity in personality seemed very powerful to me that night. I could not help but be struck by how most people seemed so much the same as they had been ten years before. Ten years had passed and yet almost everybody was so recognizable in their characteristic social mannerisms, their modes of relating to others, the way they spoke, even the things they spoke about. Michael still had the best sense of humor. Roberta was

c06.qxd 3/30/05 1:20 PM Page 203

CHAPTER 6 TRAITS ACROSS THE LIFESPAN: CONTINUITY AND CHANGE 203

still nonconventional and rebellious. Kevin was still obsessed with sports. Barbara still danced in a provocative way. Keith still seemed depressed.

My observations about Mary Anne's "poise," Robert's "social dominance," and Keith's "depressiveness" were attributions about the personality traits these people seemed to display, as seen in high school and 10 years later. Do people's dispositional traits change over time? Or do traits remain stable? Can we expect that Robert Amundson will still be socially dominant when I see him, God willing, at my 40-year high-school reunion? Will Mary Anne Cromwell go through new changes over time? What can we expect from Michael or Barbara 40 years after graduating from high school? The issue of stability and change in personality traits leads naturally to a related question: Where do personality traits come from, anyway? They don't just emerge out of thin air. Traits must develop, in some sense, over time. Is their development driven by our genes? By our environments? These are surely some of the most interesting questions in all of personality psychology. Some of the answers that are beginning to emerge from personality research may fit your expectations. But I bet some will surprise you.

THE CONTINUITY OF TRAITS

In everyday talk, attributing a trait to a person presupposes a certain degree of continuity over time. If I tell you, for example, that my nephew is a very "agreeable" person, I am implying that his high level of agreeableness has some staying power over time. Despite momentary fluctuations, I am suggesting that he tends to be agreeable. Just how long-term that staying power might be, however, is not clear from my attribution. Do I mean that he is likely still to be an agreeable person a week from now? Yes, I do mean to suggest that. What about next year? Probably. Twenty years from now? I don't know. Am I suggesting that he will continue to be a relatively agreeable person for the rest of his life? No, I am not necessarily saying that.

When people use trait terms in everyday talk, they seem to be operating from an implicit assumption that certain features of human individuality are more-or-less stable "for a while." My nephew is a highly agreeable person, and I expect that he will remain highly agreeable for a while. It is not clear how long "for a while" is. But it is long enough in time to allow me to attribute a trait. For this reason, psychologists expect personality trait scales to show high test?retest reliability in the short run (chapter 4). Traits are not supposed to change dramatically from, say, one day to the next. If they did, then they wouldn't be "traits." Therefore, a trait scale that gives wildly different scores for a given individual from one day or week to the next is thought to lack test?retest reliability, and such a scale is likely to be rejected by the scientific community. But what about a given trait for a given person over the long haul? Beyond the "for a while," how much continuity can we expect in traits from one year to the next? From young adulthood to middle adulthood? From childhood through old age?

TWO KINDS OF CONTINUITY Before we can begin to answer these questions, we need to distinguish between two very different meanings of continuity in personality traits.

The first meaning is absolute continuity. Caspi (1998) defines absolute continuity as "constancy in the quantity or amount of an attribute over time" (p. 346). In absolute terms,

c06.qxd 3/30/05 1:20 PM Page 204

204 PART 1I SKETCHING THE OUTLINE: DISPOSITIONAL TRAITS AND THE PREDICTION OF BEHAVIOR

"how much" social dominance did Robert Amundson have in 1972 and in 1982? Let us say that we administered a well-validated scale of this trait to Robert in 1972, and at that time he scored 27. On the exact same scale in 1982, he again scored 27. In this case, we could say that Robert showed absolute continuity on the trait of social dominance over a 10-year span. But the concept of absolute continuity is virtually never applied to the single individual. Instead, absolute continuity is usually understood in terms of group averages on a given trait (Caspi, 1998). As an example, we might compare the average social dominance score for the graduates of Gary Lew Wallace High School obtained in 1972, when they were 18 years of age, with the average score obtained from the same group of people 10 years later, when they were 28. Such a study would help us answer this question: How much absolute continuity do we observe on social dominance across a 10-year period in the lifespan? If the average scores for age 18 and 28 are highly similar, then we might conclude that the group, as a whole, exhibited continuity on this trait. People were no more or less socially dominant, on the average, at age 28 compared with age 18.

The issue of absolute continuity comes into important play when personality psychologists consider certain hypotheses and expectations about human development. For example, some developmental theories suggest that adolescence is a time of confusion and anxiety for many people, but that things settle down a bit as individuals move into adulthood (e.g., Blos, 1979). With respect to my reunion experience, then, we might expect that the group average scores on measures of "anxiety" or "insecurity" or even "rebelliousness" might have been higher among my peers in 1972, when we were high school students, than in 1982, when we were young adults. Our expectations would suggest that absolute continuity would not occur across this developmental time period, that people in general would change, and that these changes would be captured in changing average scores over time. I did not have the foresight to administer personality tests to my classmates in 1972 and 1982. But my overall gut impressions suggested to me that while there may have been absolute continuity on many traits, a few things did seem to have changed for the group as a

Individual differences in personality traits tend to show substantial differential continuity over time. In the same way that this woman is recognizable at three different points in her life, so too may certain personality dispositions-- e.g., extraversion, agreeableness, neuroticism--be recognizable over time.

c06.qxd 3/30/05 1:20 PM Page 205

CHAPTER 6 TRAITS ACROSS THE LIFESPAN: CONTINUITY AND CHANGE 205

whole. My nonscientific sense was that as 28-year-olds we were generally more comfortable with our lives and our identities in the overall than we had been in high school. Had someone done the study, I would have expected a slight drop in social anxiety from age 18 to age 28 for the group as a whole, and perhaps a slight increase in friendliness, too. But we will, of course, never know.

Differential continuity refers to "the consistency of individual differences within a sample of individuals over time, to the retention of an individual's relative placement in a group" (Caspi, 1998, p. 345). When I suggest that Robert Amundson seemed to be as socially dominant at age 28 as he was at age 18, I am making a statement about differential continuity. I am suggesting that, relative to his peers, Robert tends to be pretty high on the social dominance dimension. Differential continuity is always a matter of individuals' relative standing to one another on a given dimension. In the case of Robert's social dominance, I might suggest that in a random sample of, say, 100 of my classmates from Gary Lew Wallace, Robert was probably about the second or third most socially dominant person in 1972. My observations 10 years later suggested to me that he remained toward the top on this dimension, compared again with his peers. But the extent to which the trait social dominance shows differential continuity in my example is a product of not only Robert's placement on the social dominance dimension, but everybody else's placement, too. For there to be high levels of differential continuity for this trait, not only would Robert need to continue to score toward the top of the distribution over time, but other people would also need to hold their relative positions on this trait over time. In other words, if social dominance shows high differential continuity over a 10-year span, then Robert would need to continue to score very high (let's say a "10" on a 10-point scale), Barbara (who, let us say, scored an "8" in 1972) would need to continue to score moderately high, I would need to continue to score in the medium range (let's figure I was around a "5"), and Keith (who scored "2" in 1972) would need to continue to score in the low range. In simple terms, Robert, Barbara, Dan, and Keith would "stay the same" over the 10-year period relative to one another on this trait. They would each continue to hold their positions in the distribution.

The extent to which individuals continue to hold their relative positions on a trait dimension over time is typically calculated with a correlation coefficient (chapter 1). The correlation is calculated between the same individuals' scores on the same trait at Time 1 (in this case, 1972, at age 18) and Time 2 (1982, age 28). High correlation coefficients (those approaching the value of +1.0) suggest high differential continuity. If, for example, we actually had obtained social dominance scores from 100 of my classmates in 1972 and again (the same 100 classmates) in 1982, then we could calculate the correlation between those two sets of scores. Had we done so and had we obtained, let us say, a correlation value of +.80, we would have evidence for very strong differential continuity on this trait. The high value of the correlation coefficient (+.80) would tell us that people tended to hold their relative positions on the given trait over the 10-year span. Thus, a person's score on the trait in 1972 would be a good predictor of the same person's score on the same trait assessed 10 years later. By contrast, had we obtained, let us say, the modest correlation value of +.10 in this study, then we would have to conclude that social dominance, as measured by our scale, does not show strong differential continuity. Low differential continuity means that people's relative positions on the given dimension change unpredictably over time. Thus, if there were low differential continuity on the trait, it would be very difficult to predict a person's social dominance score at age 28 from his or her score at age 18, because individuals' relative positions to one another on this trait would appear to vary

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download