Read Hauser-Kuo-Cartmill_Birth.Order.and.Personality.Among.Adult.Siblings.pdf text version

Birth Order and Personality among Adult Siblings: Are There Any Effects?

Robert M. Hauser Center for Demography and Ecology University of Wisconsin-Madison Hsiang-Hui Daphne Kuo Department of Sociology University of Washington Randi S. Cartmill Department of Sociology University of Wisconsin-Madison July 1999

An earlier version of this paper was presented at the March 1997 meetings of the Population Association of America, Washington, D.C. This research was supported by grants from the National Institute on Aging (AG-9775) and the Vilas Estate Trust. It was carried out using facilities of the Center for Demography and Ecology at the University of Wisconsin-Madison, for which core support comes from the National Institute of Child Health and Human Development. Statistical tabulation and estimation were performed using SPSS. All of the data used herein are in the public domain and are available on the World Wide Web at http://dpls.dacc.wisc.edu/wls/. We thank Judith Rich Harris for helpful comments. The opinions expressed herein are those of the authors. Correspondence should be directed to Robert M. Hauser, Department of Sociology, University of Wisconsin-Madison, 1180 Observatory Drive, Madison, Wisconsin 53706; E-MAIL [email protected]

Birth Order and Personality among Adult Siblings: Are There Any Effects?

ABSTRACT Frank Sulloway's 1996 book, Born to Rebel, renewed and excited public interest in effects of birth order on life chances. Sulloway's argument draws on a widely accepted psychological theory, which posits five basic dimensions of personality (the "Big Five"): Agreeableness, Conscientiousness, Extraversion, Neuroticism, and Openness to Experience. Sulloway argues that firstborns are more antagonistic, responsible, extraverted, neurotic, and conservative. The Wisconsin Longitudinal Study (WLS) provides multiple measures from a standard scale of the Big Five personality characteristics (John's BFI-54). These provide an opportunity to test the central theoretical argument of Born to Rebel. This differs from almost all prior work because the subjects are adults, the sample is large, and we can directly compare the personalities of siblings. The Wisconsin data confirm prior findings of reliability and stability in adult personality and of gender differences in agreeableness, extraversion, and neuroticism. However, in comparisons among Wisconsin high school graduates at midlife, among their randomly selected siblings, and between graduates and their siblings, birth order is not related to any dimension of personality. The most salient finding about the personality of siblings is a moderate level of resemblance among brothers and sisters.

Frank Sulloway's 1996 book, Born to Rebel, renewed and excited public interest in effects of birth order on life chances {Sulloway 1996}. The work was hailed as superb scientific detective work by some and as the social environmentalist's answer to Herrnstein and Murray's The Bell Curve by others. Sulloway appeared on the major radio and television talk shows, and he appeared able to explain every nuance of family dynamics by reference to birth order. Despite archives full of negative findings, birth order has a strong and persistent appeal as an explanatory variable. Everyone has a birth order, and -- despite the decline of primogeniture in modern societies -- just about everyone seeks to understand their place in their family and in the larger social world by reference to birth order. The question remains whether Sulloway's theories are valid. The main theme of Born to Rebel is that the social dynamics of relationships among parents and children lead first-born children to be conservative and later-borns radical. Sulloway offers both evolutionary justifications and historical examples of this, largely put in terms of the competition among siblings for parental attention and resources. His argument draws on a widely accepted psychological theory, which posits five basic dimensions of personality (the "Big Five"): Agreeableness, Conscientiousness, Extraversion, Neuroticism, and Openness to Experience {Costa and Widiger 1994}. Sulloway argues that firstborns are more antagonistic, responsible, extraverted, neurotic, and conservative, while later-borns have opposite characteristics. According to Sulloway, the effects of birth order are large and consequential. For example, he says that they are larger than the well-established gender differences in

2 personality. They lead later borns to exceptional prominence as innovators and rebels, both in the scientific and the political sphere. Sulloway even suggests large-scale demographic implications, e.g., that declining fertility leads to growing political conservatism by dint of the larger share of first-borns in the general population {WHA 1996}. Sulloway's confirmation of these claims is based primarily on a classification of many previous studies of birth order as either confirming, disconfirming, or neutral {! Sulloway 1996: 73}.1 His review offers almost no information about the size or importance of effects. Neither does the main source of his meta-analysis, the extensive and influential 1983 review by Ernst and Angst {! Ernst and Angst 1983}. The conclusions of that earlier review were essentially negative, but Sulloway claims concordance among the subset of studies in which social class and size of sibship had been controlled.2 Of the personality studies reviewed by Ernst and Angst, almost none were based upon samples of siblings. Mainly, they were based on persons from different families, classified by their birth order. Most were small studies of college students in which social class and size of sibship were rarely controlled. The handful of studies that focused on personality differences within actual sibships were all of young children and

1

One might think that "neutral" findings are disconfirming.

Other scholars have been unsuccessful in reproducing the findings from Sulloway's "meta-analysis" of Ernst and Angst's compendium of birth order studies {Harris 1998: 367-73}.

2

3 were based on personality ratings by teachers or parents. One might legitimately wonder whether such ratings are affected by received knowledge about the association between birth order and personality, or whether the behavior of children in the home is valid as an indicator of personality or behavior in adulthood {Harris 1998: 374-75}. Sulloway also offers striking evidence of birth order effects on positions taken in a series of historical conflicts and controversies. These analyses are based on a large database that Sulloway accumulated. This evidence, too, deserves close scrutiny, but it is not the subject of the present report. Effects in the general population that are too small to merit much interest could possibly be of greater importance among historic elites. Sibling Data from the Wisconsin Longitudinal Study The Wisconsin Longitudinal Study (WLS) is a 35-year study of a random sample of 10,317 men and women who graduated from Wisconsin high schools in 1957 and of their randomly selected brothers and sisters. Survey data collected from the original respondents or their parents in 1957, 1964, and 1975 provide a full record of social background, youthful aspirations, schooling, military service, family formation, labor market experiences, and social participation. Survey data have been supplemented by mental ability tests (of primary respondents and their siblings), measures of school performance, and characteristics of communities of residence, schools and colleges, employers, and industries. In 1977 parallel interview data were obtained from 2000 siblings of the primary respondents.

4 In 1992-93 the original sample of graduates was interviewed again by telephone and mail. In the 1992-93 surveys of the 1957 high school graduates, there were 8,493 telephone respondents, of whom 6,875 also completed the mail survey. At that time, the graduates were about 53 years old, and about 70 percent lived in Wisconsin. While the WLS is regional in origin, the sample is similar to the majority of Americans of their age. Approximately two-thirds of the population born around 1939 are white, nonHispanic high school graduates {Kominski and Adams 1992}. Content of the surveys included detailed occupational histories and job characteristics; incomes, assets, and inter-household transfers; social and economic characteristics of parents, siblings, and children and descriptions of the respondents' relationships with them; and extensive information about mental and physical health and well-being. In 1993-94 the sibling sample was expanded with content similar to that in the 1992-93 graduate surveys. At that time, most siblings were between 45 and 61 years old. Unlike the original graduates, the siblings were not selected explicitly with respect to age or educational attainment. Due to budget constraints, it was not possible to give the entire eligible sample of approximately 6700 cases the full telephone-mail treatment. We administered both telephone and mail to 80% of the eligible cases and mail only with a supplement containing key items that were on the telephone survey to the remaining 20%. We have data for 4804 siblings who completed the telephone survey only, for 4045 siblings who completed the mail survey, and for 5363 siblings who completed the telephone modules and the supplementary mail questionnaires for the

5 modules containing the variables for identification and sample information, education, employment, children, and marriage. Measuring Personality For reasons unrelated to research on birth order, we obtained standard measures of the Big Five personality characteristics in the samples of WLS graduates and siblings. We administered two items from each personality dimension in the telephone interviews and five or six items from each personality dimension in the mail surveys. These data provide a uniquely appropriate and powerful setting in which to estimate the persistent effects of birth order on personality in adulthood. The BFI-54 is a relatively short instrument that assesses the Five-Factor Model of Personality dimensions, or the "Big Five" {John 1990; John 1991}. The Extraversion (vs. Social Inhibition) scale captures gregarious, energetic, and expressive features of behavior. The Agreeableness (vs. Antagonism) scale reflects essentially pro-social characteristics, describing the person who is empathic and makes an effort to establish positive relationships with others. The Conscientiousness (vs. Lack of Direction) scale captures the multiple elements of persistence and impulse control in task and achievement settings. The Neuroticism (vs. Emotional Stability) scale reflects multiple elements of negative emotionality, such as nervous tension, fearfulness, and brittleness under stress. The Openness to Experience scale refers to persons who are imaginative, curious, creative, and susceptible to absorbing experience. Research by McCrae and Costa {!McCrae and Costa 1990} at NIA suggests that the five-factor model can yield a

6 replicable and comprehensive representation of the major dimensions of personality in adulthood. Administration of the BFI-54 in a large sample of adults has provided a unique, large-scale assessment of personality at midlife. Figure 1 shows some sample items from the BFI-54. In the telephone interview, each respondent was asked whether s/he agreed with the item and then whether agreement/disagreement was strong, moderate, or slight. In the mail survey, respondents rated their agreement with each item on a six-point numerical scale labelled from "agree strongly" to "disagree strongly." In our analyses, we have scored responses on a scale from 1 to 6, reversing the scoring as appropriate for negatively worded items. Reliability and Stability of Personality Measures Table 1 gives estimates of the reliability and stability of the personality variables in the WLS data. These are based on about 6500 WLS graduates who completed the 1975 survey and the 1992-93 telephone and mail surveys and on about 3500 of their brothers and sisters who completed the phone and mail surveys in 1993-94. For each personality dimension, Alpha reliabilities for the mail surveys are similar for graduates and siblings. They range from 0.580 to 0.777. These are not very high, but there are also no more than 6 items per dimension. In the telephone survey, where there are only two items per dimension, the alpha reliabilities are substantially lower, ranging from 0.366 to 0.617. Again, the estimates are similar for graduates and their siblings. The mail surveys were usually sent out about two weeks after completion of the telephone interview. Thus, correlations between the telephone and mail measures

7 provide a check on the short-term stability of personality. These correlations range from 0.46 for Conscientiousness to 0.69 for Extraversion. They are not very high, but that is simply a consequence of the reliability of the measures. When we correct the phone-mail correlations for internal consistency reliability (using alpha), we find that the correlations are very close to 1.00 for four of the five personality dimensions. The stability is lower in the case of Conscientiousness, for which the corrected correlations are 0.873 among graduates and 0.807 among siblings. We also looked at the resemblance in personality between graduates and their siblings. These correlations are modest, but they may be interpreted directly as estimates of the share of the variance in personality that lies between families. Thus, in the uncorrected measures, the mail surveys suggest that 10 to 18 percent of the variance in the personality dimensions lies between families. In most cases, the estimates from the telephone survey are lower, as we would expect from the lower reliability of the two-item measures. As shown in the last line of Table 1, correction for unreliability raises the estimates of sibling resemblance. Siblings share 13 to 20 percent of the variance in Agreeableness, Conscientiousness, and Neuroticism. The common variance in Extraversion is slightly larger, while that in Openness is impressively large. The two corrected estimates for Openness to Experience are both in excess of 30 percent. We shall see that there is far more evidence of sibling resemblance in personality than of birth order differences in personality among siblings.

8 Birth Order and Personality Among Individuals We looked for birth order differences in adult personality in two fundamentally different ways. First, we looked separately at the data from the telephone and mail surveys of graduates and at the mail survey of siblings, following the conventional practice of classifying respondents by their birth order and gender within each size of sibship. We limited these analyses to respondents from families (including a small number of non-biological siblings) in which all siblings survived to 1975 and for whom complete data were obtained. Second, we compared graduates with their brothers or sisters. In this section, we report the first set of analyses. Figure 2 shows box plots of the distribution of Agreeableness by gender and birth order (first-born vs. later born) for sibships of sizes 1 to 6 in the mail survey of graduates (n = 4990). First, we observe substantial overlap of the distributions across all of the dimensions of the classification. That is, neither gender, nor size of sibship, nor birth order has substantial explanatory power. Second, there is a clear gender differential in Agreeableness in every comparison within size of sibship. Women are more agreeable than men. Third, we see little evidence of a birth order effect on Agreeableness. The distributions are very similar for first-born and later-born siblings. However, where the distributions appear to differ slightly, there are more comparisons (5) in which the data show less Agreeableness than more Agreeableness (2) among firstborns. This is weak evidence indeed; in fact, not one of the 10 contrasts of mean

9 agreeableness between first-born and later-born graduates is significant at even the 10 percent probability level. Figures 3 to 6 provide similar evidence about the non-effects of birth order on the other four personality dimensions among WLS graduates. The details are much the same as in the case of Agreeableness. There is almost no visible evidence of birth order differences in personality. We carried out parallel analyses of the data from the telephone survey of graduates (n = 6041) and the mail survey of siblings (n = 3161). Again, we found no substantial or consistent evidence of birth order effects on personality. By way of summary, Figure 7 shows normal quantile plots of the test-statistics obtained by contrasting the personality of first-borns with that of later-borns in the three surveys. Here, in each sample, we contrasted mean personality levels within gender for family sizes from 2 to 10. We excluded Extraversion from the summary analyses because Sulloway reports inconsistent evidence, which he attributes to factorial complexity of the Extraversion dimension {! Sulloway 1996: 73-74}. Thus, within each survey, there are 18 tests for each personality dimension and 72 tests in all. All effects were signed so a positive test-statistic supported Sulloway's prediction. Figure 7 shows that the distributions of test statistics are close to random, but there are a few too many statistics in both the upper and lower tails of the distributions. That is, the small number of larger deviates tend as much to contradict as to support Sulloway's theories.

10 Birth Order and Personality Within Families The unique value of the WLS data lies in our ability to compare the personalities of graduates and their siblings, rather than merely to array the personality data by birth order. These comparisons require careful specification of family and individual factors affecting the personality measures. Let personality characteristic v in sibships of size t for individual w be :ijkvtw = "vt + :ivt + :jvt + :kvt + ,ijkvtw, where "vt is the intercept, :ivt is the effect of response status (r = graduate vs. s = sibling), :jvt is the effect of gender (m = male vs. f = female), :kvt is the effect of birth order (p = first born vs. l = later born), and ,ijkvtw is a stochastic error. Response status enters the model because the original respondents must have graduated from a Wisconsin high school in 1957, while siblings were chosen at random from the sibships of the graduates, without respect to their age or educational attainment. Without loss of generality, we suppress the notation for personality characteristics, sibship size, and individuals, so the model for mean personality characteristics is :ijk = " + :i + :j + :k. This mean structure permits effects of response status and gender to be confounded with those of birth order. In fact, it is not generally the case that just any comparison between siblings of first and higher order will yield an effect of birth order. Using this notation, we give selected examples of estimators of the effects of birth order, gender, and response status. In the analysis, we have also considered selected interaction (2) (1)

11 effects among birth order, gender, and response status. In this paper, reported estimates are pooled across all possible estimators of each effect. Birth Order Consider two sister pairs. In the first pair the first-born sister is the graduate, so the difference between mean personality measures is :rfp - :sfl = " + :r + :f + :p - (" + :s + :f + :l) = (:r - :s) + (:p - :l). (3)

In the second pair, the later-born sister is the graduate, so the difference between mean personality measures is :sfp - :rfl = " + :s + :f + :p - (" + :r + :f + :l) = (:s - :r) + (:p - :l), so we can estimate the effect of birth order by :p - :l = [(:rfp - :sfl ) + (:sfp - :rfl )]/2. We can also estimate birth order effects from brother pairs, and from combinations of male-female and female-male graduate-sibling pairs. Gender Consider two mixed-gender pairs. In the first, the graduate is a first-born male, so the mean personality difference is :rmp - :sfl = " + :r + :m + :p - (" + :s + :f + :l) = (:r - :s) + (:m - :f) + (:p - :l), (6) (5) (4)

12 and in the second pair, the graduate is a first-born female, so the mean personality difference is :rfp - :sml = " + :r + :f + :p - (" + :s + :m + :l) = (:r - :s) + (:f - :m) + (:p - :l). Thus, the obvious estimator is :m - :f = [(:rmp - :sfl ) - (:rfp - :sml)]/2. We can obtain separate or pooled estimates of the gender effect from same-gender or mixed-gender first-born, later-born pairs and from pairs in which both siblings are later-born.3 Respondent Status Consider two pairs of sisters, in which the first-born member of the first pair is a graduate, and the first-born member of the second pair is a sibling. The mean structure is given in equations 3 and 4, above. We can estimate the effect of respondent status by :r - :s = [(:rfp - :sfl ) - (:sfp - :rfl )]/2. (9) (8) (7)

Note that we could estimate :r - :s by taking the difference between mean personality measures of the first member of each pair or by taking the difference between mean personality measures of the second member of each pair. However, either of those estimators potentially confounds the effect of respondent status with other family characteristics, while the estimator in (9) is a difference of within-family differences. We

Standard errors of estimates from these models have been obtained using standard formulas for the variances of linear composites.

3

13 can obtain similar, within-family estimators in first-born, later-born male-male pairs and pairs of mixed gender, as well as in pairs of siblings in which both are later-born. Findings Within Families Table 2 shows within-family estimates of gender effects on personality from the mail surveys of the WLS. For each personality dimension, we report the estimates separately for sibships of sizes two to six or more and, also, a pooled within-family estimate for all sizes of sibship. The estimates control for possible effects of birth order and response status. Recall that, by construction, the personality measures each range from 0 to 30. The WLS data provide consistent evidence of gender differences in three personality dimensions: Agreeableness, Extraversion, and Neuroticism. Women are more agreeable, extraverted, and neurotic than their brothers. This accords with standard findings in personality research. Table 3 shows within-family estimates of the effects of respondent status on each personality measure. The estimates control for possible effects of birth order and gender. There is no compelling theoretical rationale for effects of the contrast between graduates and their siblings. However, the sampling design differs between them. All of the original respondents graduated from Wisconsin high schools in 1957, while there is no restriction with respect to age or graduation status in the case of siblings. In fact, there are differences between graduates and their siblings in every dimension of personality. The graduates are more agreeable, conscientious, extraverted, and open,

14 and they are less neurotic than their siblings. These findings validate our decision to control respondent status in estimating effects of birth order. Table 4 shows within-family estimates of the effects of birth order on each personality measure in the WLS mail surveys. The estimates control for possible effects of gender and respondent status. To be brief, there is essentially no evidence that birth order is associated with adult personality. There is not a single significant contrast between first-born and later-born siblings within any size of sibship.4 In the pooled tests, one coefficient reaches borderline statistical significance in the case of Conscientiousness (t = 1.86). This effect lies in the direction predicted by Sulloway. Given the large number of sibling pairs (n = 2034) and the large number of statistical tests reported in Table 4, we think this one finding lends little support to Sulloway's theory. For example, Raftery {! Raftery 1995: 140} shows that, in a sample of 1000, one would need a t-statistic of at least 2.63 even to provide weak evidence in favor of the alternative hypothesis. Summary and Conclusion Frank J. Sulloway's Born to Rebel offers strong and far-reaching hypotheses about the role of birth order in personality and in social change. His theories rest on the assumption that birth order has strong, consistent, and persistent effects on personality. We have tested those assumptions using data on adult personality from a large sample

The findings are similar among male-male, female-female, and mixed-gender pairs of siblings. In detailed analyses by type of pair, the only effect that is even marginally significant occurs in the case of Conscientiousness in male-male pairs.

4

15 of midlife adults and their siblings, the Wisconsin Longitudinal Study. The personality measures -- Agreeableness, Conscientiousness, Extraversion, Neuroticism, and Openness to Experience -- have been measured using a standard psychological scale, John's BFI-54, and -- considering the reduced number of items in the WLS surveys -they display acceptable reliability and stability. The WLS personality data also reproduce standard findings about personality differences between women and men. We have looked for association between birth order and personality in two different ways. First, we classified respondents by birth order, gender, and family size and looked for net effects of birth order on personality. We carried out similar analyses using data from the WLS graduate mail and telephone surveys and from the sibling mail survey. There were no visible differences by birth order in the distributions of personality variables within gender and size of sibship. In dozens and dozens of statistical tests, birth order differences in personality appeared to be essentially random. Second, we constructed pairs of graduates and their brothers or sisters, and we looked for birth order differences in personality within families. Again, we found essentially no evidence that sibling differences in personality are associated with birth order. Thus, we find no support for the psychological underpinnings of Frank Sulloway's theories. The WLS data do show a highly significant sibling effect on personality, namely, that brothers or sisters -- regardless of birth order -- are moderately similar in personality to one another. Siblings share 13 to 20 percent of the variance in Agreeableness, Conscientiousness, and Neuroticism. The common variance in

16 Extraversion is slightly larger, while that in Openness is impressively large. Reliabilitycorrected estimates for Openness to Experience are both in excess of 30 percent. We think that future research into the sources of sibling resemblance in personality will be far more useful scientifically than further elaborations of theories about birth order.

Table 1. Reliability and Stability of Personality Variables: Estimates from the Wisconsin Longitudinal Study Size of sibship Reliability (alpha): Graduates Mail Phone Reliability (alpha): Siblings Mail Phone Phone-mail correlation Graduates Siblings Phone-mail correlation (corrected) Graduates Siblings Sibling correlation Mail Phone Sibling correlation (corrected) Graduates Siblings 0.110 0.071 0.122 0.073 0.146 0.159 0.105 0.079 0.181 0.167 Agreeableness Conscientiousness Extraversion Neuroticism Openness

0.686 0.387

0.638 0.448

0.763 0.596

0.777 0.617

0.608 0.408

0.686 0.369

0.664 0.490

0.762 0.580

0.753 0.612

0.580 0.366

0.520 0.485

0.467 0.460

0.690 0.696

0.685 0.687

0.508 0.502

1.008 0.963

0.873 0.807

1.023 1.047

0.990 1.012

1.019 1.090

0.160 0.187

0.187 0.156

0.191 0.270

0.138 0.128

0.305 0.433

Note: Estimates are based on data from about 6500 graduates and about 3500 siblings.

Table 2. Effects of Gender on Personality: Within-Family Estimates from the Wisconsin Longitudinal Study Size of sibship Two n 819 Estimate t Estimate t Estimate t Estimate t Estimate t Estimate t Agreeableness -1.76 -7.99 -1.77 -8.08 -2.04 -7.44 -1.52 -4.52 -1.93 -6.28 -1.81 -14.22 Conscientiousness 0.10 0.43 -0.05 -0.24 0.21 0.82 0.53 1.61 -0.52 -1.65 0.05 0.40 Extraversion -0.94 -3.17 -0.23 -0.80 -0.79 -2.48 -0.60 -1.35 -0.89 -2.58 -0.67 -4.27 Neuroticism -1.81 -5.72 -1.36 -4.58 -1.97 -5.62 -1.76 -3.79 -0.81 -1.95 -1.52 -8.86 Openness -0.35 -1.42 0.34 1.38 -0.29 -1.04 -0.32 -0.76 -0.24 -0.78 -0.16 -1.12

Three

739

Four

559

Five

374

Six or more

651

All

3142

Note: Estimates are based on comparisons of graduates and siblings in which respondent status and birth order are controlled.

Table 3. Effects of Respondent Status on Personality: Within-Family Estimates from the Wisconsin Longitudinal Study Size of sibship Two n 819 Estimate t Estimate t Estimate t Estimate t Estimate t Estimate t Agreeableness 0.26 1.31 0.16 0.74 1.07 4.00 0.65 1.97 0.65 2.31 0.59 4.86 Conscientiousness 0.55 2.72 0.73 3.64 0.63 2.69 0.58 1.84 0.36 1.31 0.58 4.98 Extraversion 0.45 1.67 0.65 2.51 0.42 1.40 0.14 0.35 0.70 2.22 0.47 3.32 Neuroticism -0.03 -0.11 -0.19 -0.68 -0.64 -1.86 -0.54 -1.15 -0.06 -0.15 -0.32 -1.94 Openness 0.53 2.35 0.31 1.40 0.62 2.36 0.04 0.12 0.16 0.56 0.31 2.48

Three

739

Four

559

Five

374

Six or more

651

All

3142

Note: Estimates are based on comparisons of graduates and siblings in which birth order and gender are controlled.

Table 4. Effects of Birth Order on Personality: Within-Family Estimates from the Wisconsin Longitudinal Study Size of sibship Two n 819 Estimate t Estimate t Estimate t Estimate t Estimate t Estimate t Agreeableness -0.14 -0.67 0.32 1.27 0.04 0.11 0.04 0.09 -0.51 -1.29 -0.05 -0.32 onscientiousness -0.11 -0.55 0.31 1.24 0.41 1.35 0.21 0.49 0.54 1.39 0.27 1.86 Extraversion -0.22 -0.80 0.15 0.48 -0.17 -0.43 0.60 1.11 -0.16 -0.37 0.04 0.23 Neuroticism -0.32 -1.16 -0.37 -1.11 -0.40 -0.89 -0.20 -0.31 0.68 1.27 -0.12 -0.58 Openness -0.02 -0.10 0.04 0.13 -0.17 -0.47 0.41 0.84 0.49 1.25 0.15 0.93

Three

521

Four

319

Five

162

Six or more

213

All

2034

Note: Estimates are based on comparisons of graduates and siblings in which respondent status and gender are controlled.

Figure 1. Sample Items from the Big Five Personality Inventory (BFI-54) ___________________________________________________________________________ Agreeableness I see myself as someone who is considerate to almost everyone Conscientiousness I see myself as someone who does things efficiently. Extraversion I see myself as someone who is talkative. Neuroticism I see myself as someone who worries a lot. Openness to Experience I see myself as someone who values artistic, aesthetic experiences. ___________________________________________________________________________

Figure 2. Agreeableness by Birth Order, Gender, and Size of Sibship: Wisconsin Graduates

Figure 3. Conscientiousness by Birth Order, Gender, and Size of Sibship: Wisconsin Graduates

Figure 4. Extraversion by Birth Order, Gender, and Size of Sibship: Wisconsin Graduates

Figure 5. Neuroticism by Birth Order, Gender, and Size of Sibship: Wisconsin Graduates

Figure 6. Openness by Birth Order, Gender, and Size of Sibship: Wisconsin Graduates

Figure 7 Quantiles of t-Statistics in Tests of Birth Order Effects on Personality WLS Graduates and Siblings (Mail and Telephone Surveys)

1.0 Lower tail probability of t-value 0.8 0.6 0.4 0.2 0.0 0.0

0.2

0.4 0.6 Percentile rank of probabilty

0.8

1.0

Information

29 pages

Find more like this

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

693813


You might also be interested in

BETA
Microsoft Word - mini research paper.doc
paulhus.qxd
FrontMatter.qxd
doi:10.1016/S1090-5138(03)00033-3