Initial Research Focus and Data Collection From 1993 to 1997
As part of a joint twin research project in Germany and Poland (Oniszczenko et al., Reference Oniszczenko, Zawadzki, Strelau, Riemann, Angleitner and Spinath2003; Riemann et al., Reference Riemann, Angleitner and Strelau1997; Zawadzki et al., Reference Zawadzki, Strelau, Oniszczenko, Riemann and Angleitner2001), the Bielefeld Longitudinal Study of Adult Twins (BiLSAT) initially focused on the genetic and environmental contributions to the variance in temperament and personality traits as well as to the links among these traits. Twin pairs were recruited through several media announcements (e.g., newspapers, TV, and radio) and twin clubs in Germany. Interested twins could call an installed telephone hotline. Names, addresses, and date of birth of approximately 1,500 German twin pairs (mostly from the region around Bielefeld) who decided to participate were registered between 1993 and 1997. A total of 2,404 twins returned a complete set of questionnaires from the first mailing, and 1,765 twins completed a second set (see Table 1). The sets of inventories included a demographic questionnaire, several temperament and personality questionnaires (see Table 2), and a self-report questionnaire on physical resemblance and twins’ medical history determining zygosity. The sample was heterogeneous with regard to education and employment status. For more detail of recruitment and descriptions of the initial sample, see a previous report (Spinath et al., Reference Spinath, Angleitner, Borkenau, Riemann and Wolf2002).
MZ: monozygotic; DZ: dizygotic.
aNumber of twin pairs with complete data.
bDZ twins included same-sex and opposite-sex pairs.
Self: self-reports; Peer: peer reports; PTS: Pavlovian Temperament Survey (Strumpf et al., Reference Strumpf, Angleitner and Newberry1999); FCB-TI: Formal Characteristics of Behavior — Temperament Inventory (Strelau & Zawadski, Reference Strelau and Zawadski1993, Reference Strelau and Zawadski1995); NEO-FFI and NEO-PI-R: Neuroticism–Extraversion–Openness Five-Factor Inventory and Personality Inventory Revised (Borkenau & Ostendorf, Reference Borkenau and Ostendorf1993; Costa & McCrae, Reference Costa and McCrae1992; Ostendorf & Angleitner, Reference Ostendorf and Angleitner2004); EAS: Emotionality–Activity–Sociability temperament survey for adults (Buss & Plomin, Reference Buss and Plomin1984); EPQ-RS: Eysenck's Personality Questionnaire Revised Short form (Eysenck & Eysenck, Reference Eysenck and Eysenck1991; Ruch, Reference Ruch1999); DOTS-R: Revised Dimensions of Temperament Survey (Windle & Lerner, Reference Windle and Lerner1986); BIPOL and UNIPOL: bipolar and unipolar adjective scales (Ostendorf, Reference Ostendorf1990); MPQ: Multidimensional Personality Questionnaire (Tellegen & Waller, Reference Tellegen, Waller, Boyle, Matthews and Saklofske2008); ZKPQ-III-R: Revised Zuckerman-Kuhlman-Personality-Questionnaire (Ostendorf & Angleitner, Reference Ostendorf and Angleitner1994; Zuckerman, Reference Zuckerman, De Raad and Perugini2002); ANPS: Affective Neuroscience Personality Scales (Davis et al., Reference Davis, Panksepp and Normansell2003; Reuter et al., Reference Reuter, Panksepp, Schnabel, Kellerhoff, Kempel and Hennig2005); DAPP-BQ: Dimensional Assessment of Personality Pathology Basic Questionnaire (Livesley & Jackson, Reference Livesley and Jackson2009); PANAS: Positive Affect Negative Affect Scale (Krohne et al., Reference Krohne, Egloff, Kohlmann and Tausch1996; Watson et al., Reference Watson, Clark and Tellegen1988); GIS: General Interest Scale (Brickenkamp, Reference Brickenkamp1990); GOALS: major life goals (Pöhlmann & Brunstein, Reference Pöhlmann and Brunstein1997); RVS: Rokeach Values Survey (Johnston, Reference Johnston1995; Rokeach, Reference Rokeach1973); AVQ: Austrian Value Questionnaire (Renner, Reference Renner2003); SWLS: Satisfaction With Life Scale (Diener et al., Reference Diener, Emmons, Larsen and Griffin1985; Sölva et al., Reference Sölva, Baumann and Lettner1995).
aNEO-FFI items of waves 2, 3, and 4 were included in the NEO-PI-R.
bOnly the items capturing the importance of major life goals (24 items) were included in the fourth wave.
cUnpublished German versions were translated and developed by the Bielefeld research group.
A special feature of BiLSAT is the collection of reports from well-informed peers (specific for each twin) in addition to self-reports from twins. In the first wave, almost 4,800 peers (mostly friends, relatives, spouses, and colleagues) provided ratings. This multiple-rater strategy allowed addressing several methodological problems with single-rater assessments (e.g., response styles, social desirability, self-deception, and contrast effects). As shown in the list of inventories used as first-person and third-person versions (see Table 2), the initial two waves of the study provided a rich, genetically informative multiple-rater data set based on different measures of temperament and personality. These data contributed substantially to personality research and helped to answer different specific research questions. For example, more accurate measures that were based on the combination of peer-report and self-report data (controlling for random error variance and variance in self-rater biases) indicated higher heritability of temperamental and personality traits (e.g., Kandler, Reference Kandler2012; Riemann et al., Reference Riemann, Angleitner and Strelau1997; Wolf et al., Reference Wolf, Angleitner, Spinath, Riemann and Strelau2004; Zawadski et al., Reference Zawadzki, Strelau, Oniszczenko, Riemann and Angleitner2001). Moreover, examining the twin covariance among personality traits revealed the underlying genetic and environmental basis of hierarchical personality trait structures (e.g., Jang et al., Reference Jang, McCrae, Angleitner, Riemann and Livesley1998, Reference Jang, Livesley, Angleitner, Riemann and Vernon2002; McCrae et al., Reference McCrae, Jang, Livesley, Riemann and Angleitner2001).
Further Research Aims and Data Collection From 1999 to 2009
About 5 years after the twins and their peers completed the second set of questionnaires, a third set of questionnaires was sent out to the participants. Between 1999 and 2002, a total of 844 individuals, including 338 complete twin pairs and over 1,600 peer raters, provided data at this third measurement wave (see Table 1). Besides new measures, this set of questionnaires also included different Big Five measures a second time (see Table 2) in order to focus on the genetic and environmental sources of stability and change in personality. Additionally, the third wave included measures of major life goals, attitudes, interests, values, personality disorders, and life events, realizing a third aim of the study that addresses the genetically and environmentally mediated links between personality traits and other individual attributes, as well as specific environments.
All individuals who were still enrolled in the study were sent a fourth set of questionnaires about 5 years after they took part in the third measurement wave. That is, the fourth wave ran from 2005 to 2008. A total of 433 individuals, including 202 complete twin pairs and over 800 peer raters, provided data at this wave. The fourth set of questionnaires again included a broad measure of Big Five personality traits and personality facets (NEO-PI-R, see Table 2), allowing genetically informative longitudinal analyses across three waves of assessment over a time span of 10 years. In addition, this set involved further follow-up measures of major life goals and life events, allowing for longitudinal, genetically informative analyses of these variables as well as analyses of interdependencies among genetic and environmental sources of personality, goals, and life events over time.
One year after the fourth wave and 13 years after the study's initiation, we contacted each twin who participated at the first wave for a final participation between 2006 and 2009. A total of 614 individuals, including 262 complete twin pairs and over 1,100 associated peer raters, completed the fifth set of questionnaires focusing on temperamental and personality traits. The inclusion of these data in BiLSAT allows longitudinal analyses of self- and peer reports on twins’ temperament and personality in adulthood over a period of almost 13 years. The combination of longitudinal, genetically informative, and multiple-rater data features BiLSAT as unique twin study worldwide.
Recent Major Findings
By now, a large number of published studies based on BiLSAT data exist. Some of the major findings were already presented in a previous report (Spinath et al., Reference Spinath, Angleitner, Borkenau, Riemann and Wolf2002). In this article, we focus on recent findings and analyses that have taken advantage of the multi-methodical and longitudinal features of BiLSAT. These major findings comprise (1) the underlying genetic basis of the hierarchical structure of personality, (2) the genetic and environmental stability and change in personality traits, and (3) the nature of the interrelations between personality traits and other individual characteristics or individual environments.
Multiple-rater data of twins’ Big Five personality traits and facets from the third wave have been combined with personality data from other twin samples of different nations, such as Japan, Canada, or the United States. This combination showed cross-cultural evidence that genetic factors are more reflective of the phenotypic hierarchical structure of the Five-Factor Model of personality than are environmental factors (Jang et al., Reference Jang, Livesley, Ando, Yamagata, Suzuki, Angleitner, Ostendorf, Riemann and Spinath2006; Yamagata et al., Reference Yamagata, Suzuki, Ando, Ono, Kijima, Yoshimura, Ostendorf, Angleitner, Riemann, Spinath, Livesley and Jang2006). Furthermore, the self- and peer-report data of twins’ personality have provided multi-methodical evidence for a combination of a bottom-up and a top-down model of genetic influences that appears to characterize the structure of each of the five broad personality dimensions and their corresponding facets (Kandler et al., Reference Kandler, Bleidorn, Riemann, Spinath, Thiel and Angleitner2010), whereas higher-order factors, such as the Big Two or a general factor of personality, mainly reflect artifacts (McCrae et al., Reference McCrae, Yamagata, Jang, Riemann, Ando, Ono, Angleitner and Spinath2008; Riemann & Kandler, Reference Riemann and Kandler2010).
Beyond the structural consistency across cultures and methods of measurement, the longitudinal nature of BiLSAT allows a focus on genetic and environmental influences on stability and change in personality and other traits across adulthood. Bleidorn et al. (Reference Bleidorn, Kandler, Riemann, Angleitner and Spinath2009) focused on mean-level trends and individual differences in change using a genetically informative growth curve modeling approach across the second, third, and forth waves of assessment. They found pronounced genetic contributions to individual differences in change in personality traits (measured with the NEO-PI-R) that showed large mean-level trends (i.e., Neuroticism, Agreeableness, and Conscientiousness), but more pronounced environmental effects on individual differences in individual-level trends in traits with only small or no significant mean-level changes (i.e., Openness and Extraversion). This finding indicates not only a genetic contribution to personality maturation but also environmental effects on individual-specific developmental trends.
Based on both self- and peer-report personality data, studies showed that genetic factors represent the main sources contributing to phenotypic rank-order and profile stability of personality in middle adulthood, whereas rank-order change and individual differences in personality profile change were predominantly attributable to environmental factors (Bleidorn et al., Reference Bleidorn, Kandler, Riemann, Angleitner and Spinath2012; Kandler et al., Reference Kandler, Riemann, Spinath and Angleitner2010). Moreover, it was shown that phenotypic rank-order continuity in Big Five personality traits increased as a function of stabilizing environmental trait variance and decreasing environmental occasion-specific effects across young and middle adulthood (Kandler et al., Reference Kandler, Bleidorn, Riemann, Spinath, Thiel and Angleitner2010). This indicates cumulative and stabilizing environmental effects on personality in this period of life. On that note, recent longitudinal analyses provide support for a bidirectional model of gene–environment interplay between personality traits and life events (Kandler et al., Reference Kandler, Bleidorn, Riemann, Angleitner and Spinath2012). More specifically, genetic variance in Extraversion and Openness accounted for genetic variance in positive events, whereas genetic variance in Neuroticism, Openness, and Agreeableness explained genetic variance in negative events. Controlling for these genetic factors between personality traits and life events, there were very small but significant effects from negative life events on Neuroticism, Openness, and Agreeableness accounting for a small proportion of environmental variance in these personality traits.
The third major field of research using BiLSAT data is the examination of the nature of the links between Big Five personality traits and other individual attributes. Until now, researchers have focused on the genetic and environmental links between personality and motivational variables, such as major life goals (i.e., agency and communion; Bleidorn et al., Reference Bleidorn, Kandler, Hülsheger, Riemann, Angleitner and Spinath2010) or general interests (e.g., artistic and social; Kandler et al., Reference Kandler, Bleidorn, Riemann, Angleitner and Spinath2011). These studies indicated that personality traits are systematically and primarily genetically linked to individual motivations. However, large proportions of genetic variance in goals and interests could not be accounted for by genetic factors on personality traits, providing support for the hypothesis that Big Five personality traits and motivations are distinct characteristics to describe individual differences in behavior.
Future Plans
We are currently planning to provide BiLSAT data as open source. As presented above, BiLSAT data have contributed extensively to behavior genetic research on personality and related variables. Open questions concern, for example, the nature and nurture of political attitudes and collectivistic versus individualistic values and the role of personality traits. The preparation of BiLSAT data as open source enables international collaborations and offers researchers to use the data in order to answer their own specific research questions.