Personality and Individual Dierences 42 (2007) 13591366

www.elsevier.com/locate/paid

Reliability and validity of the Danish version of the UCLA Loneliness Scale
Mathias Lasgaard
*

Department of Psychology, University of Aarhus, Jens Chr. Skous Vej 4, 8000 Aarhus C, Denmark Received 19 May 2006; received in revised form 25 September 2006; accepted 5 October 2006 Available online 28 November 2006

Abstract The objective of this study was to examine the psychometric properties of a Danish version of the UCLA Loneliness Scale (UCLA). The 20-item scale was completed along with other measures in a national youth probability sample of 379 8th grade students aged 1317. The scale showed high internal consistency, and correlations between UCLA and measures of emotional loneliness, social loneliness, self-esteem, depression, extraversion, and neuroticism supported the convergent and discriminant validity of the scale. Exploratory factor analysis supported a unidimensional structure of the measure. The results, highly comparable to the original version of the scale, indicate that the Danish version of UCLA is a reliable and valid measure of loneliness. 2006 Elsevier Ltd. All rights reserved.
Keywords: Loneliness; UCLA; Rating scales; Adolescence; Reliability; Validity; Factor analysis

1. Introduction Loneliness is a serious problem among adolescence (Jones & Carver, 1991; West, Kellner, & Morre-West, 1986) and has been associated with low self-esteem (Brage, Meredith, & Woodward, 1993; Schultz & Moore, 1988), depression (Koenig, Isaacs, & Schwartz, 1994; Lau, Chan, & Lau, 1999), anxiety (Johnson, LaVoie, Spenceri, & Mahoney-Wernli, 2001; Moore & Schultz, 1983),
*

Tel.: +45 8942 4967; fax: +45 8942 4901. E-mail address: mathiasl@psy.au.dk

0191-8869/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2006.10.013

1360

M. Lasgaard / Personality and Individual Dierences 42 (2007) 13591366

anorexia nervosa (Troop & Bifulco, 2002), and suicide ideation and behaviour (Garnefski, Diekstra, & de Heus, 1992; Roberts, Roberts, & Chen, 1998). However, very little research in loneliness has been done in adolescent populations in Denmark, which may be due to the lack of validated versions of well-established measures. The present article examines a Danish version of the UCLA Loneliness Scale (UCLA) (Russell, 1996); the most widely used self-report scale for measuring loneliness in adolescent and adult populations (Hartshorne, 1993; Russell, 1996). The original scale (Russell, Peplau, & Ferguson, 1978), based on statements used by individuals to describe feelings of loneliness, suered from methodological weaknesses. However, the scale was revised in a second version (Russell, Peplau, & Cutrona, 1980), and a third version (Russell, 1996). The two revised scales are easy to administer and have satisfactory psychometric properties, including high internal consistency, typically a P .89 in adolescent populations (e.g. Koenig et al., 1994; Mahon, Yarcheski, & Yarcheski, 2004), and good test-retest reliability (e.g. Hartshorne, 1993; Russell, 1996). Moreover, Russell (1996) and Russell et al. (1980) have reported evidence of convergent validity, indicated by strong correlations between UCLA and other measures of loneliness, and discriminate validity, e.g. supported by conrmatory factor analysis indicating that UCLA and measures of social support dene distinct factors. The developers of UCLA consider the scale to be a unidimensional measure of loneliness (Russell, 1996). However, exploratory factor analyses of the Revised UCLA (second version) have yielded a number of dierent structures, on which basis researchers have argued that the scale comprises one (Pretoirus, 1993), two (e.g. Knight, Chisholm, Marsh, & Godfrey, 1988; Mahon, Yarcheski, & Yarcheski, 1995), three (e.g. Austin, 1983; McWhirter, 1990), four (Hojat, 1982) or ve (e.g. Hojat, 1982; Neto, 1992) factors. As stressed by Russell (1996), some of the derived factors reect the direction of items, indicating that ratings are aected by acquiescence or other similar response styles. Moreover, conrmatory analyses have supported the unidimensionality of the scale (Hartshorne, 1993; Russell, 1996). Evidently, the factorial structure of UCLA remains rather controversial. UCLA has been used and validated in many dierent countries, including Argentina (Sacchi & ce, Joshi, & Pelletier, 1993), Germany (Do Richaud de Minzi, 1997), Canada (de Gra ring & Bortz, 1993), Greece (Anderson & Malikiosi-Loizos, 1992), Iran (Hojat, 1982), Portugal (Neto, 1992), Russia (Ruchkin, Eisemann, and Ha gglof, 1999), South Africa (Pretoirus, 1993), Turkey (Uruk & Demir, 2003), and Zimbabwe (Wilson, Cutts, Lees, Mapungwana, & Maunganidze, 1992). All studies found the adapted scale a sucient measure of loneliness. The goal of this study was to standardize a Danish measure of loneliness. To do so, we aimed to examine the validity and reliability of a Danish version of UCLA in a representative adolescent population by testing the internal consistency, the convergent and discriminant validity, and the factorial structure of the scale.

2. Method 2.1. Subjects The data in this study were collected from a questionnaire survey with a national probability sample of 379 youngsters aged 1317 (M = 14.1 years, SD = 0.4). The sample was geographically stratied and 39 randomly selected schools that taught 8th grade students were approached with

M. Lasgaard / Personality and Individual Dierences 42 (2007) 13591366

1361

the purpose to recruit one randomly selected class from each school. Twenty-two schools agreed to participate. Each class consisted of between 6 and 24 pupils (M = 17.2 pupils, SD = 4.0), and an average of 90% of the students were present on the day of the study. All students present participated in the study. 2.2. Procedures The study was introduced through a letter to the headmaster of the selected schools, explaining the aim of the study, and the selection procedure in case there was more than one 8th grade class in the school (the primacy of the initials of the class teacher decided the class). A letter to the class teacher explained the aim of the study and described the procedure of the data collection, the need of introduction, monitoring and support, and the condentiality procedures (e.g. sealing the return envelope in front of the pupils). Finally, a letter to each student explained the purpose, the condentiality, and the collection procedure of the study. 3. Measures The UCLA Loneliness Scale (third version) (Russell, 1996) was translated from the original English version into Danish by the author and back translated by a person holding a Master of Arts. The original scale was compared with the back-translated scale and non-uniform items discussed, which led to minor revisions of the Danish translation. A professor of clinical psychology and four psychology students at master level then reviewed the translated UCLA, also causing minor revisions. Finally, the measure was tested in a pilot study with six 8th grade students, who conrmed that the written instruction, items, and rating system were understandable. The scale consists of 20 items (11 positive and 9 negative), describing subjective feelings of loneliness, none of which refers specically to loneliness. Consequently, the scale does not directly measure states that laypeople attribute as loneliness, but rather the scale measures a theoretically dened and scientically validated understanding of loneliness. The 20 items are rated on a 4point Likert scale in accordance with the rate of frequency, ranging from never (1) to always (4). Scores on the scale range from 20 to 80 with higher scores reecting greater loneliness. With the purpose to test for convergent validity, the Peer Network and Dyadic Loneliness Scale (PNDLS) (Hoza, Bukowski, & Beery, 2000) was included. PNDLS is a new 2 8-item child and adolescence Loneliness Scale, consisting of two subscales: (1) Peer Network Loneliness Scale (PNLS) that measures social loneliness caused by the lack of involvement in a social network, and (2) Peer Dyadic Loneliness Scale (PDLS) that measures emotional loneliness caused by the absence of a close dyadic friendship. The measure was translated and revised using the same procedure as with UCLA. The two subscales were scored on a 4-point Likert scale and showed good internal consistency in the study (PNLS: a = .83; PDLS: a = .77). Since strong correlations have been found between loneliness and measures of personality and depression, it is relevant to investigate the discriminant validity of UCLA by comparing the magnitude of correlations between UCLA and such measures with the correlations between UCLA and other measures of loneliness (Russell, 1996). For that reason Rosenberg Self-Esteem Scale (SES) (Rosenberg, 1965) was used to measure self-esteem. The scale comprises 10-items, scored on a 4-point Likert scale, and has been translated into Danish by Prof. Elklit (Department of

1362

M. Lasgaard / Personality and Individual Dierences 42 (2007) 13591366

Psychology, University of Aarhus). SES showed good internal consistency in the present study (a = .88). Moreover, the subscales Extraversion and Neuroticism from the Eysenck Personality Questionnaire Revised Short Scale (EPQ-RSS) (Eysenck & Eysenck, 1991) were included to measure two well-established personality traits. Each subscale comprises 12 statements, scored on a 2point yes/no scale. The two subscales were translated and revised using the same procedure as with UCLA. Both subscales showed good internal consistency in the study (both scales a = .85). Beck Depression Inventory for Youth (BDI-Y) (Beck, Beck, & Jolley, 2004), validated in a Danish sample by Thastum, Paulsen, and Hansen (2004), was used to measure depression. The measure comprises 20 symptoms of depression, scored on a 4-point Likert scale in accordance with the degree of occurrence, and had a high internal consistency in the study (a = .94).

4. Results 4.1. Sample characteristics The great majority of the students were 14 years old, so analyses of age are based on three groups (students < 14 years (n = 19); students = 14 years (n = 315); students > 14 years (n = 43)); two students did not report their age. The gender distribution was 53% male and 47% female. Ninety-four percent of the pupils were born in Denmark, while 6% were born in other countries. Seventy-one percent of the students lived with both parents, 28% lived with one parent, and 1% lived with others. Six percent of the pupils were an only child, 41% had one sibling, and 53% had two or more siblings. Twenty-six percent of the students lived in rural areas (i.e. in the countryside but not in a village), 24% lived in a village/small town (610,000 inhabitants), and 50% of the students lived in a city (10,0001,500,000 inhabitants). The reported characteristics resemble general gures of Danish 8th grade students (cf. Lasgaard, 2006). Moreover, the sample had a distribution of public and private schools similar to the general distribution of schools. In summary, the study has a high probability of being representative of all Danish 8th grade students. 4.2. Descriptive statistics The average score for UCLA total was 37.46 (SD = 10.38; range: 2075), which is slightly above the average score of 3.283 adolescents (M = 15.1), based on data from 21 studies (summarized by Lasgaard, 2006). The mean, median (36) and modes (28, 34) were fairly similar, indicating that the distribution of scores was quite normal. However, the scores were slightly positively skewed (.72, p < .05), suggesting that rather few students scored highly on the scale. Notably, a similar pattern was reported in four studies using the original version of the scale (Russell, 1996). No signicant dierences in loneliness scores were found in relation to age, gender, country of birth, living arrangements, number of siblings, and residential location. 4.3. Reliability and validity The internal consistency of the scale was high with a Cronbachs Alpha of .92, which is comparable to results of the original scale where Cronbachs Alpha ranged from .89 to .94 across four

M. Lasgaard / Personality and Individual Dierences 42 (2007) 13591366 Table 1 Correlation matrixa Scale UCLA PNLS PDLS SES BDI-Y EPQ-RSS-E
a

1363

PNLS .73

PDLS .69 .69

SES .58 .55 .41

BDI-Y .59 .47 .38 .66

EPQ-RSS-E .57 .57 .47 .30 .28

EPQ-RSS-N .58 .47 .38 .59 .73 .41

All ps < .0005.

dierent samples (Russell, 1996). Table 1 shows the correlations between UCLA and the other included measures. As expected, and in accordance with studies including the original scale (Russell, 1996), UCLA loneliness scores were highly correlated with the two subscales of the other included loneliness measure, PNLS (r = .73, p < .0005), and PDLS (r = .69, p < .0005). This result supports the convergent validity of the scale. Moreover, UCLA was highly related to the measures of self-esteem (r = .58, p < .0005), depression (r = .59, p < .0005), and the personality traits of extraversion (r = .57, p < .0005) and neuroticism (r = 58, p < .0005). However, the magnitudes of these correlations were evidently less than the correlations found between UCLA and the two subscales of PNDLS. Also the magnitudes of the correlations were less than those reported between the original UCLA and two other measures of loneliness (r = .65; 72, p < .001) (Russell, 1996). These ndings support the discriminant validity of the adapted scale. 4.4. Factor analysis The structure of UCLA was investigated by use of factor analysis. An exploratory analysis with direct oblimin rotation (loading criteria P .30) was performed to determine the optimal factor structure for this sample. Criteria for identifying the factors were based on (1) Glorfelds (1995) version of parallel analysis with a sample size of N = 379 and k = 20 variables (the principal components must be greater than 1.505 eigenvalues for the rst component, 1.410 for the second component, 1.341 for the third component, 1.280 for the fourth component, 1.230 for the fth component, and 1.187 for the sixth component, using the 95th percentile and 5000 replications), (2) the minimum average partials method (Velicer, 1976) using available syntax (OConnor, 2000), and (3) a scree plot. A two-factor solution explaining 50.0% of the variance meeting the three criteria was identied. The eigenvalue was 8.018 for the rst factor and 1.990 for the second factor. The solution comprised one factor with 19 items, and a second factor with nine items, including nine double loadings. One item (#17) did not load on any factor, which is why the factor analysis was repeated without this item. Using the same criteria for identifying factors the second analysis extracted a two-factor solution with an eigenvalue of 9.894, explaining 52.1% of the variance. The solution comprised one factor with 10 items, and a second factor with nine items. However, the two factors reect the direction of item wordings with the 10 rst loading items all being positively worded and with the nine items on the second factor all being negatively worded. Similar ndings have been reported by Adams, Openshaw, Bennion, Mills, and Noble (1988), Austin (1983), Knight et al. (1988), and McWhirter (1990). As already mentioned, these

1364

M. Lasgaard / Personality and Individual Dierences 42 (2007) 13591366

results indicate that the ratings of UCLA are aected by acquiescence or other similar response styles, rather than dierent typologies of loneliness. Moreover, two conrmatory analyses have supported the unidimensionality of the scale (Hartshorne, 1993; Russell, 1996). Therefore, a one-factor solution (loading criteria P.30) including all 20 items was carried out, which extracted a factor that carried 19 items and had an eigenvalue of 8.018, explaining 40.1% of the variance. Notably, the excluded item (#17), which also showed the smallest loading in two samples of the original scale (Russell, 1996), loaded with .27 on the factor, just beneath the loading criteria. For that reason the item was maintained in the scale. In summary, the performed exploratory analyses support a unidimensional structure of the Danish version of UCLA.

5. Discussion This study has examined a Danish version of the UCLA Loneliness Scale. In summary, the psychometric data presented support the validity and reliability of the scale. The internal consistency of the scale was high and the reported correlations with measures of emotional loneliness, social loneliness, self-esteem, depression, and personality traits, supported the convergent and discriminant validity of the scale. Moreover, these results were highly comparable to the original scale. Interestingly, the loneliness scores of the adolescent group were slightly positively skewed, since relatively few students reported high levels of loneliness. As stressed by Russell (1996), researchers studying loneliness should bear in mind that distributional problems may aect the results of statistical tests. Contrary to other factor analyses of the Revised UCLA (second version), which for the most part has supported a multidimensional structure, the exploratory factor analyses of the Danish version of UCLA (third version) supported a unidimensional structure. This may be explained by the minor revisions of the scale, including some changes in item wording presented in the third version of the UCLA. Moreover, some of the derived factors, reported in earlier studies, reect the direction of item wording. Hopefully, future research will solve the problem of ratings reecting the direction of item wordings. One limitation of the study is that the investigated sample consists of a narrow age group, the reason being that the measures were selected from a large-scale national probability study of 8th grade students. However, the original version of UCLA has been used successfully in a great number of student and adult populations (e.g. Russell, 1996; Wei, Shaer, Young, & Zakalik, 2005), which is why it seems highly likely that the Danish version will prove a sucient measure of loneliness in research in older populations. Another limitation of the study is that most of the measures used to validate the Danish version of UCLA have not themselves been validated as translated measures. Unfortunately, rather few measures of psychological constructs have been validated in Denmark, which is why it in most cases was not possible to include standardized measures. A problem associated with the former lack of a validated Danish measure of loneliness, is that the available knowledge about loneliness is based on answers to single-item self-rating scales about the occurrence of loneliness (e.g. how often do you feel lonely?), frequently included in large-scale studies (e.g. (King, Wold, Tudor-Smith, & Harel, 1996)). However, it is very doubtful how useful such answers are, i.e. as they are more prone to eects of the willingness to report,

M. Lasgaard / Personality and Individual Dierences 42 (2007) 13591366

1365

and the understanding of what loneliness is. Moreover, a meta-analysis of reported gender dierences in loneliness indicates that studies, which use single-item self-rating scales, achieves results rather dierent from studies, which measure loneliness by using UCLA (Borys & Perlman, 1985). In conclusion, a Danish measure of loneliness has been standardized in a national probability study, and is available for future research, which could focus on predictors and consequences of loneliness, and possible mediating factors (e.g. attributions, coping, and attachment). References
Adams, G. R., Openshaw, D. K., Bennion, L., Mills, T., & Noble, S. (1988). Loneliness in late adolescence: a social skills training study. Journal of Adolescent Research, 3(1), 8196. Anderson, L. R., & Malikiosi-Loizos, M. (1992). Reliability data for a Greek translation of the Revised UCLA Loneliness Scale: comparisons with data from the USA. Psychological Reports, 71(2), 665666. Austin, B. A. (1983). Factorial structure of the UCLA Loneliness Scale. Psychological Reports, 53(3), 883889. Beck, J. S., Beck, A. T., & Jolley, J. B. (2004). Beck Youth Inventories: manual. Copenhagen: Dansk Psykologisk Forlag. Borys, S., & Perlman, D. (1985). Gender dierences in loneliness. Personality and Social Psychology Bulletin, 11(1), 6375. Brage, D., Meredith, W., & Woodward, J. (1993). Correlates of loneliness among Midwestern adolescents. Adolescence, 28(111), 685694. chelle de solitude de lUniversite SUL): Validation Laval (E ce, G.-R., Joshi, P., & Pelletier, R. (1993). LE de Gra canadienne-franc aise du UCLA Loneliness Scale. Canadian Journal of Behavioural Science, 25(1), 1227. Do ring, N., & Bortz, J. (1993). Psychometrische Einsamkeitforschung: Deutsche Neukonstruktion der UCLA Loneliness Scale. Diagnostica, 39(3), 224239. Eysenck, H. J., & Eysenck, S. B. G. (1991). Manual of the Eysenck Personality Scales. London: Hodder and Stoughton. Garnefski, N., Diekstra, R. F., & de Heus, P. (1992). A population-based survey of the characteristics of high school students with and without a history of suicidal behavior. Acta Psychiatrica Scandinavica, 86(3), 189196. Glorfeld, L. W. (1995). An improvement of Horns parallel analysis methodology for selecting the correct number of factors to retain. Educational and Psychological Measurement, 55(3), 377393. Hartshorne, T. S. (1993). Psychometric properties and conrmatory factor analysis of the UCLA Loneliness Scale. Journal of Personality Assessment, 61(1), 182195. Hojat, M. (1982). Psychometric characteristics of the UCLA Loneliness Scale: a study with Iranian college students. Educational and Psychological Measurement, 42(3), 917925. Hoza, B., Bukowski, W. M., & Beery, S. (2000). Assessing peer network and dyadic loneliness. Journal of Clinical Child Psychology, 29(1), 119128. Johnson, H. D., LaVoie, J. C., Spenceri, M. C., & Mahoney-Wernli, M. A. (2001). Peer conict avoidance: associations with loneliness, social anxiety, and social avoidance. Psychological Reports, 88(1), 227235. Jones, W. H., & Carver, M. D. (1991). Adjustment and coping implications of loneliness. In C. R. Snyder, & D. R. Forsyth (Eds.), Handbook of social and clinical psychology: the health perspective (pp. 395415). New York: Pergamon Press. King, A., Wold, B., Tudor-Smith, C., & Harel, Y. (1996). The health of youth: a cross-national survey, WHO regional publications. European Series, 69. Knight, R. G., Chisholm, B. J., Marsh, N. V., & Godfrey, H. P. D. (1988). Some normative, reliability, and factor analytic data for the Revised UCLA Loneliness Scale. Journal of Clinical Psychology, 44(2), 203206. Koenig, L. J., Isaacs, A. M., & Schwartz, J. A. J. (1994). Sex dierences in adolescent depression and loneliness: why are boys lonelier if girls are more depressed? Journal of Research in Personality, 28(1), 2743. Lasgaard, M. (2006). Ensom blandt andre: En psykologisk undersgelse af ensomhed hos unge i Danmark. [Lonely among others: A psychological investigation of loneliness in young people in Denmark]. Psykologisk Studieskriftserie, 9(1), 1131. Lau, S., Chan, D. W. K., & Lau, P. S. Y. (1999). Facets of loneliness and depression among Chinese children and adolescents. Journal of Social Psychology, 139(6), 713730.

1366

M. Lasgaard / Personality and Individual Dierences 42 (2007) 13591366

Mahon, N. E., Yarcheski, T. J., & Yarcheski, A. (1995). Validation of the Revised UCLA Loneliness Scale for adolescents. Research in Nursing and Health, 18(3), 263270. Mahon, N. E., Yarcheski, A., & Yarcheski, T. J. (2004). Social support and positive health practices in early adolescents: a test of mediating variables. Clinical Nursing Research, 13(3), 216236. McWhirter, B. T. (1990). Factor analysis of the Revised UCLA Loneliness Scale. Current Psychology, 9(1), 5668. Moore, D., & Schultz, N. R. (1983). Loneliness at adolescence: correlates, attributions, and coping. Journal of Youth and Adolescence, 12(2), 95100. Neto, F. (1992). Loneliness among Portuguese adolescents. Social Behaviour and Personality, 20(1), 1522. OConnor, B. P. (2000). SPSS and SAS programs for determining the number of components using parallel analysis and Velicers MAP test. Behavior Research Methods, Instruments, and Computers, 32(3), 396402. Pretoirus, T. B. (1993). The metric equivalence of the UCLA Loneliness Scale for a sample of South African students. Educational and Psychosocial Measurement, 53(1), 233239. Roberts, R. E., Roberts, C. R., & Chen, R. Y. (1998). Suicidal thinking among adolescents with a history of attempted suicide. Journal of the American Academy of Child and Adolescent Psychiatry, 37(12), 12941300. Rosenberg, M. (1965). Society and the adolescent self-image. New Jersey: Princeton University Press. Ruchkin, V. V., Eisemann, M., & Ha gglof, B. (1999). Hopelessness, loneliness, self-esteem, and personality in Russian male delinquent adolescents versus controls. Journal of Adolescent Research, 14(4), 466482. Russell, D. W. (1996). UCLA Loneliness Scale (version 3): reliability, validity, and factor structure. Journal of Personality Assessment, 66(1), 2040. Russell, D., Peplau, L. A., & Cutrona, C. E. (1980). The Revised UCLA Loneliness Scale: concurrent and discriminant validity evidence. Journal of Personality and Social Psychology, 39(3), 472480. Russell, D., Peplau, L. A., & Ferguson, M. L. (1978). Developing a measure of loneliness. Journal of Personality Assessment, 42(3), 290294. Sacchi, C., & Richaud de Minzi, M. C. (1997). UCLAs revised scale of loneliness: an Argentinean adaptation. Revista gica, 6(1), 4353. Argentina de Clinica Psicolo Schultz, N. R., & Moore, D. (1988). Loneliness: dierences across three age levels. Journal of Social and Personal Relationships, 5(3), 275284. Thastum, M., Paulsen, K., & Hansen, H. S. (2004). Dokumentation for den danske udgave af Beck Youth Inventory (BYI). [Documentation for the Danish version of Beck Youth Inventory (BYI)]. In J. Beck, A. Beck, & J. Jolley (Eds.), Beck Youth Inventories: manual (pp. 2935). Copenhagen: Dansk Psykologisk Forlag. Troop, N. A., & Bifulco, A. (2002). Childhood social arena and cognitive sets in eating disorders. British Journal of Clinical Psychology, 41(2), 205211. Uruk, A. C., & Demir, A. (2003). The role of peers and families in predicting the loneliness level of adolescents. Journal of Psychology: Interdisciplinary and Applied, 137(2), 179194. Velicer, W. F. (1976). Determining the number of components from the matrix of partial correlations. Psychometrika, 41(3), 321327. Wei, M., Shaer, P. A., Young, S. K., & Zakalik, R. A. (2005). Adult attachment, shame, depression, and loneliness: the mediation role of basic psychological needs satisfaction. Journal of Counselling Psychology, 52(4), 591601. West, D. A., Kellner, R., & Morre-West, M. (1986). The eects of loneliness: a review of the literature. Comprehensive Psychiatry, 27(4), 351363. Wilson, D., Cutts, J., Lees, I., Mapungwana, S., & Maunganidze, L. (1992). Psychometric properties of the Revised UCLA Loneliness Scale and two short-form measures of loneliness in Zimbabwe. Journal of Personality Assessment, 59(1), 7281.