17
Basic Human Values 3 Wednesday 17th July, 16:00 - 17:30, Room: No. 21 1 Self-other agreement in values

Basic Human Values 3 - €¦ · Basic Human Values 3 Wednesday 17th July, 16:00 ... Development and University of Tartu, ... difficult for others to infer a person’s values because

  • Upload
    ngokien

  • View
    213

  • Download
    0

Embed Size (px)

Citation preview

Basic Human Values 3

Wednesday 17th July,

16:00 - 17:30, Room: No. 21

1 Self-other agreement in values

Self- other agreement in personal values

Henrik Dobewall1; Toivo Aavik1; Kenn Konstabel2; Shalom H. Schwartz3,4, & Anu Realo1

1 University of Tartu, Estonia; 2 National Institute for Health Development and University of Tartu, Estonia; 3 The Hebrew

University of Jerusalem, Israel; 4 National Research University—Higher School of Economics, Russia

2 Self-other agreement in values

Research question

• Can we judge other people’s values accurately, or are values too subjective to assess?

• To address this question, we examined self-other agreement for personal values, both for four higher-order values and six more narrowly defined value factors.

• We also compared agreement on values with agreement on the Big Five personality traits.

Self-other agreement in values 3

Self-other agreement

• People’s self-reports of their behaviour, attitudes, and personality may be affected by various response biases, for instance, socially desirable responding (Paulhus, 1991).

• How do we know, then, if a person truly endorses Benevolence values highly or rejects Power values?

• One possibility is to collect data through an independent measurement method, using opinions of other people (e.g., peers, spouses, siblings, parents etc.) who know the person well.

Self-other agreement in values 4

Agreement correlations in other related constructs

• An examination of the convergent (Campbell & Fiske, 1959) or consensual (McCrae, 1982) validity (i.e., the correlation between self and other-reports) has a remarkable history in the fields of

• well-being (e.g., Dobewall et al., 2012; r = .55 / Schneider & Schimmack, 2009; r = .42),

• Big Five personality traits (e.g., Connolly et al., 2007; r = .36 / Konstabel et al., 2012; McCrae et al., 2004; r = .40 to r = .70)

• or affectivity traits (Watson et al., 2000).

Self-other agreement in values 5

Previous research I

• The use of other-ratings in value research is relatively scarce (for exceptions see Rentfrow & Gosling , 2006; Paryente & Orr, 2010).

• Lee and colleagues (2009) reported – using the Schwartz Value Survey – self-other correlations for Openness to Change vs. Conservation values (r = .42) , Self-Transcendence vs. Self-Enhancement values (r = .52) and the ten basic values (ranging from r = .18 for Achievement to r = .49 for Power).

Self-other agreement in values 6

Previous research

• None of these studies, however, took the examination of self-other agreement as their focus. Several did not report many of the agreement correlations (e.g., Murray et al., 2002) or interpret them meaningfully (e.g., Lee et al., 2009).

Self-other agreement in values 7

Subjectivity of values • Personal values might be too privately held personal concerns

(McAdams, 1995) or “too individually subjective” (Hitlin & Piliavin, 2004, p. 359) to be judged by others.

• Rokeach (1973) and Schwartz (1992) both suggest that it is difficult for others to infer a person’s values because a value may be expressed in a variety of behaviours and any single behaviour may express multiple values.

• Moreover, values refer to motivation, not to action, so observers must infer them indirectly.

• Five Factor Theory of personality: Values, are so-called characteristic adaptations (formed through the interaction of traits with the environment; McCrae & Costa &, 1999) better assessed by direct observation than traits (Allik & McCrae, 2002).

Self-other agreement in values 8

Correction for measurement error I

• In order to obtain more accurate comparisons between self-other agreement in values and in personality traits.

• Schwartz (2005) explains that the specific values measured with the ESS Human Values Scale (PVQ21) have low reliabilities because each scale has only two items (three for universalism), and these items were constructed and selected to cover the broad conceptual components of each value rather than to express the same narrowly defined content.

Self-other agreement in values 9

Correction for measurement error II

• Internal consistency coefficients that measure scale reliability (Cronbach, 1951) can be used to estimate the proportion of true score variance.

• Function of two parameters: (1) the average inter-correlation among the set of items and (2) the number of items in the scale (Simms & Watson, 2007).

• Random error can be reduced by obtaining multiple responses (here, two informants) (Schmidt & Hunter, 1996).

• Principle of aggregation (Rushton et al., 1983): Aggregate sets of indicators estimate constructs in a more stable and representative way than any single item can do because measurement errors will, to some extent, be averaged out.

Self-other agreement in values 10

Method

Sample 1 Measures

• Portrait Value Questionnaire (PVQ21; Schwartz et al., 2001)

• 4 higher-order values used.

• ‘Short Five’ personality inventory (S5; Konstabel et al., 2012).

Data

• N = 101 for self-ratings and N = 96 for other-ratings.

• Mean age of 25.7 years (SD = 7.9) / 26.9 years (SD = 9.5).

• Acquaintance 9.7 years (SD = 8.2).

Sample 2

Measures

• Estonian Value Inventory (EVI; Aavik & Allik, 2002)

• 6 culture-specific factors.

Data

• N = 97 for self-ratings and other-ratings.

• Mean age of 21.8 years (SD = 4.3) / 26.7 years (SD = 10.9).

• The average time of acquaintance was 3.4 years.

Self-other agreement in values 11

• Some gender bias

Internal consistency coefficients and self-other agreement correlations for higher-order values (PVQ21) and Big Five personality domains (S5)

Self-other agreement in values 12

Cronbach

Alphas

Self-Other

Agreement

Self-

ratings

Other-

ratings

Observed Corrected No. of

Items

PVQ21

Openness to Change .75 .79 .50 .65 6

Conservatism .69 .63 .43 .65 6

Self-Enhancement .74 .81 .49 .63 4

Self-Transcendence .57 .79 .45 .67 5

Median .72 .79 .47 .65

S5

Neuroticism .90 .86 .60 .68 12

Extraversion .86 .86 .74 .86 12

Openness .70 .79 .51 .68 12

Agreeableness .70 .80 .52 .69 12

Conscientiousness .83 .85 .51 .61 12

Median .83 .85 .52 .68

Internal consistency coefficients and self-other agreement correlations of the value factors from the Estonian Value Inventory (EVI)

Self-other agreement in values 13

Cronbach

Alphas

Self-Other

Agreement

Self-

ratings

Other-

ratings

Observed Corrected No of

Items

EVI

Self-Realization .75 .78 .34 .44 8

Self-Enhancement .73 .81 .50 .65 8

Benevolence .80 .80 .37 .46 8

Hedonism .80 .84 .58 .71 8

Conservatism .73 .79 .54 .71 8

Broadmindedness .69 .66 .50 .74 8

Median .74 .79 .50 .68

Discussion

• When corrected for attenuation due to measurement error, self-other agreement on the higher-order values was substantial (median r = .65) and similar to that for the Big Five personality traits (median r = .68).

• Self-other agreement was also substantial for the six more narrowly defined value factors (median r = .68).

• Without correction, self- and other-ratings of values share only 22% or 25% of their variance, suggesting a more modest reading of the findings.

• Agreement coefficients were somewhat smaller but robust based on uncentered value scores.

Self-other agreement in values 14

Can informants assess targets’ mean value ratings (MRAT) with any accuracy?

• Style vs. substantive meaning (Schwartz et al., 1997; Beierlein et al., 2011)

• MRAT is a relatively stable individual characteristic.

• Informants are unlikely to know how targets use a response scale (see Schwartz, 2005).

• Informants might have a sense of the extent to which targets describe themselves as strongly endorsing few or many values, or to which they describe themselves in a socially desirable way (Lönnqvist et al., 2007).

• Result: Sample 1; r = .28 (.40 disattenuated) / Sample 2; r = .35 (.38 disattenuated) .

Self-other agreement in values 15

Conclusions

• Informants can judge targets’ values just as accurately as they judge their well-being, personality traits, emotions, and other individual attributes.

• The results suggest that other-ratings of personal values can be used to validate (see e.g., Sandvik et al., 1993) and complement (see e.g., Fogarty et al., 2013) self-report value measures.

Self-other agreement in values 16

Acknowledgments

• This research was supported by grants from the Estonian Ministry of Science and Education (IUT2-13 and PUT78). The work of the fourth author was partially supported by the International Laboratory of Socio-Cultural Psychology of the Higher School of Economics, Moscow. We thank Diana Boer, Jan-Erik Lönnqvist, and Delaney Michael Skerrett for their helpful comments on earlier drafts of this article.

• This study was supported by the European Social Fund’s Doctoral Studies and Internationalization Program, DoRa.

Self-other agreement in values 17