Page 534 - Cultures and Organizations

Notes 499

36. Correlation r 0.65** across twenty-two of the twenty-three CVS countries
(all except Thailand, for which the necessary WVS data were not available).
37. In the new, 2008 version of the Values Survey Module (VSM08), we included
both monumentalism and long-term orientation items, expecting they might merge
into one dimension.
38. Heine, 2003; Minkov, 2007, pp. 164ff.
39. “Service to others” appeared only in the WVS prior to 2005. For the sake of con-
sistency, Misho excluded the 2005–08 wave for all three items. From the 1995–2004
period, he used the latest data for each item.
40. We followed Misho’s method, which he borrowed from Inglehart: using per-
centages of people who selected a particular position on the item scale, not national
averages. In accordance with that method, we used the positive extremes: percentages
of people who selected options such as “very important,” “very proud,” or “very much
like me.” The reason for adopting this method is that when it is used with nationally
representative samples, it appears to yield results having the highest predictive powers
with respect to many external variables.
41. The correlations between these items and LTO-CVS were as follows: thrift r
0.53 (p 0.013, n 21); national pride r 0.64 (p 0.002, n 21); service to oth-
ers r 0.70 (p 0.008, n 13).
42. Factor-analyzed, the three items yielded a single factor with an eigenvalue of
2.10, explaining 70 percent of the variance. The item loadings were as follows: service
to others 0.94; national pride 0.86; thrift for children 0.70. Because service to others
had quite a few missing values, we used linear regression on the two other variables
to predict the missing factor scores. Both items were significant predictors in stepwise

2
regression with a highly reliable cumulative R of 0.97. After adding the predicted
scores to those obtained from the factor analysis, we transferred the scores to a 0–100
scale, reversing the sign.
43. As mentioned in note 39, Misho excluded the 2005–08 wave. For nine countries
for which one or more scores from 1994–2004 were missing but 2005–08 data were
available, we used the latter anyway. In Table 7.4 these countries are marked with an
asterisk (*).
44. Across twenty-one common cases, r 0.72*** (excluding Hong Kong and Thai-
land, for which missing data in the 1995–2004 period had been replaced by data from

2005–08).
45. Importance of tradition was measured in the 2005–08 wave of the WVS with
item v89, which—as with the other items in the same section—is based on the work
of Shalom Schwartz. Item v89 is worded as follows: “Using this card, would you please
indicate for each description whether that person is very much like you, like you, some-
what like you, not like you, or not at all like you? (Code one answer for each descrip-
tion): Tradition is important to this person: to follow the customs handed down by
one’s religion or family.” Our new measure of LTO correlates with the percentages of
respondents who chose “very much like me” at r 0.56*** (n 37). Importance of
perseverance is measured in the WVS as a desirable trait for children, just like “thrift”
in our new LTO score. The item is coded as A039 prior to 2005 and v18 afterward.