Individual differences within the psychological atlas of the world Lazar Stankov Institute for Positive Psychology and Education, Australian Catholic University, Strathfield, NSW 2135, Australia abstract article info Article history: Received 9 October 2015 Received in revised form 11 January 2016 Accepted 14 January 2016 Available online 5 February 2016 This paper presents findings based on 8883 participants from 33 countries. It employs mixture modeling (latent profile analysis) to classify individuals into latent classes/groups. The analyses are based on 12 factor scores from the domains of social attitudes (3 factors), social axioms (5 factors) and social norms (4 factors). Five latent clas- ses were identified and most countries tend to have individual members from each class. The three largest groups consisting of 75% of the total number of participants differed in terms of the mean levels on factor scores. Group 1, labeled Liberal, had low mean factor scores on 11 out of 12 measures. Group 2, labeled Moderate, had average factor scores and Conservative Group 3 had high scores. Participants from each group were more common in some and less common in other countries. European countries plus Australia and Canada had the largest number of individuals belonging to Liberal Group 1. Conservative Group 3 had large number of individuals from South Asia and South-East Asia, Sub-Saharan Africa and Latin America. All other countries in our sample, including the USA, Russia and China had the largest number of people from the Moderate Group 2. © 2016 Elsevier Ltd. All rights reserved. Keywords: Cross-cultural differences Social attitudes Social axioms Social norms Psychological continents 1. Dimensions of cross-cultural differences vs taxonomy of individuals (and countries) A common approach in contemporary multivariate cross-cultural studies is dimensional in nature. It employs either individuals or countries as units of analysis, carries out factor analyses to establish the main dimensions along which these units differ, establishes factorial invariance and compares the overall standing of different cultural units on these dimensions. Such an approach has been used in studies of personality (Poortinga, van de Vijver, & van Hemert, 2002), social attitudes (Stankov & Lee, under review), values (Bilsky, Janik, & Schwartz, 2011), social axioms (Bou Malham & Saucier, 2014; Stankov & Saucier, 2015) and social norms (Hofstede, 2001; House, Hanges, Javidan, Dorfman, & Gupta, 2004), among others. The aims of dimensional analysis have been to find out a) on which psychological dimensions (factors) there exist large country-level differences (see Saucier et al., 2015; Stankov, 2011a, 2011b; Stankov & Lee, 2008, 2009, under review; Stankov, under review) establish the ranking of particular cultural units (e.g., countries or world regions) on these dimensions (see Stankov, 2007; Stankov & Saucier, 2015; Stankov, Lee, & van de Vijver, 2014). The present study is typological in nature. Its emphasis is on the identification of groups of participants that are similar in terms of their response patterns on a set of psychological measures. It sought to identify how many distinct groups/clusters/classes/types there are in a particular sample of participants and what are the patterns of means of each group on the measures employed. Thus, some groups of individuals could be relatively high on a particular measure, such as Extraversion, and low on measures of, say, Conscientiousness and Agreeableness, while other groups may show the opposite trend. This approach is probabilistic in the sense that it provides estimates of prob- abilities that a given unit (e.g., individual) belongs to a particular group. Also, it allows for the possibility that most, if not all, societies have mem- bers belonging to each identified group but some societies may have a larger number of representatives of a particular type. How much evi- dence, if any, is there that a large number of Southern Europeans are Ex- troverted and a large proportion of Northern Europeans are, say, Conscientious? Such evidence may open the possibility of drawing a psychological atlas to describe the distribution of human populations in the world today. These two approaches are complementary and neither is completely new. As is well known, the origin of factor analysis that is at the core of di- mensional approach is in the early work of Spearman (1904). Although Cattell (1943) was among the first to employ Tryon's (1939) cluster anal- ysis in the study of personality, recent work on the development of typol- ogies is usually linked to the publication of Lazarsfeld and Henry's (1968) book on latent structure analysis and to the work on latent variables mix- ture modeling (see Muthen, 2001; Lubke & Muthen, 2005). These latter procedures have become available in the Mplus package (see Muthén & Muthén, 2014) and are employed in the present paper. 2. Substantive issues: cross-cultural differences on non-cognitive constructs Although cross-cultural research has been concerned with perfor- mance on tests of intelligence and achievement, much of the work in re- cent years has been on non-cognitive constructs. In this paper, we focus Personality and Individual Differences 94 (2016) 180–188 E-mail addresses: lazondi@rocketmail.com, Lazar.Stankov@acu.edu.au. http://dx.doi.org/10.1016/j.paid.2016.01.027 0191-8869/© 2016 Elsevier Ltd. All rights reserved. Contents lists available at ScienceDirect Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid