Cross-cultural psychologists, and other scholars who are interested in the joint effects of cultural and individual-level constructs, often collect data and are interested in hypotheses that involve multiple levels of analysis simultaneously. For example, in cross-cultural research, it is not uncommon to collect data from numerous individuals in numerous countries (or cultures). Such data structures are frequently referred to as multilevel or hierarchically nested, or simply nested data structures because observations at one level of analysis (e.g., individuals) are nested within observations at another (e.g., culture). Within a multilevel framework, questions of interest could be couched in terms of cultural differences in means of individual-level measures such as Life Satisfaction, within-culture relationships between individual-level measures such as Life Satisfaction and Individualism, and between-cultural differences in such within-culture relationships.

When analyzing such nested data structures, the possibility that relationships among constructs can vary across levels of analysis must be taken into account. That is, relationships between two variables at the between-country level (e.g., relationships among country-level aggregates, sometimes referred to as ecological correlations) may or may not be the same as the relationships between these two variables within countries (e.g., individual-level correlations). In fact, relationships at the two levels of analysis are mathematically independent (e.g., Nezlek, 2001), and it is inappropriate to draw conclusions about within-culture relationships from between-culture analyses. This inappropriateness is highlighted by the possibility that within-country (i.e., individual-level) relationships may vary across countries, undermining the validity of any estimate of “the” individual-level relationship, simply because there may not be a single, uniform individual-level relationship.