2 results
11 - Multilevel Modeling and Cross-Cultural Research
- Edited by David Matsumoto, San Francisco State University, Fons J. R. van de Vijver, Universiteit van Tilburg, The Netherlands
-
- Book:
- Cross-Cultural Research Methods in Psychology
- Published online:
- 05 June 2012
- Print publication:
- 11 October 2010, pp 299-345
-
- Chapter
- Export citation
-
Summary
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.
11 - Naturally occurring variability in state empathy
- from Part II - Empathy and related concepts in health
-
- By John B. Nezlek, Department of Psychology, College of William & Mary, Astrid Schütz, Department of Psychology, Chemnitz University of Technology, Paulo Lopes, Department of Psychology, University of Surrey, C. Veronica Smith, Department of Psychology, University of Delaware
- Edited by Tom F. D. Farrow, University of Sheffield, Peter W. R. Woodruff, University of Sheffield
-
- Book:
- Empathy in Mental Illness
- Published online:
- 17 August 2009
- Print publication:
- 29 March 2007, pp 187-200
-
- Chapter
- Export citation
-
Summary
Introduction
A traditional and important distinction in the study of most individual differences is that between a trait and a state. For psychologists, traits are relatively stable and enduring characteristics that tend to be considered more as causes of behaviours (including the selection of situations) than as outcomes. The adverb ‘relatively’ is included as a qualifier because traits may change over an extended period of time (e.g. over a decade or a lifetime), but for most intents and purposes, it is assumed that traits do not change. In contrast, states are presumed to change and are more often thought of as outcomes or reactions to circumstances, although it is entirely possible to think of states as causes of behaviours including the selection of situations and circumstances. There is no fixed time period over which a state must or should exist – the length of a state can vary from minutes to hours, perhaps encompassing an entire day. Regardless, the assumption is that states change, and such changes are meaningful and represent something other than random fluctuation.
The state–trait distinction is particularly important because it cannot be assumed that constructs at the two levels of analysis function in the same ways. That is, state- and trait-level phenomena may be governed by, or reflect, different psychological processes (e.g. Tennen et al., 2005). Consistent with this distinction, state- and trait-level relationships are mathematically independent (e.g. Nezlek, 2001).
![](/core/cambridge-core/public/images/lazy-loader.gif)