A cohort refers to a group of individuals who have common characteristics such as age, experience, location, or generation. Historically, the term was used to describe a Roman military unit. Currently, the term is used more loosely, and the grouping characteristics of a cohort can be quite varied. The most typical type of cohort in developmental psychology is referred to as an age cohort or birth cohort and can include either a particular year of birth (e.g., 1990) or a span of years such as the “baby boomer” or “generation X” generations. These age or birth cohorts are likely to share common cultural, historical, and social influences. However, time of birth is not the only grouping characteristic of a cohort. Other types of cohorts include groups of individuals who have experienced a significant life occurrence such as 9/11 or a group of individuals who started medical school at the same time. Overall, cohorts are of concern in many areas of research other than developmental psychology, such as economics, health, and sociology, and are also of concern to demographers.
In developmental psychology, cohorts represent a methodological concern because age and cohort can be confounding variables. Thus, in many studies there is a risk of a cohort effect. A cohort effect occurs when the results are affected by the particular cohorts used in the study rather than representing true age effects. For example, in life span studies of intelligence, age and cohort are often confounded because older adults are typically less educated than younger adults, so although differences in intelligence are found across the life span, these differences do not reflect true age-related effects. In general, most developmental studies use the traditional cross-sectional or longitudinal designs, and both of these designs are susceptible to cohort effects. In the cross-sectional design, groups of different ages are included, but differences that are seemingly age related may in fact be due to unique experiences to one or more of the age, or cohort, groups. For example, in a study of reading development in 6-, 8-, and 10-year-olds, a new reading curriculum may have been implemented just as the 6-year-old group started school, and therefore their reading scores are as high as those of the 8-year-old group. It would then be erroneous to conclude that there are no age-related changes in reading development between the ages of 6 and 8 because a cohort effect may have occurred. In the longitudinal design, data are collected from one group across time, but that group may have had unique experiences such that conclusions from the study are not generalizable. In an ongoing study of computer literacy, for example, the study began in 1980 with a group of 10-year-olds tested at regular intervals. This study would yield data on 20-year-olds from 1990 that would not necessarily be reflective of the typical computer literacy of individuals who turned 20 in 2004. Once again, a cohort effect may have occurred because the data from the study are not applicable to the present day.
To control for effects of cohorts, several research designs are available, but the issue of which design is most appropriate in developmental psychology is contentious. Sequential designs are a common type of research design used to control for cohort effects. In sequential designs, a combination of cross-sectional and/or longitudinal designs is used. The cohort sequential design utilizes two longitudinal designs with data collection starting at two different times. The time-sequential design utilizes two cross-sectional designs with data collection at two different times. The cross-sequential design utilizes a longitudinal and cross-sectional design that includes groups of different ages that are followed longitudinally. All of these designs can be used to try to control for cohort effects.
References:
- Achenbach, T. M. (1978). Research in developmental psychology: Concepts, strategies, New York: The Free Press.
- Baltes, P. B. (1968). Longitudinal and cross-sectional sequences in the study of age and generation ef Human Development, 11, 145–171.
- Glenn, D. (1977). Cohort analysis. Beverly Hills, CA: Sage.
- Levitt, M. J. (2003). Methods of studying aging. Retrieved from http://www.fedu/~levittmj/agmethod.html
- Miller, S. A. (1998). Developmental research methods (2nd ).Upper Saddle River, NJ: Prentice-Hall.
- Schaie, K. W. (1965). A general model for the study of developmental problems. Psychological Bulletin, 64,92–107.
- Woolfe, L. M. (n.d.). Theoretical perspectives relevant to developmental psychology. Retrieved from http://www.webster.edu/~woolflm/designs.html