Linda J Waite1,2, Louise Hawkley2, Ashwin A Kotwal3, Colm O'Muircheartaigh2,4, L Philip Schumm5, Kristen Wroblewski5. 1. Department of Sociology, University of Chicago, Illinois, USA. 2. NORC at the University of Chicago, Illinois, USA. 3. Division of Geriatrics, Department of Medicine, University of California San Francisco, USA. 4. Harris School of Public Policy, University of Chicago, Illinois, USA. 5. Biostatistics Laboratory, Department of Public Health Sciences, University of Chicago, Illinois, USA.
Abstract
OBJECTIVES: In this article, we seek to provide assistance to those who might want to use data from the National Social Life, Health, and Aging Project (NSHAP) Rounds 1-3 to compare those born in different cohorts. We outline 2 theoretical models that underlie the design of NSHAP-the life course model and the birth cohorts model-and review examples of social and political changes that may have differentially affected cohorts of older adults. Then we present 2 ways that NSHAP data might be used to compare cohorts, show examples of analyses of cohort differences in measures in NSHAP, and discuss features of the data that might affect their use for this purpose. METHODS: Round 3 of the NSHAP added a group of respondents born between 1948 and 1965, the Baby Boom. Together with data from an earlier cohort, interviewed in Rounds 1-3, these data allow analysis of birth cohorts of older adults in the United States. We show examples of some approaches. RESULTS: Our age-matched cohort differences approach included all observations where the respondent was aged 57-67 at the time of interview in different time periods (3,816 observations overall; 2,316 for the Silent Generation cohort and 1,500 for the Baby Boom cohort). Our second approach, age, period, and cohort effects, models the effects of age and birth year using restricted cubic splines, with one model excluding the linear effect of birth year, and the other excluding the linear effect of period. We present examples of analyses using each of these methods. DISCUSSION: We describe features of the NSHAP data of which researchers should be aware when conducting cohort analyses with these data. Published by Oxford University Press on behalf of The Gerontological Society of America 2021.
OBJECTIVES: In this article, we seek to provide assistance to those who might want to use data from the National Social Life, Health, and Aging Project (NSHAP) Rounds 1-3 to compare those born in different cohorts. We outline 2 theoretical models that underlie the design of NSHAP-the life course model and the birth cohorts model-and review examples of social and political changes that may have differentially affected cohorts of older adults. Then we present 2 ways that NSHAP data might be used to compare cohorts, show examples of analyses of cohort differences in measures in NSHAP, and discuss features of the data that might affect their use for this purpose. METHODS: Round 3 of the NSHAP added a group of respondents born between 1948 and 1965, the Baby Boom. Together with data from an earlier cohort, interviewed in Rounds 1-3, these data allow analysis of birth cohorts of older adults in the United States. We show examples of some approaches. RESULTS: Our age-matched cohort differences approach included all observations where the respondent was aged 57-67 at the time of interview in different time periods (3,816 observations overall; 2,316 for the Silent Generation cohort and 1,500 for the Baby Boom cohort). Our second approach, age, period, and cohort effects, models the effects of age and birth year using restricted cubic splines, with one model excluding the linear effect of birth year, and the other excluding the linear effect of period. We present examples of analyses using each of these methods. DISCUSSION: We describe features of the NSHAP data of which researchers should be aware when conducting cohort analyses with these data. Published by Oxford University Press on behalf of The Gerontological Society of America 2021.
Authors: J Shepherd; S M Cobbe; I Ford; C G Isles; A R Lorimer; P W MacFarlane; J H McKillop; C J Packard Journal: N Engl J Med Date: 1995-11-16 Impact factor: 91.245
Authors: Jennifer Karas Montez; Jason Beckfield; Julene Kemp Cooney; Jacob M Grumbach; Mark D Hayward; Huseyin Zeyd Koytak; Steven H Woolf; Anna Zajacova Journal: Milbank Q Date: 2020-08-04 Impact factor: 4.911