Longitudinal study

A longitudinal study is a research design of the empirical research on the study of social and individual change processes. In a longitudinal study, the same empirical study is, in contrast to a cross-sectional study, conducted at multiple time points and compared the results of each survey wave.

Forms of longitudinal studies

A distinction longitudinal studies in trend studies (also: replicative Survey ) and panel studies.

In a trend study the same study will be conducted at multiple time points, each with different samples. With the help of trend studies it is possible to change at aggregate level (ie at the level of the total sample, which in turn is supposed to represent the population) to understand. However, it is not possible to derive changes at the individual level of trend data. If this is done yet, we speak of a so-called ecological fallacy.

In contrast to the trend study in the panel study is done (also Panel Survey, a panel survey ) the collection at multiple time points with the same sample. In this way, also intra-individual changes can be recorded ( so-called internal fluctuation ). The aggregated values ​​also make it possible - as with the design trend - the conclusion to inter-individual, that is, the total sample concerned, changes ( so-called net changes ).

Assessment and problems

A panel study is more meaningful than a trend study: As mentioned above, it is not possible to derive changes at the individual level of trend data. When a panel design, however, also intra -individual change can be detected.

The two forms of longitudinal studies have to deal with the other problems: In a trend study, the sample in each survey wave must be comparable. For example, if the proportions of different population groups in the individual waves are different, the results can be significantly distorted.

This danger is at a panel study not because always the same sample is used. However, the problem of panel attrition occurs: This refers to mortality, removals, denials and similar failures from the panel, which must be compensated by replacing people. Often, the failures are not random, but systematically, that is, the panel mortality of certain populations or risk groups is higher than in others.

To keep the panel mortality as low as possible, it is necessary to regularly maintain the panel, that is, the address register needs to be updated, losses compensated by suitable replacement people. Since this process is very complex, a panel study is more expensive than a trend study. May be distorted a panel study also characterized in that the measuring instrument ( questionnaire ) obsolete by the socio- cultural change. Another source of error are so-called panel effects such as an increased awareness by participating in the panel.

See also: Problems with panel studies

Known studies

  • Berlin Aging Study (1990-1993, BASE; Baltes, Helmchen, Steinhagen -Thiessen and others; especially geriatric issues )
  • The National Educational Panel Study (NEPS ) at the Institute of educational science longitudinal research Bamberg ( INBIL )
  • The Bonn longitudinal study normal aging (from 1965 BOLSA; Hans Thomae, Ursula Lehr and others)
  • German Ageing Survey ( DEAS ) ( from 1996)
  • Framingham Study ( Framingham Heart Study )
  • Life in Germany ( German Socio -Economic Panel - SOEP), a longitudinal study of private households in Germany.
  • Minnesota now study
  • Munich longitudinal study LOGIC
  • New England Centenarian Study ( study of centenarians )
  • The North American " Nurses' Health Study " ( Boston). Beginning with the 1976 biennial nurses are interviewed and clinically examined for T.. Have been explored thus far, for example, dietary behavior, risk factors for heart disease and cancer ( esp. breast cancer ) and the change in cognitive abilities with age.
  • Zurich Longitudinal Studies
  • Saxon longitudinal study
  • Bremer longitudinal study on the transition from school to work ( 1989-2001 ).
  • The study pairfam (2008 - 2022, 12,402 participants)
  • Generations and Gender Survey ( GGS )
  • National Longitudinal Surveys (NLS )
  • National Survey of Families and Households ( NSFH ) ( Interviews 1987-88, 1992-94 and 2001-2003)
  • Netherlands Kinship Panel Study ( NCPs )
  • Longitudinal Studies of Generations
  • Panel Study of Income Dynamics ( PSID )
  • Survey of Health, Ageing and Retirement in Europe ( SHARE)
  • R V: " The fears of the Germans." Since 1991, annually spend about 2,400 people aged 14 and over information about their personal fears, but also about worries that they make to society, economy and politics.
  • Programme for the International Assessment of Adult Competencies ( PIACC )
Cohort study Sample (statistics) Paul Baltes Nun Study Long-term experiment
498408
de