What is Longitudinal Research?




Longitudinal research refers to research that investigates events or phenomena over an extended period of time. Longitudinal research studies can be as brief as 1 or 2 years, for example, when evaluating the effects of a particular learning strategy, technique, or treatment. They can also run over several decades, as when examining changes in adult personality and health behaviors. They allow changes to be assessed over a variety of levels and take into account the natural development and growth that occurs across the life span. For example, consider a study that follows children over a several-year period to assess the consequences of media exposure, as opposed to a study done over a few months to evaluate the effects of exposure to violent images on levels of overt aggression. The first would be longitudinal, whereas the second study would not be considered a longitudinal study unless the children or age cohort was repeatedly assessed over a longer period of time.

Types Of Longitudinal Studies

There are four principle types of longitudinal studies: trend studies, cohort studies, panel studies, and case-based studies. All four types tend to be descriptive, in that they generally do not manipulate variables (a variable is any clearly described item or construct that can be observed and analyzed), but describe how selected variables change over time and how these changes are related to other variables. Trend, cohort, and panel studies typically use instrumental measures, such as surveys, questionnaires, pencil-and-paper tests, and other standardized instruments, supplemented by descriptive measures. Casebased studies typically use descriptive measures, such as  interviews,  projective  techniques,  observations, and narratives and are supplemented by instrumental measures.

Academic Writing, Editing, Proofreading, And Problem Solving Services

Get 10% OFF with 24START discount code


Trend studies examine changes within a defined sample population that does not stay constant. For example, if you are interested in changes in the amount of television or video exposure in preschoolage children, you could take a survey sample from preschool-age children repeatedly, over several years, and look at the trend. The children from whom you draw the sample would be constantly changing, and you would be measuring the patterns of change in the viewing  exposure  of  a  defined  sample  population,3to 5-year-old children, over time.

Cohort studies examine changes within a defined sample population that is stable. For example, if you are interested in changes in employment and health status of American Vietnam-era combat veterans, you could periodically sample from a directory of U.S. military personnel who were listed as serving in combat between 1962 and 1974. The sample may contain different individuals at each data collection point, but it would represent the same population cohort at each time.

Panel studies examine changes over time within a selected sample that remains constant. For example, you are interested in describing the changes that occur over time as people age. You could select a group of people and have them complete a variety of measures at repeated intervals over several decades. The people in the study remain the same at each data collection point,  with  natural  attrition  (e.g.,  an  individual’s death) being the primary cause of the panel sample changing.

Case-based studies examine changes within organizations, groups, or individuals, often with regard to some intervention. The intervention could be training, a new policy or procedure, an educational program, or therapy. For example, you are interested in changes in an organization resulting from the introduction of a training program for a new technology. You could collect a set of measures from a sample of employees or students before the introduction of the program and then repeat the measures at set intervals. Similar to a cohort or panel study, a case-based sample would remain stable (in the case of employee or student samples in an organization) or the same (in the case of group or individual participants).

Advantages Of Longitudinal Studies

Longitudinal research is an effective and powerful method for investigating developmental variables and is essential in understanding outcomes for most naturally occurring or socially imposed interventions. One of the chief advantages of longitudinal research is that, since the same subjects or cohorts are being followed and repeatedly sampled over time, changes that are observed can be attributed to individual changes, rather  than  variations  over  individuals  (referred  to as sample variance or statistical error). This allows a reliable description of patterns of change in individuals or groups, and a description of the direction and magnitude of causal relationships between variables, that would be available no other way.

Disadvantages Of Longitudinal Studies

Longitudinal studies are considered nonexperimental, meaning that they are not conducted using randomized selection of participants who are compared with a matched control group through manipulation of a variable to expose the effect of that variable on the (experimental) participants. Nonexperimental studies are considered inferior by experimentalists, who place primary importance on the causal inference that randomization, control, and manipulation theoretically provide. However, as mentioned earlier, longitudinal studies can be used to describe the direction and magnitude of casual relationships, even though they are not as theoretically precise as experimental studies in determining causal relationships.

From a more practical perspective, the central disadvantage to longitudinal research is the length of time it takes to complete a study, with the attendant problems of attrition (in both participants and research assistants), costs, and measures. Conducting a research project that follows a group of participants over the course of years or decades requires significant investment of time from both the participants and the research team, significant financial support to maintain contact with  participants  and  involvement  of  investigators, and significant foresight on the part of the investigators in selecting measures that will not become antiquated during the course of the study.

Examples Of Longitudinal Studies

The Study of Adult Development at Harvard has followed two panels of people over a long period of time: a panel of 268 Harvard students selected from the sophomore classes between 1939 and 1942 representing a socially advantaged, predominately male group and a panel of 456 disadvantaged inner-city males from Boston born about 1930. This Harvard-based study has focused on the physical and psychological health of the panel members, using social histories, biennial questionnaires, face-to-face interviews (every 15 years), and complete physical examinations (every 5 years). It has provided numerous insights into the components of healthy aging as well as insights into social and cultural constraints on aging, adaptation to stress, habits and coping skills, and childhood risk factors.

The Seattle Longitudinal Study, a study of adult development from midlife through old age, has followed  a  panel  of  people  since  1956. The  research has focused on psychological development during the adult years. The original panel of 500 participants was randomly selected from a community health program and ranged in age from early twenties to late sixties. The study has continued since 1956, with data collected in 1963, 1970, 1977, 1984, 1991, and 1998. At each interval, a new group of people randomly selected from the same community health program have been asked to participate. As of the 1998 data collection, nearly 6,000 people have participated at some time in this study. Of the original participants from 1956, 38 people remain who have now been in the study for 42 years. The study collected data from the primary participants as well as siblings and offspring over the years, and it has been the source of numerous academic publications, as well as congressional presentations and policy discussions regarding issues related to aging, retirement, and other public policy considerations.

Numerous other databases have been created over the past several decades. Many agencies and institutions make these databases available to qualified researchers at no cost or for a nominal fee.

Summary

Longitudinal  research  studies  investigate  events or phenomena over extended periods of time. They follow the same or a similar group of people and are essential in understanding human change and development. However, they demand significant investments of time and resources and are usually descriptive because they do not provide the precise cause-and effect logic of experimental studies. They provide reliable descriptions of patterns of change and of direction and magnitude of change that would be available no other way.

References:

  1. Bijleveld, , Kamp, L., Mooijaart, A., Kloot, W., Leeden, R.,& Burg, E. (2004). Longitudinal data analysis: Designs, models, and methods. Thousand Oaks, CA: Sage.
  2. Institute for Personality and Social Research at the University of California at Berkeley, http://ls.berkeley.edu/dept/ipsr/ IPSRArchiveWeb/Archivhtm
  3. Menard, (2002). Longitudinal research. Thousand Oaks, CA: Sage.Murray Research  Center,  http://www.radcliffe.edu/murray/index.php
  4. National Archive of Computerized Data on Aging, http://www.icpsr.umich.edu/NACDA
  5. Schaie, W. (2004). Developmental influences on adult intelligence: The Seattle Longitudinal Study. New York: Oxford University Press.
  6. Seattle Longitudinal  Study,  http://geron.psu.edu/sls/index.html
  7. Vaillant, G. (2002). Aging well: Surprising guideposts to a happier life from the landmark Harvard Study of Adult Development. New York: Little, Bro