Longitudinal Design

The Structure of Longitudinal Studies

All longitudinal studies involve at least two points of measurement, and the better ones involve more than just two. The points of measurement or observations are often referred to as waves in longitudinal research. So, a three-wave study would measure participants at three different points in time. The time in between waves of measurement is called the interwave interval. If the hypothetical three-wave study had a 1-year interwave interval, it would take 2 years to complete (i.e., the “baseline” or T1 measure, the T2 measure at T1 + 1 year, and the T3 measure at T1 + 2 years). Selecting the correct interwave interval is of vital importance in longitudinal research. If the interval is too short, the researcher may never see any change in participants, and if it is too long, people may have changed and then changed back to their original state. In either case, the researcher could erroneously conclude that people were stable when in reality, things were quite the contrary.

Longitudinal studies are sometimes referred to as quasi-experiments because they examine the effect of some variable that the researcher is interested in (e.g., smoking or poor conflict management skills) but that the researcher could not actually manipulate in an experimental setting. This is because in some cases (e.g., smoking) it would not be ethical to assign people to the experimental treatment condition, and in other cases (e.g., poor conflict management skills), it is simply not possible to effectively manipulate the causal variable.

There are several different types of longitudinal study designs that are used to study different types of change.

Trend Studies

The trend study examines changes in the general population over time. In the trend study, a different sample of participants, drawn from the population of interest, is measured at each point of observation. The U.S. Census Bureau measures households and living arrangements in the U.S. population every 10 years. An examination of cohabitation rates as documented in the 1970, 1980, 1990, 2000, and 2010 censuses would amount to a trend study. Such an investigation would allow researchers to examine trends in cohabitation to see if it is on the rise or decline in the U.S. population. Trend studies are very useful for describing changes that might occur or be ongoing in the general population (e.g., consumer spending habits) over extended periods of time.

...