Reviewed by Sep 30, 2020| Updated on
Longitudinal data, also called panel data, is a series of repetitive observations of the same topics, collected over a period of time from a larger population and is useful to quantify change. Longitudinal data varies from cross-sectional data, since it tracks the same subjects over a period of time, whereas cross-sectional data samples different subjects (whether people, businesses, countries or regions) at every point of time.
Since it has many advantages over repetitive cross-sectional data, longitudinal data is usually used in financial and economic studies. For example, since longitudinal data tracks how long things last, it can be used to see if the same group of people remain unemployed during a recession, or if different people move in and out of unemployment. This may help to determine the factors that affect unemployment the most.
With the historical simulation approach, the longitudinal analysis may also be used to determine the value at risk (VaR) of a portfolio. It simulates how much the value of the current portfolio would have varied over past periods, using historically observed asset movements in the portfolio during those times. It gives a forecast of the highest possible loss over the next period.
For event studies, longitudinal data is also used to examine which factors cause abnormal stock returns over a period, or how stock prices respond to announcements for merger and earnings. It can also be used to assess deprivation and income inequality by monitoring individual households.
Since standardized test scores are longitudinal in schools, they can be used to evaluate teacher's effectiveness and other factors that influence student performance.