Definition · Plain-language
Longitudinal study
A longitudinal study observes the same variables, and usually the same subjects, repeatedly over an extended period to track how they change over time.
The step most authors miss
Doing CRediT right? Don’t stop at the statement.
A CRediT statement credits you inside one paper. The recognition CRediT was built for happens when those roles are tied to you, persistently. Sign in with your ORCID — free — and claim your CRediT contributions on casrai.org, the home of the standard. They become a verified, portable part of your identity, not a line that disappears into one PDF.
Free: claim your contributions, then export a journal-ready CRediT statement, schema.org structured data, JATS XML, CSV or BibTeX — and preview your public profile. A membership publishes that profile publicly and verifies the journals you serve.
Following change over time
A longitudinal study is defined by repeated observation: the same variables are measured on more than one occasion, usually over an extended span, often on the same individuals. This lets researchers see how things develop — how attitudes shift, how a disease progresses, how children mature — rather than capturing a single moment. Because measurements are ordered in time, the design can establish that a presumed cause preceded its effect, which is one of the conditions for inferring causation, and it can distinguish genuine change within individuals from differences merely between them.
Panel, cohort and trend designs
Longitudinal research takes three main forms, distinguished by who is followed. A panel study tracks exactly the same individuals over time, measuring each repeatedly, which gives the richest picture of individual change. A cohort study follows a group sharing a defining characteristic or event — such as people born in the same year — though it may sample fresh members of that cohort at each wave. A trend study repeatedly samples a population over time without following the same people, revealing population-level change. Each balances the depth of individual tracking against cost and practicality differently.
Attrition and other challenges
The strengths of longitudinal research come at a price. The defining problem is attrition: participants drop out between waves through relocation, disengagement, illness or death, and if those who leave differ systematically from those who stay, the remaining sample becomes biased. Longitudinal studies are also expensive, slow to yield results, and exposed to practice effects from repeated testing and to historical changes during the study period. Researchers manage these with strong participant-retention strategies, statistical methods for missing data, and clear documentation. Despite the cost, the ability to observe change directly makes the design uniquely valuable.
Key facts
At a glance
- Definition: repeated observation of the same variables over an extended time period
- Contrast: a cross-sectional study is a single snapshot in time
- Panel study: follows the same individuals at every wave
- Cohort study: follows a group sharing a characteristic or event
- Trend study: re-samples a population over time without the same people
- Main challenge: attrition — biased dropout of participants between waves
Common misconceptions
What people often get wrong
Often heard: A longitudinal study just means a study that lasts a long time.
Actually: Duration alone is not enough. What defines it is repeated measurement of the same variables over time; a single long study collecting data only once is still cross-sectional.
Often heard: A longitudinal study must follow the very same individuals throughout.
Actually: Panel studies do, but cohort and trend designs may sample different members of the same population at each wave. Repeated measurement over time, not identical participants, is the defining feature.
Often heard: Attrition is a minor inconvenience that does not affect the results.
Actually: Attrition is a serious threat. If those who drop out differ systematically from those who remain, the surviving sample becomes biased, distorting estimates of change and any conclusions drawn from them.
Going deeper
