Longitudinal or panel data surveys involve repeated surveys of the same individuals, households, firms, or governments at different points in time.
Longitudinal data are often used to examine individual life courses, enabling researchers to investigate the associations between different events or circumstances in a person life. For example, you would use a longitudinal dataset to study the effect of birth weight on later educational attainment.
One example of a longitudinal dataset is the British Cohort Study which is following the lives of the 17,000 babies born in the UK between 5th and 11th April 1970. The study talked to mother and midwife about the birth of each of these babies. At ages 5, 10 and 16, class teachers and parents were interviewed. Now the study-subjects are adults and they are interviewed every 3 or 4 years about their work, relationships, health, family, politics and values. The data is this study is all anonymised before being released to researchers. ESDS Longitudinal provides access and support to the major UK longitudinal studies.
Qualitative data is data that is non-numeric, for example, interview transcripts, focus groups, oral histories, case notes, records of meetings, diaries and mass observations. Difficult to digitalise, qualitative data often includes some diverse data types such as audio, video, photos and text. ESDS Qualidata provides access and support for a range of social science qualitative datasets.
Mesodata, are data that fall into an intermediate conceptual layer between aggregate and survey level databases. A mesodata database is typically a survey dataset that has been partly aggregated, or aggregated to very small geographic units, or a mixture of survey and aggregate data. It often retains the granularity required to show some of the relational or hierarchical structures of the original survey database.