USER Log-in

Australian Data Archive

Creating and Managing Longitudinal Data

The creation, management and analysis of longitudinal data is often more complex than that of cross-sectional data, for a vaerity of reasons. This section of ADA Longitudinal provides advice and guidelines on how to create and manage longitudinal data files, and links to methods and techniques for longitudinal data analysis.

Creating Longitudinal Data

The preparation and management of longitudinal quantitative data is a complex process, given the periods over which such data is collected, and the potential changes in staff, processes and procedures that can occur over the life of a longitudinal study. We aim here to provide advice and recommendations on the creation and management of longitudinal data based on the archive’s experience and preferences with such data.

The expression “data preparation” is very broad and its interpretation can vary between disciplines. Generally speaking ‘data’ refers to some kind of information with either quantitative or qualitative attributes. A set of recorded interviews, results of blood tests, information gathered from a consumer questionnaire and video recordings are only a few examples of data. Therefore, the activities summarized by ‘data preparation’ differ depending on the type of data being managed.

This section provides advice on procedures for preparing longitudinal quantitative data files, including:

Linking data across waves

Variable and value labelling

Managing missing data

Data cleaning

Using Longitudinal Data

Longitudinal research is important for many disciplines, as longitudinal data facilitates the investigation of research questions related to change and stability. In contrary to a cross-sectional study, where the data set contains observations on multiple variables observed at a single point in time for each unit in the study, the data set of a longitudinal study consists of repeated observations of similar items over a period of time.

Longitudinal studies can reveal shifting attitudes and patterns of individual behavior that might go unnoticed with other research approaches. In general, longitudinal studies provide data suitable for sophisticated statistical analysis of change over time and might enable researchers to predict cause-effect relationships. Additionally, longitudinal studies are particularly useful to analyze and predict long-term or cumulative effects which are normally hard to analyze in a cross-sectional study.

This section provides advice, links and materials on using longitudinal data in your research.

Funding Partners

Collaboration Partners

logo_arc logo_nrc Logo_anu logo_uwa logo_uts logo_um logo_uq