Skip to content

Unit 2.8 Dataset metadata

Overview

Unit study time

  • 10 minutes

Intended Learning Outcome

By the end of the unit, you will be able to ...

  • Describe what dataset metadata provides and how it differs from study metadata.

What is dataset metadata?

Dataset metadata provides the who, what, when, why, where and how of a singular dataset. It provides specific information about how the data within a dataset was collected.

For larger research projects, you may produce multiple datasets. These datasets could have different areas of interest or be carried out in different locations or by a certain researcher. As such, having metadata for each dataset helps you manage vast amounts of data and saves you time in the future.

Many of the metadata elements we use to describe the overall study can be reused at the dataset level.

What metadata elements we capture at the study level could we also capture at the dataset level?

Repeated metadata elements to capture for datasets

Title The title of the dataset

Creator The creator of the particular dataset

Subject e.g. keywords or topics

Description e.g. a description of the dataset and what it includes

Contributor

e.g. people or organisations who contributed to the research process

Date e.g. the date range of when the data for that dataset was collected

Type

Format The format that the dataset is stored in

Language The language the dataset is stored in

Relation Any other publications or resources that are related to that dataset

Coverage The geographical coverage of the dataset

Access rights The access rights of the individual dataset

We can also capture additional metadata elements that we didn't cover in the study metadata to give further information about a dataset.

What additional metadata elements should we collect for a dataset?

New metadata elements to capture for datasets

Kind of data

Study The study that produced the dataset

Case quantity The case quantity refers to number of data instances that were recorded (for tabular data, this is the number of rows)

Variables How many variables are in the dataset

Last Updated When the dataset was last updated

Test your knowledge

What is meant by “dataset metadata”?

  1. Information about individual variables only
  2. Information describing an entire dataset, including its structure, content, and context
  3. The values within the dataset
  4. The coding scheme for a single variable
Reveal answer

Dataset metadata provides an overview of the dataset as a whole, including its structure, scope, and key characteristics.

Two datasets contain similar variables, but one includes multiple waves and the other is a single extract. Why does this matter?

  1. File formats may differ
  2. The datasets cannot be used together
  3. Variable names will change
  4. Dataset structure affects how the data can be analysed and compared
Reveal answer

Differences in dataset structure (e.g. longitudinal vs single collection) affect how data should be analysed and compared.

Why is dataset metadata important for reuse and sharing?

  1. It reduces storage requirements
  2. It ensures identical results
  3. It provides essential information for others to understand and work with the dataset
  4. It eliminates all errors
Reveal answer

Dataset metadata helps others understand how the dataset is structured and how it can be used, supporting effective reuse.