Collecting Data

Data Collection Basics

Data collection refers to the process of gathering, measuring, and evaluating information on variables of interest to answer a specific question.
The results of data collection are usually displayed in diagrams, graphs, or tables.
The success of any statistical analysis substantially depends on the accuracy of data collection.

Types of Data

Primary data: This is data that is collected directly from first-hand experience. It can come from surveys, experiments, or observations.
Secondary data: This is data that has already been gathered and recorded by someone else. This can include census data, existing data sets, or records.
Both primary and secondary data can be either qualitative (categorised by qualities and attributes) or quantitative (numerical).

Methods of Data Collection

Surveys and questionnaires: This can be a simple and efficient way to gather information. However, designing a good questionnaire that will yield useful information can be complex.
Observations: Directly observing and recording information can be more reliable than self-reported data, but it can be more time-consuming and difficult to categorise.
Experiments and trials: These can give extremely precise information, but can be complex and sometimes ethically problematic.

Sampling

Collecting data from an entire population is often impractical or impossible. Hence, a sample, a smaller group chosen to represent the whole population, is used.
It’s crucial to use a sample that is representative of the population for the data to be reliable.
Random sampling is typically the best way to avoid bias.
Stratified sampling involves dividing the population into subgroups (“strata”) and randomly sampling from each group.

Data Quality

It is important to consider the reliability and validity of collected data.
Reliability refers to the consistency of a measure.
Validity refers to the authenticity and truthfulness of research findings.
Good data should be complete, consistent, and accurate.
The main sources of errors in data collection include inconsistent instrumentation, subject variability, data entry errors and experimenter bias.

Data Collection Ethics

When collecting data, especially from people, it’s important to respect privacy and confidentiality.
It’s also important to ensure that participation is voluntary and based on informed consent.
Misuse of data can have serious ethical and legal consequences.