Survey Research

Systematically gathering data from a sample

Survey research is a quantitative design that collects standardized information from a sample to describe a population's characteristics, attitudes, or behaviours. Data are gathered through questionnaires or structured interviews. Core concerns include sampling adequacy, question design, mode effects, and nonresponse bias — all of which directly threaten validity. The total survey error framework provides a systematic conceptual basis for identifying and minimizing these threats throughout the research process.

Definition and Scope

Survey research aims to collect standardized data from a sample chosen to represent a defined population. The researcher gathers information from individuals through questionnaires or structured interviews, and the findings are generalized back to the population. The design can serve descriptive purposes — revealing the distribution of a variable — or analytical purposes, examining relationships among variables. Because no manipulation is involved, survey research is observational in nature and typically supports correlational rather than causal conclusions.

How It Works: Types and Key Steps

Surveys may be cross-sectional — capturing data at one point in time — or longitudinal, following the same sample over time. Key steps are: (1) specifying research questions, (2) defining the target population and sampling strategy, (3) designing and pilot-testing questionnaire items, (4) choosing a data collection mode (online, mail, telephone, or face-to-face), (5) fielding the survey, (6) analysing nonresponse, and (7) analysing and reporting findings. Decisions made at each step have a cumulative impact on total survey error and the credibility of conclusions.

A Concrete Application Example

An educational researcher wants to measure university students' digital literacy levels. Using stratified random sampling, the researcher selects 500 students across faculties and administers an online questionnaire containing Likert-type items. Ambiguous items identified during piloting are revised. The nonresponse rate is kept below 20 percent to minimize bias. Data are analysed with descriptive statistics and factor analysis. The findings provide an evidence base for curriculum development policy, illustrating how survey research translates a broad population question into actionable, generalizable knowledge.

Common Pitfalls and Good Practice

Common pitfalls include leading or double-barrelled question wording, inadequate sample size, low response rates, and mode effects — the same question yielding different answers across administration modes. Nonresponse bias is more frequently overlooked than sampling bias, yet systematic differences between respondents and non-respondents can seriously distort findings. Good practice requires piloting the instrument, independent expert review, follow-up reminders to boost response rates, and transparent reporting of sampling procedures and response rates so readers can assess generalizability.

Key terms

Total Survey Error: The cumulative set of error sources from sampling, measurement, and nonresponse in a survey.
Nonresponse Bias: Distortion in findings because respondents differ systematically from non-respondents.
Mode Effect: The influence of data collection mode (online, mail, phone) on survey responses.
Cross-Sectional Design: A survey design in which data are collected from the sample at one point in time.
Sampling Frame: The list or registry from which sample members are drawn from the target population.