What is it?
Survey research is essentially a method of collecting standardized information from a sample of people that represent a population through structured questions. Survey research is a frequently used method to collect objective as well as subjective data, that is, data about the inner states of respondents, such as their attitudes, beliefs, values and behaviours. Surveys are particularly suited if one wants to generalize the results to a broader population.
When to use it?
Survey research is particularly suited to:
- Learn about consumers' beliefs, attitudes and opinions around a topic.
- Test and validate research hypotheses originated by qualitative approaches (e.g., focus group).
- Study the impact of time or of an intervention program in consumers' beliefs, attitudes and values.
- Track the developmental changes and the psychological impact of life events, by using longitudinal survey designs.
- Assess differences between subgroups (e.g., male vs. female) and relationships between variables (correlational research).
The use of surveys provides a number of advantages relative to other types of research, namely:
- Provide a great deal of evidence at a relatively small monetary cost (especially if performed through the internet).
- Provide a way to assess people’s motivations, beliefs, opinions and attitudes.
- Less time and assistance from the researcher(s) are required in data collection.
- Provide a way to collect large amounts of data relatively quickly.
- Ensures standardization of measurement given that all respondents are asked precisely the same questions in the same manner, thereby eliminating the potential for interviewer bias.
- People at distant locations can be reached by post or internet surveys.
- Offer anonymity, making them more desirable for studying sensitive topics.
Disadvantages and limitations
Surveys have some disadvantages and limitations relative to other types of research, namely:
- Do not provide a direct measure of behaviour, but only the participant’s self-reports of that behaviour, that can be prone to bias.
- Self-reports may not be truthful or can be prone to social desirability effects (i.e., people presenting a certain image of themselves to be viewed in a positive way by other people).
- Not suitable for examining implicit or less manifest motivations and attitudes (e.g. associated with racial stereotypes).
- Respondents may perceive response categories as limited, artificial and constraining and/or have difficulty in understanding, due to lack of experience with close answer formats.
- Response tendencies may bias the results.
- If response rates are low, it becomes difficult to generalize to larger populations because of nonresponse bias.
- Usually, results are correlational, thus causality cannot be inferred.
- Can be costly whenever samples with specific characteristics and/or representative samples (e.g. of a country) need to be recruited.
- Sophisticated statistical techniques may be required to analyze large data sets.