Design and conduct a survey

https://doi.org/10.1016/j.ctim.2004.12.005Get rights and content

Introduction

Surveys are used to collect information on a broad range of subjects of interest in fields as diverse as health, sociology, psychology, economics, market research, business and law.1 They consist of a number of written questions and are completed by ‘respondents’. A survey is used when large numbers of respondents are required, when questions are relatively simple and easy to respond to, and when the study does not require detailed clarification or exploration of responses. A large number of surveys have been conducted in complementary medicine, many focusing on trying to find out why people use complementary approaches (and their characteristics).

My experience has been that a survey is often chosen for an undergraduate or masters dissertation project, perhaps on the assumption that a survey is easy to carry out. This is not true: a worthwhile survey needs care and skill in designing and conducting; although it may not take much time for respondents to complete, yet it is an intrusion and therefore has ethical consequences; and it should only be considered if the outcome is likely to be valuable by contributing genuinely new knowledge.

Before you think about conducting a survey you need to carefully consider the purpose of your research. Your research question should guide your method, rather than the other way round.

Section snippets

Step 1. Defining the objectives

It may seem obvious that defining the research objectives are the first step in carrying out a survey. However, as with any research approach, this may be difficult, as ideas that are general and abstract need to be refined into very specific questions. Focus your study objectives so that you can choose the most appropriate method of data collection. For example you may want to find out how patients experience a treatment and if they are satisfied with it. It is hard to really explore patient

Survey sampling and response

Information in a survey is obtained from a sample of participants, drawn from the population (all the people you are interested in). In order to be able to make generalisable statements about the target population the sample should represent that population, that is, share the same characteristics. You need a systematic procedure for sampling your population so that individuals with particular characteristics do not have a higher chance of getting into your sample than others do, as this may

Survey analysis

Descriptive statistics are used to summarise and present the data, focussing on general trends and important points. Such data are often presented as graphs or tables. In contrast, inferential statistics are concerned with hypothesis testing and making statements about the population based on your sample. This will involve testing for differences or relationships between variables. There are a number of good resources on survey analysis.2, 12, 13 Krishanmurty et al.10 have produced an

Conclusion

Carrying out a survey is not simple, and needs careful preparation particularly if you need to design your own questionnaire. The input of a statistician from the outset is highly recommended and helps to avoid the pitfalls of collecting data that proves to be useless and impossible to analyse. If possible, use questionnaires that have been tested for validity and reliability, and which may be appropriate for your population. Think carefully about your sampling procedures, and how you will

First page preview

First page preview
Click to open first page preview

References (14)

  • K.J. Thomas et al.

    Use and expenditure on complementary medicine in England: a population based survey

    Complement Ther Med

    (2001)
  • A. Fink

    How to design surveys

    (1995)
  • J. Peat

    Health science research: a handbook of quantitative methods

    (2002)
  • A. Bowling

    Measuring health: a review of quality of life measurement scales

    (1997)
  • A. Bowling

    Measuring disease

    (2001)
  • J. Richardson

    Developing and evaluating complementary therapy services. Part 2. Examining the effect of treatment on health status

    J Altern Complement Med: Res Paradigm, Practice, Policy

    (2001)
There are more references available in the full text version of this article.

Cited by (13)

View all citing articles on Scopus
View full text