Survey Design and Analysis

About the course

Back to course listing

This course is aimed at those who need to plan and design questionnaire-based surveys and those who need to analyse the data collected.

The first day covers the vital activities of designing a quality survey and obtaining a good sample. It covers the following topics.

  • Defining your sample
  • Choosing the right sampling methods to obtain a representative sample (e.g. stratified sampling, cluster sampling, multi-stage sampling, …)
  • Deciding on the best way to collect your data (e.g. online, face-to-face, postal questionnaire, …)
  • Designing questions to get good quality answers
  • Deciding on the sample size needed to get estimates that have sufficient accuracy
  • Using effect sizes and power to decide the sample size needed for statistical tests
  • The software G*Power will be used to demonstrate sample size calculations

The second day will cover summarising the results of your survey, presenting the results and undertaking statistical tests. It concentrates on those analyses which are particularly appropriate for the sort of data commonly collected using questionnaires. It covers the following topics.

  • Creating graphs and charts
  • Calculating summary statistics
  • Testing for differences between groups
  • Investigating relationships between variables
  • Fitting regression models
  • The software Excel and SPSS will be used to demonstrate the topics

See our full range of courses (and options for bespoke courses) here or contact the course organiser, Professor Neil Spencer, on 01707 285529,

Why choose this course?

This course will cover the skills needed to design and analyse questionnaire-based surveys, focussing on the practical application of statistical ideas and methods. The software G*Power, Excel and SPSS will be used to demonstrate the topics covered.

Entry requirements...

No prior knowledge of statistics is required.

Course fees

Dates: We are currently assessing whether a quality online delivery of this course is possible.