Quantitative Research 1 Sampling and Surveys Dr N

Quantitative Research 1: Sampling and Surveys Dr N L Reynolds

Quantitative Research 1 Lecture objectives • To provide an overview of the sampling process • To understand how sample sizes are determined • To examine the impact of sampling on the analysis of quantitative data • To look at what is needed for good descriptive research • To highlight some issues and stages in the design of a research instrument

Importance of Sampling and Surveys • Sampling is the “who” of research: brilliant questions (or observations) of the wrong sample equals research that does not fulfill its research objectives • Surveys provide a means to ask large samples the same set of questions efficiently and effectively. They allow “populations” to be described. • Research instruments provide a template for data collection; well designed research instruments make the researchers job easier

Key Issues 1. 2. 3. 4. How do should we go about sampling? How can sampling affect the data? What is a survey? How is a good (quantitative) research instrument designed?

Key Issue 1: The sampling process 1. Define the population of interest and, if possible, identify the sampling frame(s) 2. Determine the sampling objectives 3. Select the sampling procedure 4. Determine the sample size 5. Select the sampling elements 6. Collect the data from the designated elements

Key Issue 1: Sampling procedures Non-probability methods • • Convenience sampling Judgment sampling Snowballing Quota sampling Probability methods • Simple random sampling • Stratified sampling • Cluster sampling NB: Estimation of sampling error requires a probability sample

Key Issue 1: Determining sample size • Practical – Resources available (time, money, personnel) • Historical • Research objectives – Planned analysis… • Statistical – Sampling distribution of the statistic (homogeneity of the population), precision, confidence

Key Issue 1: Calculating sample size • Using standard deviations n = Z 2 σ2 E 2 • Using proportions n = Z 2 [P (1 -P)] E 2 n = sample size Z = desired confidence interval (within 1, 2 or 3 standard deviations of the mean on normal curve) E = acceptable level of sampling error σ = standard deviation in the population P = proportion in the population

Key Issue 2: Impact of sampling on analysis • Non-parametric and parametric tests • Requirements of specific data analysis technique – T-tests, multivariate regression, factor analysis, conjoint analysis, … • Managerial confidence in the results

Key Issue 3: What is survey research? • Direct communication with the respondents (sample) • Collecting data using a research instrument (usually a questionnaire)

Key Issue 3: Survey research methods The Research Instrument • Level of structure of the research instrument • Level of disguise concerning the subject of the research Basic Data Collection Methods • Face-to-face • Telephone • Mail • Electronic

Key Issue 3: Which method should be used? • • • Absolute & variable costs Speed of responses Sample issues Response rates Data quality – Reliability & validity

Key Issue 4: Questionnaire Design 1. 2. 3. 4. 5. 6. 7. Develop question topics Select question and response formats Select question wording Determine sequence Design layout and appearance Pilot test Undertake survey End

Key Issue 4: Individual Questions • Is the question necessary? • Several questions or just one? – double questions • Can respondents answer the question? • Will respondents answer the question? • Is the question clear and unbiased? – type of language used – ambiguity in item wording – leading the respondent to a particular response 14

Sampling And Surveys: Their contribution to your dissertation • All will use samples of some sort – people, organisations, countries, events, things (papers) – having knowledge of the sampling process ensures your ability to justify how you have sampled • All will read papers where sampling has been used; being able to critique the sampling will contribute to the critical analysis of the study concerned. • Most will use a research instrument for their dissertation – for interviews, for focus groups, for surveys – knowing how to design the instrument well will contribute to the quality of the research conducted • All will read papers that use a research instrument of some kind; being able to critique the design will contribute to the critical analysis of the study described