Type Alan Bryman author names here Social Research

Type Alan Bryman author names here Social Research Methods Chapter 18: Sampling in qualitative research Slides authored by Tom Owens

What is purposive sampling? • The form of sampling typically used in qualitative research. • Strategic in nature: individuals/cases are selected on the basis of their relevance to research questions. • Not possible to extrapolate results to the general population. • Important to have clear and specific criteria to determine inclusion/exclusion of units of analysis. Pages 416 -418 Bryman: Social Research Methods, 4 th edition

Types of purposive sampling: 1. Theoretical sampling • Glaser & Strauss (1967) – Probability sampling inappropriate for qualitative research – Aim to discover categories and their properties – Ongoing process: collecting, coding and analyzing data, deciding where to look next (purposive sampling) and developing grounded theory – Theoretical saturation point is reached when categories and concepts are dense enough; no more data collection is needed. Key concept 18. 3 Page 419 Bryman: Social Research Methods, 4 th edition

The process of theoretical sampling Figure 18. 1 Page 420 Bryman: Social Research Methods, 4 th edition

Types of purposive sampling: 2. Generic purposive sampling • Criteria for selecting individuals/cases determined by research questions • These criteria are typically determined in advance of sampling and remain fixed throughout the process • Appropriate individuals/cases are identified • A sample of these is then selected. • In mixed methods research, survey findings can be used as the basis for purposive sample selection. Page 422 Bryman: Social Research Methods, 4 th edition

Types of purposive sampling: 3. Snowball sampling? A sampling technique that can be useful when there is no sampling frame, making probability sampling impossible A small group of people are selected initially, on the basis or research relevance Then these participants propose others with similar characteristics or experience Can be useful when networks of individuals are the research focus. Page 424 Bryman: Social Research Methods, 4 th edition

Sampling in ethnography • Often a combination of convenience and snowball sampling • May involve purposive sampling • Researcher has to glean information from whoever is prepared to divulge it • Stratified sampling might be possible Page 424 Bryman: Social Research Methods, 4 th edition

Sample size • Difficult to establish, in advance, how many people should be interviewed, • Impossible to say, in advance, when theoretical saturation point will be reached • Larger samples are needed when the qualitative study has a broad scope and when inter-group comparisons are required by the research • A minimum number of twenty to thirty interviews is recommended by Warren (2002) • Great care is needed when making inferences from a purposive sample. Pages 425 -427 Bryman: Social Research Methods, 4 th edition

Sampling is not just about people • Documents can be selected for their relevance to research questions • Time periods need to be sampled if observations are made at different times of the day or on different days of the week • Contexts need to be sampled if observations are made in different locations (Hammersley & Atkinson, 1995) Page 437 Bryman: Social Research Methods, 4 th edition

Sampling in ethnography • Often a combination of convenience and snowball sampling • May involve purposive sampling • Researcher has to glean information from whoever is prepared to divulge it • Stratified sampling might be possible • Not just people – Sampling time units (observing at different times) – Sampling contexts (observe in different locations) (Hammersley & Atkinson, 1995) Pages 441 -444 Bryman: Social Research Methods, 4 th edition