1 Research Methods CJ490 Susan Wind Welcome!
2 Sampling The MOST important part of research process
3 Sampling Importance If poor sampling, can’t generalize to a population Statistical techniques may be used incorrectly Policy based on research with poor samples may not be valid
4 What is Sampling? ??? ANYONE?
5 Sampling is…. Ways to identify a subset of the population that has the same characteristics of the population
6 Sampling “Population” Group of elements from which sample is drawn May be based on a concept, such as Drug Users Domestic Violence Victims May be specific and have a list Drug uses in a program Victims in a DV shelter
7 POPULATION SAMPLE 5 ELEMENTS
8 Sampling Types Probability Non-Probability
9 Probability Sampling Rely on random selection; chance Mathematical formula can identify the likelihood that a specific element would be selected Formula can estimate how much the sample will differ from the population
10 Non-Probability Sampling Probability of an element being chosen cannot be determined Often used in social science research Population list is usually not known
11 Strengths of Non-Probability Sampling Can be used if no list for population, which is not uncommon Particularly useful for exploratory research Also useful for field research
12 Weaknesses of Non- Probability Sampling Can’t know if sample accurately represents the population
13 Non-Probability Types Availability Sampling Quota Sampling – accidental sampling with requirements for a number of respondents in specific group(s) Purposive or Judgment Sampling Snowball Sampling
14 DV victims Survey You want to study why some women in a shelter for victims of domestic violence return to their batterer. What kind of non probability sampling will you use? Explain your answer.
15 Availability Sampling Non-probability sampling technique Participants just “happen” to come in contact with interviewer Interviewer controls the time and place of the sample, but not who might be available to participate There is no way of knowing if the sample represents the whole population
16 Snowball Sampling Start with few members Ask those sampled to refer others Excellent for exploratory designs Focus is generally on qualitative information
17 Choosing a Sampling Technique Purpose of study – MOST important factor Scope of study Large study – Probability sampling is preferred and often “doable” Small study- depends on other factors Population characteristics: Extremely important: Is there a list of everyone in the population? Must have a list to do probability sampling Cost. Convenience
18 Questions? Virtual Office Imail Phone if an emergency Contact information in welcome and in an announcement
19 Research Methods CJ490 Susan Wind Thanks for Participating!