SOCB05 – Lecture 4 – May 30 2013
Logic of Sampling
Who do you want to generalize to? – Theoretical Population
What population can you get access to? – Study Population
How can you get access to them? – Sampling Frame
Who is in your study? – Sample.
Things to consider:
o Can I study everyone?
o How do I know who to study?
o How do I pick my sample?
o Will it represent everyone? Does it need to?
o What limitations can I put on my sample?
We need a process to select participants/observations on which to carry out a
Studying a whole population may be cumbersome/impossible.
o Need a small group that approximates the population
Two Types of Samping
o was born in the 1940s, and remains the most common for gaining
large representative samples.
Not always the best because some topics that are not suitable or
o Most common in social research, particularly quantitative
o Representative of larger population. Elements of a population stand
equal chance of selection
o More complex.
o Common in qualitative studies
o Vulnerable/hard to reach populations?
o May be unrepresentative. Elements of the population do not stand
equal chance of selection
o Usually used in complex phenomenon that are not seeking
o Used on groups that are vulnerable, or hard to reach.
Example: Abused children, prostitutes, etc. People that there
isn’t really a list of people fitting the category.
Methods of Non-Probability Sampling
o Aka “reliance on available subjects”
o Cheap, convenient, and easier to find.
o Use of those available at a particular place and time.
o Useful for pretesting questionnaires or other social measurements o Not representative of a larger population
Difficult to generalize
o Example: Picking people who walk past a store right near an entrance
where the public transit is located.
o Sample selected on the basis of your own knowledge of the population
you intend to study and your research question
o Useful when studying a small subset of a larger population, where
members of the subpopulation are easily identified.
o Already recruited respondents provide the researchers with assistance
in locating other members of the population under study.
o Useful when members of a population are particularly hard to locate
Example: Homeless people.
o Exploratory purposes, as an introduction into the field.
o Representativeness is problematic
People tend to refer people that are like them. So there will be a
bias because the participants will be similar to each other.
o Usually starts with one connection, who will share others that they
know of. And it just keeps going from there.
o The non-probability version of probability sampling.
o Begins with a table describing the characteristics of the target
o Try to assign equal proportions of people who belong to different
groups to your sample.
o Units are selected into the sample based on a specified list of
Methods of Probability Sampling
Used when researchers want precise, statistical descriptions of large
“to provide useful descriptions of the total population, a sample of individuals
from a population must contain essentially the same variations that exist in the
Representativeness= the quality of a sample of having the same distribution of
characteristics as the population from which it was selected
o Provides generalizability
When its not representative, it introduces bias:
Not using a probability sampling method
Having an inadequate sampling frame
o Allows for use of inferential statistics
o Equal Probability of Selection Method: When everyone has an equal
chance of being selected. Population: Theoretically specified aggregation of the elements in a study.
Basically, all of the people that you are interested in. If you could study
everyone, who would be included in your study?
Sample: Portion of the population from which information is colle