Define the formal probability and informal non-probability sampling approaches

Question 1

Introduction

Answer 1

There are two general types of sampling approach: formal sampling and informal sampling.

Question 2

Formal/Probability sampling

Answer 2

Formal sampling (also known as probability sampling) is based on the assumption that every member of the sample has a known and non-zero probability of selection (these do not have to be equal, although this is the case with simple random sampling, the most common form of probability sampling). A sampling frame is an indispensable part of formal probability surveys.

Definition of the sampling frame is particularly important as an incomplete or inaccurate frame will lead to a poor sample, one that is biased in that repeated samples will consistently under-represent some members of the target population and over-represent others.

Question 3

Informal/Non-probability sampling

Answer 3

Informal sampling (also known as non-probability sampling) refers to any sampling procedure that does not give specified values to the selection probabilities, and thus probabilities of selection of some members of the target population are unknown and / or may be zero. Non-probability sampling methods are used when a sampling frame for the population is not available, when a probabilistic approach is not necessary, or for convenience/ethical reasons. Informal sampling methods normally avoid the time, cost, and other difficulties involved in making a frame, but in doing without a frame they run the same (and perhaps greater) risks of giving biased samples.

Question 4

Researchers need to understand both approaches

Answer 4

Researchers need to understand both formal probability and informal non-probability sampling approaches so that they can:

• choose between different methods and select combinations appropriate to the particular theoretical and practical demands of each research study
• recognise the strengths and weaknesses (particularly the latter) of sampling in undertaking their own research and in assessing the validity of other people’s research.

Question 5

Probability sampling to avoid bias

Answer 5

Thus, the probability methods of sampling are designed to avoid selection bias (ie provide a representative sample) so that if a number of different samples are taken, then although these samples may have different means, an average of the means of all possible samples will be equal to the population mean.