Module Components

• Learning Activities

Topics

• Sampling Overview
• Sampling for Quantitative
• Sampling for Qualitative
• Measurement
• Levels of Measurement
• Reliability and Validity of Measures
• Estimating Reliability and Validity
• Modue 3 Self-Review

NRSG 790 - Methods for Research and Evidence-Based Practice

Module 3: Research Methods

Sampling for Quantitative Research

Overview

Sampling is the process of selecting a portion of a population about which information is collected. Researchers sample from the population that is accessible and meets predetermined inclusion and exclusion criteria to participate in the study, and generalize to a target population.

Bias

Researchers work with samples because they are not able to study all members of a population. Samples may or may not be representative of the population, resulting in the possibility of bias. Sampling bias is the over or under representation of a population characteristic, risking construct and external validity.
Sample size is an important factor in achieving statistical conclusion validity. The larger a sample, the more likely it is to be representative with less bias. A large sample does not negate the potential error introduced through a flawed design.
Power analysis a priori is typically used to estimate sample size needs. Power is a function of the effect size, alpha, and sample size. For example, if a researcher designs a study with a desired power of .80, a significance level of .05, and a moderate effect (.40), the unknown variable, the sample size can be determined.

Techniques

A representative sample cannot be guaranteed without obtaining information from the population; therefore researchers aim to minimize bias through study design and sampling technique. There are two main categories of sampling, probability sampling and non-probability sampling.
Probability Sampling Techniques: These use some sort of random sampling in an effort to ensure that the characteristics of the sample studied resemble those of the population of interest. These techniques are typically used in quantitative research designs. While not practical, probability sampling is the best method to obtain a representative sample.

Type of Saampling	Strategy
Simple Random	Each member of the population has an equal chance of being selected
Systematic Random	Beginning with a random starting point, every kth case is selected from a list
Stratified Random	Each case is first assigned to a group (stratum), such as male or female. Equal or proportional numbers are then randomly selected from each group
Cluster	Selecting a random sample of a group (nursing schools) and then sampling from the group (nursing students)

Review this source for additional details on probability sampling http://www.socialresearchmethods.net/kb/sampprob.php

Non-probability Sampling Techniques: These do not use random sampling therefore all cases do not have an equal chance of being included in the study; however, that doesn’t necessarily mean that the samples produced do not represent the population of interest. Non-probability techniques are practical, however carry an increased  risk of bias and threat to generalizability. Some of these techniques are used in both quantitative and qualitative research designs.

Type of Saampling	Strategy
Convenience	Select cases based on availability
Quota	Convenience subgroup (men) selected from a stratum (gender)
Purposive	Judgment sampling, uses researchers knowledge of the population to select the sample (clinical experts in a field)
Snowball	Network sampling, included cases refer others who meet inclusion criteria

Review this source for additional details on non-probability sampling http://www.socialresearchmethods.net/kb/sampnon.php