Evidence-Based Practice for Nurses

to smaller clusters or subsets of a population. First, researchers identify all ele-

ments that could be included in a study. Then they determine logical subsets

of the populations using random selection techniques. The participants are

then randomly selected within those groups. For example, a researcher wants

to study baccalaureate nursing education in the United States. Using cluster

sampling, the researcher first randomly selects 10 states that will be included

in the study. Then all baccalaureate nursing programs in each of the 10 states

are identified. The researcher then randomly selects three schools from each

state and obtains the names of baccalaureate students enrolled in each of the

programs selected. Sampling is concluded by randomly selecting 30% of the

names on each list.

Although cluster sampling is an effective method to sample large populations,

there are limitations. In the preceding example, suppose the researcher decided

to select 50 students from each program rather than selecting 30%. This could

result in oversampling of students from small programs and undersampling

of students from large programs. Selecting a percentage of students provides

a more representative sample.

Systematic Random Sampling

A fourth type of probability sampling is systematic random sampling. In this

method, researchers select subjects by creating a numbered list of elements and

then selecting every kth element. Sometimes every odd- or even-numbered

element may be chosen. To determine the sampling interval, researchers must

know two things: the desired size of the sample and the size of the sampling frame.

The formula for determining sample interval is as follows (Gray et al., 2016):

k=Sizeof samplingframe

Sizeof sample

For example, a researcher desires to study infant feeding patterns and has

access to a well-baby clinic. At this clinic, 300 infants are registered as patients.

The researcher needs 20 subjects for the study. Using systematic sampling, the

researcher would obtain a list of names of all infants seen in the clinic and

assign each name a number. The sampling interval would be calculated by

dividing 300 by 20, which equals 15. After randomly selecting a starting point

on the list, every 15th (kth) name would be invited to participate in the study.

Sampling bias may be introduced with this method. This method should

not be used if there is a pattern inherent in the list of elements. For example,

if every 10th name on the list is a baby with formula intolerance, these infants

would either be over- or undersampled, resulting in a biased sample.

KEY TERMS

systematic random

sampling: Sampling

method in which

every kth element

is selected from

a numbered list

of all elements

in the accessible

population; the

starting point on

the list is randomly

selected

sampling interval:

The interval (k)

between each

element selected

when using

systematic random

sampling

294 CHAPTER 11 Using Samples to Provide Evidence