Another factor that researchers need to consider when conducting quantita-
tive studies is attrition rate, also known as dropout rate. Typically, there will
be subjects who agree to participate in a study and, for whatever reason, do not
complete the study. Attrition is a threat to internal validity, and the researcher
should attempt to create a study design that minimizes loss of subjects. One
method to address attrition is to increase the number of subjects in the sample
so that even with losing subjects, the researcher will have a sufficient sample
size for the research study (Albert et al., 2012).
The design of the study also influences decisions about sample size. The
number of variables being investigated is an important factor. Typically, the
greater the number of variables being tested in a study, the larger the sample
size needed to detect changes in the variables. The sensitivity of instruments
used to collect data also affects the sample size. A very precise instrument
typically requires fewer subjects than does a less precise instrument (Bloom &
Trice, 2011). In addition, decisions about sample size are influenced by practical
matters such as cost, convenience, and feasibility.
There is one accepted rule for determining sample size when using quanti-
tative research designs (Bloom & Trice, 2011). The rule of 30 is used by many
quantitative researchers (Gray et al., 2016). This rule states that in order to
have a sufficient sample size to adequately represent the target population,
there needs to be a minimum of 30 subjects in each group being studied. For
example, in an experiment using a control group and an intervention group, a
sample size of 60 would be indicated. The rule of 30 should be considered the
minimal number of subjects in each group.
A more powerful and accurate method to determine sample size for quan-
titative studies is to conduct a power analysis. A power analysis is a statistical
method used to determine the acceptable sample size to detect the true effect
or difference in the outcome variable (Houser, 2011). When nurses read that
a power analysis was conducted and used correctly, they can have greater as-
surance that the sample size was appropriate for the study and be confident
applying the findings to the target population (Polit & Beck, 2014).
Two factors must be established to conduct a power analysis: significance
level and effect size. The significance level is the alpha level established prior to
the beginning of the investigation. A vast majority of nursing researchers use
p = .05 as the significance level. The effect size is an estimate of how large a dif-
ference will be observed between the groups (Hayat, 2013). When researchers
expect that the effect size is large, fewer subjects are needed to detect differences
between the groups. If the effect of an intervention is small, a larger sample is
needed to statistically demonstrate that the intervention was effective. A review
of relevant literature on the research topic can assist the researchers in identify-
ing the effect size (Albert et al., 2012).KEY TERMS
attrition rate:
Dropout rate; loss
of subjects before a
study is completed;
threat of mortality
power analysis: A
statistical method
to determine
the acceptable
sample size that
will best detect
the true effect of
the independent
variable
significance level:
The alpha level
established before
the beginning of a
study
effect size: An
estimate of how
large a difference
will be observed
between the groups300 CHAPTER 11 Using Samples to Provide Evidence