6 Evidence Appraisal: Research 129nursing literature, the p-value for determining statistical significance is generally
set at p < 0.05. Though p-values indicate statistical significance (i.e., the results
are not due to chance), effect sizes increase readers’ comprehension of the size of
the differences found (clinical significance). Nursing research results are increas-
ingly reporting effect sizes (clinical significance) and confidence intervals (preci-
sion measure for effect size) to more fully interpret results and guide decisions for
translation.
Effect size is an important concept in EBP because it provides a common metric
for summarizing evidence in meta-analysis. When combining studies in a meta-
analysis, the effect size measures the strength of the relationship or association
between variables in the research studies’ combined populations. In quantitative
studies other than meta-analysis, effect size is a more precise measure of the mag-
nitude of the difference between groups (Sullivan & Feinn, 2012). Both are es-
sential to the interpretation of results. An effect may be the result of a treatment,
a decision, an intervention, or a program. The most commonly used estimate of
the size of an intervention effect is Cohen’s d, whereby 0.8 is considered to be a
large effect, 0.5 is interpreted as a medium effect, and 0.2 equates to a small ef-
fect (Cohen, 1988). Consider that an experimental group of 20 new graduates is
given a competency test on assessment of cardiac function after having a simula-
tion experience. Their average score is 80 with a standard deviation (SD) of 10;
the control group, which did not have a simulation experience, scores 75 with a
standard deviation of 12. In this example of two independent groups, the effect
size is calculated as follows ([d = mean 1–mean 2/ SD of either group]): 80 minus
75 divided by 10 equals 0.5, which is a medium effect size.
Confidence intervals (CI) address one key EBP question for appraising the evi-
dence: How precise is the estimate of effects (Polit, 2010)? What exactly is the
confidence interval (CI)? This measure of precision is an estimate of a range of
values within which the actual value lies. The CI contains an upper limit and a
lower limit. A 95% CI is the range of values within which an investigator can be
95% confident that the actual values in a given population fall within the upper
and lower limits of the range of values. Consider a study on the effect of sucrose
analgesia used to manage pain during venipuncture in newborn screening