It is also important to differentiate between statistical
significance and clinical significance. For example, suppose
a drug lowered cholesterol levels on average from 195 to
- Analysis indicated that this decrease was statistically
significant; however, because any cholesterol value below
200 is considered to be within normal range, there is no
clinical significance to this finding. In another example,
consider an experimental study involving guided imagery.
Children in the guided imagery group had average pain
ratings of 4.2, while children in the control group had average pain ratings of
5.2. Although the difference between these means is not statistically signifi-
cant, it may be clinically significant to have pain ratings a whole point lower.
Furthermore, because guided imagery was noninvasive and inexpensive and
children reportedly enjoyed it, it might be worth incorporating as a practice
intervention. When appraising evidence, it is wise for nurses to keep in mind
that statistical significance and clinical significance are different.
Appraising the results section of research reports can be challenging. An
understanding of the material presented in this chapter can assist you in meeting
this challenge. Information in Box 13-2 and Figure 13-14 should be considered
when evaluating data analysis.
Be familiar with symbols used.
Read the title of the table first.
Pay attention to the labels of columns and rows.
Observe when headings for subsamples and subscales are indented.
Follow columns carefully both horizontally and vertically.
Attend to significant findings.
Alpha levels are either reported with a notation or under the table
with an asterisk (*).
Pay attention to superscript or footnote markings.
Do not skip tables, because information can be clearer after studying
a table.
Recognize that there can be typos and math errors.
Remember that subsamples may be omitted from the table
(i.e., sample indicates total subjects and number of women, leaving
reader to calculate number of men).BOX 13-1 Tips for Reading Statistical Tables
FYI
Although many readers frequently skip over
the tables when reading research articles,
this is not a good practice. Data presented
in tables may contain evidence on which
practice changes can be made, and thus may
be one of the most important components
of a report.13.9 What Does All This Mean for EBP? 371