this text to elaborate on all the tests listed in the table, information about some
of the most common tests is worth remembering. When answering questions
about statistically significant differences between groups, nurses should be fa-
miliar with Chi square, t tests, and analysis of variance (ANOVA). They should
also be familiar with Pearson’s r and multiple regression, which are tests used to
determine if there is a statistically significant correlation among the variables.
Testing for Differences Between Groups
As shown in Table 13-12, a variety of tests is used to analyze data for the
purpose of determining if there is a statistically significant difference between
the groups. When deciding which test to use, researchers must consider the
number of groups to be included in the analysis and at what level variables
were measured. Pilot studies frequently involve one group, whereas classic
experiments and quasi-experiments can include two or more groups. Before
analyzing data, researchers need to consider whether the groups are dependent
or independent to select the correct tests to perform. For example, suppose
a researcher is studying perceptions of health in married couples before and
after a class about stress reduction. The researcher tests both men and women
who are married to one another. Because individuals in a marriage have many
characteristics in common, the data collected should be analyzed using infer-
ential tests for dependent groups. Data from designs using subjects as their own
controls, for example, measuring blood pressures before and after exercise, are
also treated as data from dependent groups. More likely than not, researchers
collect data from independent groups, for example, test scores from two groups
of diabetic patients.
The Chi Square Statistic
Chi square is a very commonly used statistic (Hayes, 1994; Plichta & Kelvin,
2013). Calculated when analyzing nominal and ordinal data, it is a nonpara-
metric test. One reason Chi square is used so often is because it is very useful
for finding differences between the groups on demographic variables. For
example, suppose that a researcher is studying the effect of aromatherapy on
blood pressure and randomly assigns 200 individuals to one of two groups.
The experimental group contains 54 women and 46 men, while the control
group has 46 women and 54 men. By performing a Chi-square test, the
researcher can determine whether the groups are alike on the extraneous
variable of gender. If the groups are not significantly different on gender, it
can be assumed that changes in blood pressure are more likely a result of the
intervention than of gender.
When Chi-square statistics are used, the frequencies that are observed
during the study are compared to the frequencies that would be expected to
KEY TERM
Chi square: A
common statistic
used to analyze
nominal and
ordinal data to
find differences
between groups13.8 Using Statistical Tests to Make Inferences About Populations 365