We want to know if a nurse’s perception of EBP
depends on their organizational level. A chi-square test is the appropriate
method for comparing two nominal or ordinal variables. The null hypothesis
for this test is that perception is independent of organization level.
The alternative hypothesis is that perception depends on organization
level.
If the null hypothesis
is true, that there is no difference in perception by organization
level, then we would expect to see the same proportion in each rating
category for both the nurse professionals and the nurse leaders.
If the proportions are not close between the nurse professionals and
the nurse leaders, we would suspect that their perceptions of EBP
differ. As with other hypothesis tests, we need to ascertain that
the differences we see are statistically significant and not due to
sampling error. We will use a chi-square test statistic. This measures
the discrepancy between the frequencies we observed in our sample
and what we would expect under the null hypothesis. The chi-square
test statistic follows a chi-square distribution.
When comparing two nominal/ordinal
variables, it is customary to present the data in a contingency table.
In our case, we have two organization levels and five response levels,
so our contingency table will have two rows and five columns. Each
“cell” of the table contains the frequency for each
combination of organization and response level. An important assumption
in conducting a chi-square test for independence is that no more than
20% of the cells can have expected frequencies of five or less. We
also assume that the survey responses are independent. The chi-square
test concerns the independence of the two factors – organization
level and perception. The assumption of independence has to do with
how the data was collected, i.e., a participant’s response
is not influenced or related to the response of another participant.