Evidence-Based Practice for Nurses

Researchers determine whether the results were obtained by chance us-
ing inferential statistical tests, sometimes known as tests of significance.
Mathematical calculations are performed, usually by computers, to obtain
critical values. These values are plotted on a normal distribution, and de-
terminations of whether findings are statistically significant or nonsignifi-
cant are made. Statistically significant critical values fall in the tails of the
normal distributions, usually three standard deviations from the mean or
where about only 0.3% of the data occur. Thus, when critical values are in
that area, researchers believe it is appropriate to claim that the findings did
not happen by chance.

13.7 Reducing Error When

Deciding About Hypotheses

When is there enough testing to make a decision?

CRITICAL THINKING EXERCISE 13-3

KEY TERMS

statistically

significant: When

critical values fall in

the tails of normal

distributions; when

findings did not

happen by chance

alone

nonsignificant:

When results of

the study could

have occurred by

chance; findings

that support the

null hypothesis

1. Inferential statistical tests are used to: (Select all that apply.)
   a. make assumptions about the population.
   b. describe the sample with means and standard deviations.
   c. test hypotheses by asking if there are differences between the groups.
   d. select a sample.
   e. determine whether results occurred by chance.

TEST YOUR KNOWLEDGE 13-6

How did you do? 1. a, c, e

At the end of this section, you will be able to:

‹ Distinguish between type I and type II errors

‹ Describe alpha levels commonly used in nursing research

Researchers do many things to reduce error so that nurses can have confidence
in findings. They reduce error by selecting designs that fit with research ques-
tions, controlling the independent variable, carefully measuring variables, and

13.7 Reducing Error When Deciding About Hypotheses 355