Evidence-Based Practice for Nurses

when in reality it does not. Patients begin using the device, which inaccurately
measures blood sugar. Because nurses evaluate the implementation of this new
device, they eventually realize that the device is not effective as the researcher
claimed. The type I error could result in accusations of harming diabetics with a
fraudulent measuring device. Because patients who used the device might have
been harmed, they might want to sue the researcher. However, if a type II error
is made, the researcher throws away the device for measuring blood sugar even
though in reality it is superior to other devices. In this situation, the researcher
misses the opportunity to market the device and earn money, and diabetics miss
the opportunity to benefit from the measuring device. Although neither scenario
is desirable, most researchers would choose to miss the opportunity to make
money rather than to harm patients and risk the legal implications.

Researchers must make decisions about how much risk they are willing to tolerate.
When interventions are complex, expensive, invasive, or have many side effects,
such as a new procedure for cardiac surgery, researchers are usually less willing to
make type I errors. When interventions are simple, inexpensive, or noninvasive,
such as a new teaching method, the tolerance for a type I error is increased.

Researchers use statistics to adjust the amount of risk involved in making
type I and type II errors. It is helpful to remember that type I and type II errors
have an inverse relationship. When type I error is increased, type II error is
decreased. Risks for these errors are adjusted by selecting the alpha level, which
is the probability of making a type I error. Alpha level is designated at the end
of the tail in a distribution (see Figure 13-12). In nursing research, the alpha

KEY TERM

alpha level:

Probability of

making a type I

error; typically

designated as .05 or

.01 at the end of the

tail in a distribution

FIGURE 13-12 Placement of Alpha Level on Normal Distribution

Type II

Error

Type II

Error

Type I Error Type I Error

Alpha = 0.05 Alpha = 0.01

13.7 Reducing Error When Deciding About Hypotheses 359