6 Evidence Appraisal: Research 107
Quasi-Experimental Studies
Quasi-experimental studies are similar to experimental studies in that they try to
show that an intervention causes a particular outcome. These designs are used
when it is not possible to meet two of the three requirements for an experimental
study. Quasi-experimental studies always include manipulation of the indepen-
dent variable. They differ from experimental studies because it is not possible
to have either a control group or random assignment of subjects. For example,
an investigator can assign the intervention (manipulation) to one of two groups
(e.g., two medical units), with one serving as the intervention group and one serv-
ing as the control group. The investigator cannot randomly assign participants to
the units.
In cases where a particular intervention is effective, withholding that interven-
tion would be unethical. In the same vein, it would not be practical to perform
a study that requires more human, financial, and material resources than are
available. At times, neither patients nor geographical locations can be random-
ized. Consider the investigator who wants to study the effect of bed-exit alarms
on patient falls. It would not be easy to randomize the use of bed-exit alarms to
individual patient rooms or to individual patients. Nursing staff likely would not
agree to deactivate bed-exit alarms on at-risk patients whose beds are equipped
with this safety feature. Quasi-experimental designs that an EBP team may un-
cover during the course of its search include nonequivalent control (comparison)
group posttest only and nonequivalent control (comparison) group pretest–
posttest. The term nonequivalent means not only that assignment is nonrandom
but also that the researcher does not control assignment to groups. This design
is most often used when doing research with intact groups. Hence, groups may
be different on important variables such as health status or demographics, and
group differences may affect outcomes. Other quasi-experimental research study
designs are not included in this chapter. EBP team members should refer to a re-
search text when they encounter any unfamiliar study designs.
