POVOJ NINO v TIRE OVEN NE

SECTION 1.3 PLANS FOR COLLECTION AND ANALYSIS OF DATA

Preceding paragraphs described a guantitative independent variable.
If the treatment levels consisted of unmodulated radiation, amplitude-
modulated radiation, and pulse-modulated radiation, the treatment is
designated as a gualitative independent variable. The different treatment
levels represent different kinds rather than difierent amounts of the indepen-
dent variable. The distinction between guantitative and gualitative treat-
ments is important in connection with trend analysis. The specific levels
of a gualitative independent variable employed in an experiment are
generally of direct interest to an experimenter. The levels chosen are usually
dictated by the nature of the research hypothesis.

CONTROL OF NUISANCE VARIABLES

In addition to independent and dependent variables, all experiments
include one or more nuisance variables. Nuisance variables are undesired
sources of variation in an experiment that may affect the dependent
variable. As the name implies, the effects of nuisance variables are of no
interest per se. In the radiation example, potential nuisance variables
include sex of the rats, variation in weight of the rats prior to the experi-
ment, presence of infectious diseases in one or more CageS where the rats
are housed, temperature variation among the cages. and differences in pre-
vious feeding experiences of the ratS. Unless controlled, nuisance variables
can bias the outcome of an experiment. For example, if rats in the radiated
groups suffer from some undetected disease, difierences among the groups
would reflect the effects of the disease in addition to radiation effecis—
if the latter effects exist.

Four approaches can be followed in controlling nuisance variables.
One approach is to hold the nuisance variable constant for all subjects.
For example, use only male rats of the same weight. Although an experi-
menter may attempt to hold all nuisance variables constant, the probability
is high that some variable will escape his attention. A second approach,
one that is used in conjunction with the first, is to assign subjects randomly
to the experimental conditions. Then known as well as unsuspected sources
of variation or bias are distributed over the entire experiment and thus
do not affect just one ora limited number of treatment levels. In this case
an experimenter increases the magnitude of random variation among
observations in order to minimize systematic eflects, that is, the effects
of nuisance variables that bias all observations in one or more treatment
levels in the same manner. Random variation can be taken into account
in evaluating the outcome of an experiment, whereas it is difficult or
impossible to account for systematic nuisance effects. A third approach to
controlling nuisance variables is to include the variable as one of the treat-
ments in the experimental design. This approach is illustrated in Section
1.4 in connection with a Latin sguare design.

The above three approaches for controlling nuisance variables
illustrate the application of experimental control as opposed to the fourth