Page 71 -

56 CHAPTER 3 Experimental design

3.4 INVESTIGATING MORE THAN ONE INDEPENDENT
VARIABLE
3.4.1 FACTORIAL DESIGN

Factorial designs are widely adopted when an experiment investigates more than one
independent variable or factor. Using this method, we divide the experiment groups
or conditions into multiple subsets according to the independent variables. It allows
us to simultaneously investigate the impact of all independent variables as well as the
interaction effects between multiple variables.
The number of conditions in a factorial design is determined by the total number
of independent variables and the level of each independent variable. The equation for
calculating the number of conditions is:

n
C = Õ Va
a=1
where C is the number of conditions, V is the number of levels in each variable, and
∏ is the product of V 1 through V n .
The best way to explain a factorial design and this equation is through an exam-
ple. Consider running an experiment to compare the typing speed when using three
types of keyboard (QWERTY, DVORAK, and Alphabetic). We are also interested in
examining the effect of different tasks (composition vs transcription) on the typing
speed. This suggests that two independent variables are investigated in the experi-
ment: type of keyboards and type of tasks. The variable “type of keyboards” has three
levels: QWERTY, DVORAK, and Alphabetic. The variable “type of tasks” has two
levels: transcription and composition. Therefore, the total number of conditions in
this experiment is calculated according to the following equation:
Numberofconditions =´ =32 6
Table 3.2 illustrates the six conditions in this experiment. In the first three condi-
tions, the participants would all complete composition tasks using different kinds of
keyboard. In the other three conditions, the participants would all complete transcrip-
tion tasks using different keyboards. When analyzing the data, we can compare con-
ditions in the same row to examine the impact of keyboards. The effect of the tasks
can be examined through comparing conditions in the same column. As a result,
the effect of both independent variables can be examined simultaneously through a
single experiment.

Table 3.2 A Factorial Design
QWERTY DVORAK Alphabetic
Composition 1 2 3
Transcription 4 5 6