Page 83 -

68 CHAPTER 3 Experimental design

RESEARCH DESIGN EXERCISES

1. Read the following scenarios. Identify actions or conditions that may induce
systematic errors in each scenario and explain the direction of the impact (i.e.,
whether the observed data will be pulled up or down from the actual value).
Scenario 1: In an experiment that investigates how novice users learn to use the T9
method to enter data into a PDA, a participant has actually used T9 for over a year.
Scenario 2: An experimenter is introducing a website to a participant. The
experimenter says, “My team has spent six months on this site. The site is like
our baby.”
Scenario 3: In an experiment that examines how individuals with severe motor
disabilities interact with computers using a brain-computer interface, all
participants recruited are healthy individuals without any disability.
Scenario 4: In an experiment that examines speech-based dictation techniques, the
experimenter forgets to switch the speech profiles between experiment sessions, so
a participant used another person's speech profile to complete the dictation tasks.
Scenario 5: In an experiment that examines the design of an e-commerce
website, participants complete multiple tasks to retrieve specific information on
the site. However, the network speed is very slow and the participants have to
wait significant amounts of time for each page to be loaded.
2. Read the following scenarios. Discuss the type of experiment design (between-
group, within-group, or split-plot) that is appropriate for each scenario.
Scenario 1: A study investigates whether people who have attended a security
training program generate and use more secure passwords than people who have
not received any security training.
Scenario 2: A research team examines the effectiveness of joysticks and
trackballs for selecting static targets and moving targets.
Scenario 3: A research team examines whether the gender of an online agent
affects the perception of trust for young users, middle-aged users, and older users.
Scenario 4: A research team examines whether virtual teams who use
video conferencing are more productive than teams who use phone-based
teleconferencing.
Scenario 5: A study examines the effectiveness of three menu structures. The
first structure has two levels, with 8 items in the first level and 64 items in the
second level. The second structure has three levels, with 4 items in the first level,
16 items in the second level, and 64 items in the third level. The third menu has
n
six levels, with 2 items in the first level and 2 items in the nth level.

REFERENCES
Broome, J., 1984. Selecting people randomly. Ethics 95 (1), 38–55.
Cooper, D., Schindler, P., 2000. Business Research Methods, seventh ed. McGraw Hill,
Boston, MA.