Seatwork-5-chapter-8
Seatwork-5-chapter-8
Seatwork-5-chapter-8
Physical variables are elements of the testing conditions that need to be controlled (the
room, the noise, etc.)
2. How is elimination used as a control procedure? Give two examples of variables that
could be controlled by elimination.
Elimination is a technique to control extraneous variables by removing them from an
experiment.
If noise might confound the
results, we test in a quiet location or soundproof rooms.
If we do not want interruptions, we hang a sign on the door saying, “Do not disturb.
Experiment in progress.”
3. What is constancy of conditions? Give two examples of variables that could be controlled
through constancy of conditions.
If room color might confound results, have subjects take test in exact same room.
keep lighting the same, keep comfort of chairs the same, keep everything the same
except for your manipulation
have voice recording of participant instructions talk at the same speed, stay at the same
volume, etc. for both conditions
If you only have access to one quiet room and a loud room, randomly assign participants
to a room
randomly assign subjects in each room to get treatment A, then the rest will get B.
5. You are doing a study at a local school. Because of the way things are scheduled, you
can have one small testing room in the morning and another, much larger testing room in
the afternoon. If you have two treatment conditions (A and B), how can you assign
subjects to the testing rooms so that the type of room will not lead to confounding in your
experiment?
You can use balancing to control for effects of the two different rooms and times of day.
Randomly assign half the subjects to the small room (the other half will go to the larger
room). Randomly assign half the small-room subjects to A and half to B; do the same for
the larger room.
6. What are demand characteristics? How do they affect our data? How can they be
controlled?
Demand characteristics are the aspects of the experimental situation itself that demand
or elicit particular behaviors; can lead to distorted data by compelling subjects to
produce responses that conform to what subjects believe is expected of them in the
experiment.
8. Dr. L. is planning a large-scale learning experiment. He would like to have 100 rats in
one treatment group and another 100 in the other group. Because he needs so many
rats, he says, “Well, I can’t test all these animals by myself. I’ll ask Dr. P. to help me. He
can run the animals in the one group while I test the animals in the other group.”
a. Knowing what you know about confounding, is Dr. L.’s solution a good one? What can
happen if one experimenter tests all the subjects in one group while another tests all the
subjects in another group?
Dr. L’s idea is a poor one because each treatment condition would be run by a different
experimenter, confounding the experiment. It would not be possible to tell if effects were
produced by the two different treatments or by the two different experimenters.
b. Given what you know about balancing procedures, work out a better plan for Dr. L.
Instead, randomly assign half the rats in each condition to Dr. L and the other half to the
additional experimenter.
11. In what ways are volunteer subjects different from nonvolunteers? What difference
does it make?
Volunteers are usually more liberal, have higher IQs, and be less authoritarian. This
affects their responses and doesn't give a good population representation.