Niles Hilsenberg

Problem Set 1
Psy 303
1)
Breastfeeding Fights Arthritis
A. This article implies that the act of breastfeeding reduces a womans risk for rheumatoid arthritis.
Specifically, the article suggests that if a women were to breastfeed their child for two or more years,
this would cause a 50% reduction in likeliness of contracting arthritis.
B. An alternative explanation for this phenomena would be that women who breastfeed on average
work less physically strenuous jobs than women who formula feed, thereby reducing their incidence of
arthritis later in life.
CIn order to better test for causality, you could use the following longitudinal experiment:
CI
Create 2 groups, A & B
Group A is a breast feeding condition. Group B is a formula feeding condition.
Randomly assign a sample of women with children who are 1 month old
to either condition A or B.
Pretest for arthritis using a reputable arthritis scale.
Recall each member of the study once a year on a random month of the year for a follow up arthritis
assessment.
After 40 years, compare results. If the groups show no significant differences in incidence of arthritis,
than a preventative relationship between breastfeeding and arthritis is highly unlikely.
Bumper Stickers Cause Road Rage
A. This article implies that bumper sticker increase your aggression/road rage while driving.
B. An alternative explanation could be that people who feel the urge to own bumper stickers strongly
enough to buy them, are inherently more aggressive. Another explanation is that owners of bumper
stickers are statistically younger, and younger drivers are statistically more aggressive.
C. In order to better test for causality, you could use the following experiment:

Create 2 groups, A & B

Group A is a bumper sticker condition. Group B is a no sticker condition.
Randomly assign assign a sample of 50 males and 50 females to either condition A or B.
Pretest for aggressive driving
Then blindfold participants and lead them into their assigned car. Having been blindfolded,
participants will not know whether or not their car has a bumper sticker on it.
Asses aggressive driving after
Compare results.

Crying over spilled semen

A. This article implies that women who engage in unprotected sex are happier because semen
...contains powerful and potentially addictive mood-altering chemicals that presumably make women
happier.
B. An alternative explanation would be that women who are happier tend to prefer unprotected sex,
meaning semen has nothing at all to do with the equation.
C. In order to better test for causality, you could use the following experiment:
Create 3 groups, A, B, and C
Group A is an internal ejaculation group. Group B is an external ejaculation group. Group C is a no
ejaculation group
Obtain a random sample of 50 very lucky men.
Randomly assign a sample of 100,000 women to either condition A, B or C
Pretest for suicidal tendencies and happiness.
After experiment has been completed test again for suicidal tendencies and happiness
Compare results.
Offer apologies to men who were randomly assigned to group C.
2)--------------------------------------------------------------------------------------------------------------------------
2I .
Independent Variable- Fear
Dependent Variable- Test performance
Confounding Variable- J uice (drinking juice vs. not drinking juice)
Eliminate this confound by giving both groups juice to drink, yet only deceiving one group into
believing there is a laxative in their juice.
2II
Independent Variable- Effectiveness of singalong method to teach math
Dependent Variable- Mathematical achievement
Confounding Variable- There was no control video; one class watched a video the other did
not. Also an additional confound could be the time of day that the class was held (A.M vs P.M)
A better study would use a control video; a video shown to the second class that had no mathematical
teaching value whatsoever. As it stands now, no inference can be made about causality because only one
class watched a video each morning. It could simply be that, watching a video reduces stress and helps
children learn better. In addition a trial should be run using two midday classes, taught by the same
teacher. By replicating the A & B conditions more carefully, it creates a better experimental design that
de-confounds more effectively.
2III
Independent Variable- Handling; Early handling
Dependent Variable- Friendliness
Confounding Variable- Amount of time handled
The kittens in group A were handled for 11 weeks. The kittens in group B were handled for 7 weeks.
Adjust the amount of time handled for each group to be congruent lengths of time. (Ex) Kittens in
group A are handled from 2-9 weeks of age. Kittens in group B are handled from 10-17 weeks of age.
Each group is handledfor 7 weeks, however one group is handled earlier. This will help determine if
early handling really is better for the kittens.
2IV
Independent Variable-Persuasiveness in recruiting psychology majors
Dependent Variable- Number of students declaring psychology major
Confounding Variable- Term; number of declarations is measured during different terms.
Historically in academia, rates of major enrollment differ naturally throughout the school year.
Reduce this confound by having all professors teach aclass in fall, winter, and spring. Then analyze the
results and calculate averages using SPSS to determine which professor is really the most persuasive.
3)--------------------------------------------------------------------------------------------------------------------------
3I
Independent Variable- Note takers
Dependent Variable- Effectiveness of note takers
Threat to Internal Validity- Biased Assignment; Students are given the choice when they
first come to the center
To improve this study, students should not and can not be given the choice of whether or not they want
a note taker. This must be randomly assigned.
3II
Independent Variable- Noxious odor
Dependent Variable- Reaction time
Threat to Internal Validity- Differential attrition; ...half the people in the noxious odor
condition are so disgusted that they run out of the room before the task ends.
In order to keep people from fleeing the scenario a different noxious odor could be used, such as sulfur
or an Axe brand body spray. If dropout rate continues to be high, it might be necessary to use a less
noxious odor all together that more people can tolerate. This scenario would entirely rule out using Axe
body spray, something must less noxious must be used.
3III
Independent Variable- Ostentatious verbosity, ie how pedantic the writing is
Dependent Variable- Reading comprehension
Threat to Internal Validity- Instrumentation; The research materials/measurements influence
conditions differently. One condition uses English literature of an unspecified grade level. The
other condition uses a 5
th
grade science textbook. The instrumentation is flawed on the grounds
that one condition uses a science text and one condition uses an English text. It could also be
bias in that the maturity of the subject matter may not be equal (5
th
grade literature vs X reading
level, none was specified for the English condition. College textbooks can be on advanced
subjects, and still maintain laymans terms.)
This experiment could be improved by standardizing the subject matter and grade level of the reading
material for the ostentatious and non-ostentatious conditions.
4)--------------------------------------------------------------------------------------------------------------------------
Color Formatting Example/Key
a. Describe specifically how you would manipulate the independent variable. Be sure to identify
appropriate control or comparison groups (each study must have 1 treatment group and at least 2
comparison groups).
b. Explain how such control/comparison groups enhance the ability to infer causality (i.e., what
alternative
explanations do they address?)
c. What pattern of results would provide the strongest support the hypothesis? That is, if the study
worked
out perfectly, what would the pattern of group means look like (e.g., which group would be highest,
lowest, etc.)?
4A. Country Music and Patriotism
Experiment
1.Create groups A, B, C, and D . Group A is a country music condition. Group B is a techno music
condition, and Group C is a monotone condition. During the C condition, a continuous E note tone is
played throughout testing. Group D is a no music condition.
2.Groups are all tested in identical rooms. They are first pretested for patriotism using a reputable
assessment measure of patriotism. After the surveys are complete, the appropriate music, tone, or lack
there of (conditions A and B, C, and D respectively) is played for 30 minutes. Participants are then re
administered a congruent measure of patriotism.
3.Testing in identical rooms controls for the subconscious effect of the room on patriotism.
4.Introduction of the group A condition, allows us to asses whether or not country music has the
potential to influence patriotism.
5.Introduction of the B condition allows us to asses if country music specifically influences patriotism,
or if it is music in general that influences patriotism.
6.Introduction of the C condition allows us to asses whether it is music in general, or musical tones in
general that have an effect on patriotism.
7.Introduction of condition D allows us to asses whether it is musical tones in general that effect
patriotism, or whether the pretest has sensitized participants to become more patriotic.
8.Ideal results for the study are as follows: Group A would have a statistically significant increase in
patriotism. Groups B, C, and D would all show no change in patriotism. The mean patriotism score
would increase from pretest to post-test in only condition A, not conditions B, C, or D.
4B. Owning A Dog Increases Exercise
Experiment
1.Create groups A, B, C and D. Group A is a Dog condition. Group B is a Ferret Condition. Group C is
a human condition, in which participants are randomly assigned a human companion for the duration of
the testing. Group D is a no companion/pet condition.
2.Throughout a period of 1 year a random sample of participants are randomly assigned to one of the
four conditions. In condition A, participants are given a randomly assigned breed of dog to live with
them for a year. In condition B, participants are given a randomly assigned breed of ferret to live with
them for a year. In condition C participants are given a randomly assigned ethnicity of human to live
with them for a year. In condition D, participants are given no companion or pet.
3.Participants are pretested for the amount of time they exercise on average using a reputable exercise
frequency assessment. After one year of living through the conditions of their randomly assigned
experimental group, the group is recalled for final testing.
4.Randomizing breeds of animals and ethnicities of human companions helps control for differences in
breeds, and racism amongst participants.
5.Introduction of the group A condition allows us to asses whether or not dogs have the to potential to
increase exercise frequency.
6.Introduction of the group B condition allows us to asses whether it is dogs specifically that influence
exercise frequency, or if it is just animals in general that increase exercise.
7.Introduction of the group C condition allows us to asses whether it is animals that increase exercise
frequency or companions in general.
8.Introduction of the Group C condition allows us to asses whether it is companions in general that
increase exercise frequency or if people naturally excise more over the course of time (maturation.)
9.Ideal results for the study are as follows: The mean exercise frequency scores of participants in group
A would show a significant increase during the pretest and the final test. Results would be very strong
and validif groups B, C, and D showed no change in exercise, however a slight increase in exercise
frequency in groups B and C could be interpreted as mediating evidence to support the claim group A
attempts to validate. Therefore if the mean exercise frequency score increased in groups B and C, these
would be acceptable, even possibly favorable results.
4C. Sunglasses cause invisible feelings
1.Create groups A, B, C and D. Group A is a sunglasses condition. Group B is a regular (non
prescription, clear) glasses condition. Group C is a contact lens condition. Group D is a no glasses
condition.
2.For one week, participants are randomly assigned to one of the conditions whereby they will be
required to wear the particular eye accessory they are assigned to from 11am-11pm during their normal
daily interactions, except in the case of condition D in which they will not wear any eye accessory.
3.Participants are pretested for feelings of invisibility using a a assessment measure that rates feelings
of invisibility.
4.Introduction of the group A condition asses whether or not sunglasses effect feelings of invisibility.
5.Introduction of the group B condition asses whether its sunglasses specifically that effect feelings of
invisibility or if it is just glasses in general.
6.Introduction of the group C condition asses whether glasses effect feelings of invisibility, or if its just
eye accessories in general that increase these feelings.
7.Introduction of the group D condition asses if eye accessories in general effect feelings of invisibility
or if participants have developed a sensitivity to the invisibility pretest assessment measure.
8.Ideal results for the study are as follows: The mean invisibility scores from the Group A condition
would increase significantly. All other group means would remain constant. A slight increase in Group
B scores could be interpreted as mediating results, that support the sunglasses hypothesis. Depending
on what background research the researcher is using, that may in fact be a favorable finding.