Deck 1: Introduction to Social Science Research Principles and Terminology

Question 1

Suppose I want to know whether adult men and women in America differ in their ability to remain faithful to their romantic partner (i.e., not cheat on them). So I select a random sample of 200 college undergraduates (100 men, 100 women) from around the country and ask them if they have ever cheated on a romantic partner. Sixty percent of the men in my sample admit to having cheated whereas 45% of the women said they had been unfaithful in the past. Please answer the following questions based on this study.
a. What is the dependent variable in this study?
b. What is the independent variable in this study?
c. How is the independent variable measured (i.e., interval-ratio, nominal/categorical, ordinal)?
d. In the question I said that I selected my sample(s) randomly. What exactly does that mean?
e. What is/are the population(s), as specified by the research question?
f. What is/are the sample(s)?
g. Do you think the sample is a good representation of the population? Explain your answer.
h. What type of study is this (experimental, quasi-experimental, correlational, observational, etc.)? Tell me why you think so?
i. Is this study descriptive, inferential, or both? Explain your answer.
j. Do you believe the results of this study will generalize? Explain your answer.

Answer 1

A. Faithfulness, or whether one cheated on a partner.
B. Sex/gender (male, female)
C. Nominal/categorical
D. It means that every member of the population, i.e., every adult American man and woman, had an equal chance of being selected to be in the sample.
E. Adult men and women in America. Or, two populations: adult American women and adult American men.
F. 200 college students, half male, from around the country. Or 100 women and 100 men (2 samples).
G.No. College students tend to be younger, wealthier, and more educated than most American adults. May influence mating opportunities, morality. College brings large numbers of men and women together without parental supervision, and that can certainly influence cheating behaviors.
H. Correlational. Nothing is being manipulated so it cannot be experimental or quasi-experimental, and nothing is being observed. It is survey data.
I. Meant to be inferential because want to make inferences about the population from the sample. But could argue it is both as sample statistics will be described.
J. I would say no because the sample is a bad representation of the population. On the other hand, I do not have evidence that wealth, age, education level, and the college environment are related to willingness to be unfaithful, so maybe results will generalize. I believe the research indicates that college students are less faithful than older Americans, but I am not sure. I would have to look it up.

Question 2

Suppose that I want to know how many hot dogs the average American adult eats in a year. So, I go to a Giants baseball game and ask the first 100 fans I meet how many hot dogs they eat in a year. I find that the average member of my sample eats 250 hot dogs a year.
a. What type of sample have I selected (i.e., random, convenience, representative)?
b. Who is my population supposed to be, according to my research question?
c. What type of study is this, descriptive or inferential?
d. Do you think my results will generalize to the larger population? Tell me why or why not.

Answer 2

A. Convenience. I took the first 100 fans I met at the game. There is nothing systematic about this method.
B. American adults, because that is what I say in the research question.
C. Inferential, because I am hoping to use data from my sample to reach conclusions about the larger population.
D. No. My sample came from a baseball game. What do people eat at baseball games? Hot dogs. So this is a sample of baseball fans, and baseball fans tend to eat hot dogs. My sample will probably eat more hot dogs than the general population.

Question 3

If someone says to you "I do not think the results of your study will generalize" what does that mean, and what might limit the generalizability of a study?

Answer 3

That means the results found in my sample will not apply to the larger population. Reasons for this can include lots of things: not statistically significant, non-random sampling methods, researcher biases in measurement, etc