Correlation
Statistics
Experimental Design
Validity, reliability, and sampling
Miscellaneous
100
When do you compute a chi-square test for independence?
When you want to examine the relationship between two nominal or dichotomous variables
100
What analysis would you run in order to describe the age of your sample in your participants section? Would this be an inferential or descriptive analysis?
Frequency. It is descriptive
100
Researchers want to examine the effect of food on math test scores. Participants are randomly assigned to eat an apple, or nothing, and then take a mathematics test. What type of design is this? What statistical analysis should be computed?
This is a simple experiment. You should run an independent-samples t test.
100
What is the difference between internal and external validity?
Internal validity is the degree to which we can say that we found an accurate relationship among variables, in that the change in DV was caused by the change in the IV, while external validity is the degree to which we can say that the results of a study are accurate for different types of people in different settings assessed with different procedures.
100
What is a pilot study?
A preliminary study that uses a small number of participants to test out the measures and procedures of the study.
200
When do you compute a Pearson’s r?
When you want to examine the relationship between two interval or ration variables
200
If you have conducted an independent- samples t test and Leven’s test is not significant, which row of output do you use and report in your results section?
You would use the first/top row of the output
200
Turn the previous experiment into a multiple groups experiment. To recap the previous experiment, researchers wanted to examine the effect of food on math test scores. Participants were randomly assigned to eat an apple, or nothing, and then take a mathematics test. How would you turn this into a multiple groups experiment? Give an example of a multiple groups experiment based on the study described above.
Answers will vary. There should be at least three groups to the IV (perhaps eating an apple, eating chocolate, or eating nothing).
200
What is random selection (hint, related to sampling)?
It is a process of selecting a sample in which all members of a population or sub-population have an equal chance of being selected.
200
What is an operational definition? How might you operationally define “happy”?
An operational definition is the explicit explanation of a variable in terms of how it is measured or manipulated. Answers will vary. Might be something like “smiles at least 5 times a day”, “laughs at least twice a day”
300
You want to analyze the relationship between gender and GPA. What analysis will you compute if there are only two gender categories in the data?
You will use an independent-samples t test
300
If you ran a chi-square goodness of fit for pants type, comparing how often people wear khakis to how often they wear jeans, and obtained a p value of 0.61, what could you say about how often people wear khakis as opposed to jeans.
The p value is not statistically significant, therefore we would retain the null hypothesis that people wear khakis as often as they wear jeans.
300
Describe the procedure of a simple experiment that could be used to test whether wall color affected mood
Answers will vary. Participants could be assigned to sit in different colored rooms and report their mood on a survey, where they answer questions such as “I am feeling content right now” and “I am feeling angry right now”, measured on a 1-5 scale, 1 = strongly disagree, 5 = strongly agree.
300
What is snowball sampling?
A non-probability/convenience sampling strategy where participants recruit others into participating in the study as well.
300
A researcher wants to know if women wearing different clothes for a job interview impacted their probability of getting the job. The participants were shown images of the same woman wearing two outfits (a dress and a pants suit) and then were asked to fill out a questionnaire determining whether or not they would hire her. What is a null and alternative hypothesis?
Answers will vary Null hypothesis: participants will be just as likely to hire the woman when she wore the dress to the interview as when she wore the pants suits Alternative hypothesis: participants will be more likely to hire the woman when she wore the dress to the interview than when she wore the pants suit
400
When you square the correlational coefficient (r) what information do you have?
The proportion of variance in y accounted for by x
400
You want to see if there is a relationship between age and number of doctor’s visits per year. You run a Pearson’s r and obtain the results r = .15, p = .04. What does this data tell you about age and number of doctor’s visits per year?
The results are significant, and we reject the null hypothesis. There is a weak positive correlation between age and number of doctor’s visits per year. In other words, as age increases, number of doctor’s visits increase, although this relationship is weak.
400
A researcher wants to examine the relationship between gender and perceived trustworthiness. Participants are randomly assigned to read one of two scenarios: one scenario in which a woman is described as being a single mother, having two children ages 3 and 5, and working at the local gas station. The second scenario portrays a man as being a single dad, having a 13 year old daughter, and working as an investment banker. What are some potential problems with this study?
The scenes should only differ in the IV being manipulated: i. e. both subjects should have the same number of same aged children, and the same occupation. Otherwise, variability in perceived trustworthiness could potentially be due to one of these other factors.
400
Explain the “history” threat to internal validity, and how you help minimize the effects.
History is any event that occurs or environmental condition that is experienced between the pretest and the post-test might be the reason for the change in DV. The threat increases the longer the time lapse between pretest and post-test. A highly controlled laboratory setting can decrease this threat.
400
Type I or II error? You run a Pearson’s r and find that r = .6, p = .04. What type of error are you at risk of making if you interpret the results that are shown, and what is the probability that you are making that error?
A Type I error – the probability of rejecting a true null hypothesis. The probability of that is 4%.
500
You wanted to examine the relationship between hours a day spent on social media and depression, so you computed a Pearson’s r and got the results r = .29, p = .035. What do the results tell you about the relationship? Are the results significant?
There was a moderate positive correlation between hours per day spent on social media and depression. The results were significant, thus we would reject the null hypothesis.
500
You want to examine the relationship between ethnicity (Asian and Hispanic in this situation) and annual income, so you run an independent-samples t test and obtain the results t(20) = 2.22, p = .045, rpb2 = .12. What conclusions can you make from this information?
Reject the null hypothesis. There is a significant correlation between the IV and the DV, and there is a medium effect size
500
A researcher wants to examine the effect of living in apartments verses living in houses and life satisfaction. He recruits participants from each situation and has them fill out a survey that measures life satisfaction. What is wrong with this “experiment”?
This is a correlational study, not an experiment. The IV was not manipulated, so the variation in DV could be due to any number of factors that exist in the participants’ lives.
500
Name and define one type of threat to internal validity that is due to participant characteristics
Answers will vary. Could be any two of the following: Attrition – groups are differentially affected by attrition, meaning that the drop-out rate is higher in one group than the other Selection – How the participants were selected for the groups impacted the results Selection interactions – the way that participants were selected for the groups interacted with another threat resulting in a difference between groups
500
What is power and what are some strategies to increase it?
Power is the ability to reject the null hypothesis when it is, in fact, false. You can increase it by increasing sample size, increase homogeneity of the sample, and making sure your measures are sensitive.
M
e
n
u