To Err is Human
A Harvard psychologist wants to test Harvard and Yale students to see how fanatically attached they are to their schools' sports teams. She recruits 30 Harvard students who are in the audience at a football game. To recruit Yale students, she contacts a friend who is teaching a philosophy class there, and his students sign up for her research.
At the end, she finds that Harvard students support their sports teams more fanatically than Yale students. What could be a potential problem here?
Non-random sampling, biased/unrepresentative sample
What is the difference between positive correlation and negative correlation?
If two variables are positively correlated, higher scores on one variable co-occur with higher scores on the other variable.
If two variables are negatively correlated, higher scores on one variable are associated with lower scores on the other variable.
A researcher wants to find out if a certain drug increases people's visuospatial abilities. To this end, they put the participants in a maze and record the time it takes for them to get out. Then the participants take a drug, and an hour later, they are brought to the same maze again. The researchers find that it takes the participants less time to complete the maze after taking the drug. When they submit their paper to a journal for submission, a reviewer points out that people may have improved in the navigation task not because of the drug, but just because it was the second time they were doing it and they were more familiar with the maze. How can the researcher amend their experiment to respond to this criticism?
The researcher can switch from a within-subjects design to a between-subjects design.
(Bonus: They can also use a different maze of course.)
Descriptive statistics describes general tendencies in the data ______________ statistics tells us how confident we can be in generalizing the data to the wider population.
Inferential
What is informed consent?
Before testing human subjects, a researcher has to obtain their consent. In order for the consent to be valid, the prospective subjects need to be informed about the experimental procedures and its potential risks and benefits. If they consent to participating in the research after being informed about and understanding the aforementioned points, they are giving informed consent.
The CEO of a company that employs 200 people boasts that their employees' average annual salary is $100k.
This information is accurate, but upon closer inspection, it turns out that the 10 top executives of the company each make $1M, while each of their other 190 employees make ~53k on average.
What would be a less misleading way to summarize the salaries at this company?
Using median instead of mean
What do we call the variable that the researchers manipulate in an experiment (e.g., varying the temperature of the room across conditions)?
Independent variable
How would you eliminate a potential confounding factor?
Adding a new condition and/or counterbalancing the conditions/trials
What is the name of the statistical procedure for combining the results of different studies that examine the same topic?
Meta-analysis
What is an operational definition?
The definition of a variable in terms of the specific procedures used to produce or measure it.
What is the main disadvantage of a correlational study?
We cannot make causal inferences.
A patient who has suffered a very rare brain damage and started seeing the world in black and white is referred to a neurologist for examination. The neurologist later publishes a paper on this particular patient and their symptoms. This study is an example of _____.
Case study
Active control
If the analysis shows that we are extremely unlikely to have obtained a result by chance alone, the results are said to be ___________ ___________.
statistically significant
What is the bidirectionality problem?
If a correlation is observed between variables X and Y, we may not be able to say that X causes Y or Y causes X, because either causal pattern might be possible. This is known as the bidirectionality problem.
Alice is filling a personality questionnaire given to her by a researcher. One of the questions asks her if she can stay calm under pressure. Alice is usually a very nervous person but does not like to admit it, as she fears that it will make people think she is a mentally unstable person. So she answers that she can stay calm. Therefore the researcher reaches inaccurate conclusions about her personality. Which bias is Alice's response an example of?
Social desirability bias
What is it called when the effects of Variable X on Variable Y depend on a third Variable Z? (Hint: A researcher finds that caffeine causes palpitations, but only in people who are used to a low caffeine intake. So the baseline/habitual caffeine consumption modulates the relationship between caffeine and palpitations.)
Interaction
What is a "double-blind procedure" useful for?
It can help eliminate errors that spring from the experimenter's/coder's own biases.
What is the difference between the highest and lowest scores in a distribution called?
Range
What is the law of parsimony?
Also known as Occam's razor, this law says that when two (or more) theories can explain a phenomenon equally well, choose the simpler one.
John reads in the newspaper about a recent social psychology experiment that showed that women were more likely to ask for directions when they were disoriented compared to men. John thinks that this is because society expects men to be independent and self-sufficient, which makes them more likely to avoid asking for help.
These findings are later challenged by another study that shows the opposite results. Men are more likely to ask for directions than women! John now thinks it must be because women feel more vulnerable than men and are afraid of talking to strangers.
What is the bias that explains John's causal attributions to the findings that he reads?
Hindsight bias
What is the difference between a hypothesis and a theory?
A hypothesis is a predicted relationship between two variables, which is usually in the form of an if-then statement.
A hypothesis can be supported or rejected by data.
If a certain family of hypotheses receive repeated confirmation, they can be organized into a theory that summarizes known facts and makes further testable predictions.
What are so-called "unobtrusive measures" useful for?
They help reduce observer effect by keeping the participants unaware that their behaviors are being recorded/measured.
__________ _____________ describes how much the scores differ from the sample mean.
Standard deviation
What do internal validity and external validity mean?
Internal validity is about the reliability of the results (the causal connection that the results point to), external validity is about the generalizability of the results.