RM Final Flashcards
(92 cards)
The possibility that a third variable is the reason for the relationship between x and y is a problem with what type of validity?
Internal Validity
What is a bi-variate correlation? What type of claim is usually made from them?
An association claim, between two variables
What is the relationship between moderators and external validity?
When generalizability is affected by moderators, your association may not generalize to all groups of people
Lisa finds a strong positive correlation between impulsivity and rash decision-making.
r = 0.8, p < 0.05.
It turns out there were several pathological gamblers in her sample. Pathological gamblers score unusually high on impulsivity measures. When she takes these people out, what will happen to the correlation? Increase or decrease?
Decreases; removing the outliers will weaken the correlation
What are the three rules for causal claims?
- Temporal Precedence (IV comes before DV)
- Covariance (As A changes, B changes)
- Internal Validity (there is no third variable that describes the relationship and cannot tell from associations)
What type of research design involves measuring the same variables, from the same people, across different time points? There is no manipulation involved.
Longitudinal Design
This diagram illustrates what kind of correlation?
GPA in 3rd grade__r=.45__GPA in 9th grade
Autocorrelation - the association of the variable with itself across time
When a relationship between two variables depends on the level of a third variable, that kind of third variable is called___?
Moderator - effect is stronger in one subgroup vs. another
This diagram illustrates what kind of correlation?
TV use in 7th grad__r=.35__Substance use in 12th grade
Cross-Lag Correlation - the association of an earlier measure of one variable with the later measure of another
Dr. Bazinga conducts a study examining whether there is a link between exam scores and sleep. But, he wants to see if any other variables (time studying, age) can explain the relationship. How might you interpret the result of sleep?
As sleep increases, exam scores increase, even after controlling for time studying and age
This method assigns people to groups in experiments. It ensures that features of people are equated except on the independent variable
Random Assignment
_________ is a simple study that uses separate group of participants that generally completed before conducting the study of primary interest in order to confirm the effectiveness of manipulation
A pilot study
Eric designs a study to examine drink preferences of university freshmen. In the study, all participants drink a cup of coffee then rate their enjoyment of the coffee, then drink a cup of tea and rate their enjoyment of the tea, and finally drink a cup of milk and rate their enjoyment of the milk. Eric tells Theresa about the plans for his study and she says she is concerned that he could have a problem of ___ effect in his study. How can Eric fix this problem?
Order effect or Counterbalancing (presents levels of the IV to participants in different orders)
In a business class experiment on the endowment effect, Theo is comparing the value of a coffee mug to someone who owns it and is selling it to someone who is buying it. The endowment effect is the tendency of sellers to value something they own more than buyers do. Participants are randomly assigned to be buyers or sellers of a mug with their first name on it. Buyers select the maximum price they would pay for the mug. Sellers select the minimum price they would accept for the mug. Theo then compares the buyers’ and sellers’ prices. What kind of design is this?
Posttest Only
Expert dancers in foxtrot and tango were recruited from dance studios and participated in Nick’s study. Their task was to listen to classical music while their electrical brain activity (EEG) was measured. Nick then compared foxtrot dancers’ EEG alpha activity to that of tango dancers’. He found that tango dancers had more alpha activity in the brain than foxtrot dancers. Is Nick’s method correlational or experimental? If experimental, what design?
Correlational (no random assignment, no manipulation of the IV)
Which of the four validities should you focus most on, when interrogating experiments?
Internal Validity
Threats to IV: Not as extreme over time
Regression
Threats to IV: Participants number decreases across testing
Attrition
Threats to IV: Order effect
Testing
Threats to IV: Measuring device effectiveness declines
Instrumentation
Threats to IV: Expectations impact interpretations
Observer Bias
Threats to IV: When participants figure out the aim of the experiment
Demand Characterisitics
The expectation of the treatment impacts the results
Placebo Effect
Dr. Lucas is examining the effect of a talking treatment on hyperactivity in adolescent boys in a local high school. He randomly samples students from this high school and measures their hyperactivity using self-report tests. Dr. Lucas then engages this group in his talking intervention for 2 hours a week every week for 6 months and measures hyperactivity again. He finds that his group is less hyperactive than 6 months before. Is there a threat to internal validity present? If so, which threat?
Maturation (due to change emerging spontaneously over time)