Variables & Measurement
What do we call something that can take on different values or categories?
A variable.
“Twenty-one percent of Canadian adolescents reported vaping.” What type of claim is this?
A frequency claim.
Which validity asks whether the researchers actually measured or manipulated the construct they intended to study?
Construct validity.
What three criteria must be satisfied to support a causal claim?
Covariance, temporal precedence, and internal validity.
When high values of X tend to occur with high values of Y, what type of association is this?
A positive association.
A researcher asks 1,000 Canadians how many hours they sleep each night and reports the average. What type of claim is most likely being made?
A frequency claim.
A study reports that eating more family meals is associated with fewer eating-disorder symptoms. A journalist says family meals “prevent” eating disorders. What changed between the study and the headline?
An association claim was turned into a causal claim.
Question: A researcher records participants' height without changing it. Is height measured or manipulated?
Measured.
“Higher coffee consumption is associated with lower depression scores.” What type of claim is this?
An association claim.
Which validity asks whether the results can be generalized to other people, places, situations, or times?
External validity.
What does temporal precedence mean?
The presumed cause occurs before the presumed effect.
When high values of X tend to occur with low values of Y, what type of association is this?
A negative/inverse association.
Researchers find that people who exercise more report lower stress. Both variables were measured rather than manipulated. What type of claim is supported?
An association claim.
What is the difference between random sampling and random assignment?
Random sampling → external validity/generalizability.
Random assignment → internal validity/causal inference.
A researcher randomly assigns participants to receive either 10 mg, 20 mg, or 30 mg of a medication. What kind of variable is dosage?
A manipulated variable.
“Taking a deep breath reduces anxiety.” What type of claim is this?
a causal claim.
Which validity asks whether the numbers provide adequate evidence for the statistical claim?
Statistical validity.
In an experiment, the researcher controls or assigns participants to different levels of what variable?
The independent variable (IV).
What does a zero association mean?
There is no systematic relationship between the variables; knowing one does not improve prediction of the other.
Researchers randomly assign participants to exercise or control conditions and later measure stress. What type of claim could this design potentially support?
A causal claim.
A study has extremely high internal validity but uses a very narrow sample of participants in an artificial laboratory setting. What validity may be limited?
External validity.
“Life satisfaction” is the abstract concept a researcher wants to study, while a five-item questionnaire is what they actually use to measure it. What are these two levels called?
The conceptual variable/construct and its operationalization/operational variable.
A study finds that students who sleep more tend to have higher grades. The researcher does not manipulate sleep. What can the study establish?
An association, but not that sleep causes higher grades.
Which validity asks whether the presumed cause, rather than another variable, produced the outcome?
Internal validity.
Why does random assignment strengthen internal validity?
It makes groups more similar on other characteristics, reducing alternative explanations/confounds for differences in the outcome.
In an association study, what does “prediction” mean?
Using one variable's values to mathematically estimate another variable's values; it does not necessarily mean predicting what will happen in the future.
A survey estimates that 39% of adolescents vape, with a confidence interval of 37–41%. What does the interval tell you?
It indicates the precision/uncertainty of the population estimate; a narrower interval indicates greater precision.
Researchers report a point estimate but no confidence interval. What aspect of statistical validity is harder to evaluate?
Precision/uncertainty of the estimate.
Question: Everyone in a study is Canadian. Citizenship could vary in the real world, but it has only one value in this study. What is citizenship here?
A constant
A headline says, “Social Media Causes Depression,” but the underlying study simply measured social-media use and depression at the same time. What is the fundamental problem?
The study can establish covariance/association, but not temporal precedence or internal validity, so it cannot support the causal claim.
A study has an excellent measure, a large representative sample, and precise statistics, but researchers cannot rule out alternative explanations for why X and Y are related. Which validity is the major problem?
Internal validity.
Researchers randomly assign children to three self-distancing conditions, then measure how long they persist on a boring task. Identify the IV, DV, and why the design can support a causal claim.
A strong correlation is found between X and Y. Name two reasons this still does not establish that X causes Y.
There may be reverse causality/lack of temporal precedence and/or a third variable/alternative explanation.
A study claims that a treatment causes depression to decrease. Researchers manipulated treatment exposure and randomly assigned participants, but the depression measure poorly represents depression. Which validity threatens the conclusion, and why?
Construct validity. The study may have good causal design, but if the outcome was poorly measured, researchers cannot confidently conclude that the treatment changed depression itself.
A researcher measures age and finds that older participants have higher scores on a memory test. Why can't the researcher simply conclude that age causes the difference?
Age was measured, not manipulated, so the study cannot establish the necessary causal criteria, particularly temporal precedence/internal validity through experimental control.