Correlation Research
The best graph format to examine an association claim between two quantitative variables.
What is a scatterplot?
Studies the effects of at least two independent variables
What is a Factorial Design?
Actively lying to research participants
What is deception by commission?
Pre-existing group differences
What are selection effects?
Can be just as interesting as experiments that show group differences
What are null effects?
r = .50
Arithmetic means for each level of the independent variable averaging over levels of the other independent variable
What are marginal means?
A document that outlines the purpose, procedures, confidentiality, risks and benefits of the research
What is informed consent?
The only way or at least the best way to investigate a causal claim.
What is an experimental design?
A third factor that could additionally inflate variability within groups.
What is situation noise?
Implications for when a small effect might still be important.
What are life-or-death implications?
When the effect of one IV depends on the level of another IV.
What is an interaction?
A scientific approach to answering questions often involving human participants to test health care treatments or drugs.
What is research?
All potential 3rd variables that researchers keep constant in experiments
What is a control variable?
The likelihood that a study will yield a statistically significant result in our sample when the IV has an effect in real life.
What is power?
These are hard to detect with a straight line model.
What are curvilinear relationships?
Used to evaluate the interaction effect in the middle part of the summary table.
What are difference in differences of means?
In 1976 in Eldridge, Maryland, scientists met to discuss basic ethical principles regarding research with human participants which lead to this document.
What is the Belmont report?
Eliminate alternative explanations (confounds) by keeping all variables (except the independent and the dependent variable) constant.
What is a well-designed experiment?
Measurement error, individual differences, and situation noise
What are the three sources of high within-group variability?
1.What is the effect size? How strong is the association?
2.Is the correlation or the mean differences statistically significant?
3.Are there any outliers affecting the findings?
4.Is there a range restriction?
5.Is the association linear or curvilinear?
What are the 5 questions to ask when investigating aspects of the statistical validity of an association?
Needed to ensure confidence that there is an interaction.
What is a statistical test?
Children, people with intellectual or developmental disabilities, prisoners.
Who are vulnerable populations with reduced autonomy?
Occur when exposure to one level of the independent variable influences responses to the next level.
What is an order effect?
Weak manipulations, insensitive measures, ceiling and floor effects, design confounds
Why we might not have enough between-groups difference?