Reproducibility
This term refers to the fact that only significant results get published.
What is the file drawer problem?
Theories are tested through sets of questions, methods, assumptions referred to as what?
What are Experimental Paradigms?
This type of t-test checks the probability that the result is in either tail of the distribution.
What is a two-sided t-test?
Collecting data from people repeatedly over time suffers from this major problem.
What is attrition?
Children, Crowd workers, Incarcerated people fall into this IRB category.
What is a Vulnerable population?
This term describes checking results by using the same methods with new data.
What is 'replicable'/'replicability'?
This term refers to how we measure a psychological construct.
What is operationalization?
This type of t-test allows for the samples to have unequal variance.
What is Welch's t-test?
This type of design exposes all participants to both the experimental and control conditions.
What is within subject design?
The right to ask questions, stop at any time, and decide for themselves, refers to this aspect of the consent process.
What is Respecting Autonomy?
This describes the process of time-stamping details about an experiment in advance.
What is 'pre-registration'?
WEIRD stands for this in research.
What is Western, Educated, Industrial, Rich, Democratic?
Many statistical tests assume data has this type of distribution.
What is a normal distribution?
This is something that consistently differs between datapoints besides the experimental manipulation and affects internal validity.
What is a confound?
**Bonus** Name one type of confound
This type of pressure can encourage participants to continue in a study even if they don't want to.
What is coercion?
The central limit theorem describes this distribution.
What is a normal distribution?
A correlation is this type of effect.
What is observational (not causal)?
This type of linear model allows more than 2 types of a categorical variable.
What is an ANOVA?
Increasing amounts of treatment resulting in increasing amounts of improvement demonstrate this relationship.
What is a dose-response relationship?
This collection approach, if used, must be disclosed to the participant after the experiment.
What is deception?
FAIR stands for this in data science.
What is Findable, Accessible, Interoperable, Reusable?
Anscombe's quartet demonstrates the importance of doing this with your data.
What is visualizing it?
A correlation of these values mean you can perfectly predict one variable from another.
What is -1 and +1?
Participants can get better (or worse) over time. In experimental design terms this is called...
What are carry-over or practice effects?
Identical rats learned differently if students were told their rat was smart or dumb, and demonstrated the importance of this aspect of collection design.
What is experimental masking/blinding?