Non-Experimental Designs
Because researchers only observe and measure things as they already are, non-experiments can't establish this.
What is causality?
In a poll of South Orange voters, all registered voters in town make up this, while the 500 people actually called are the sample.
What is the population?
The more hours students spend studying, the higher their exam scores tend to be. This is the direction of that correlation.
What is positive?
Researchers flip a coin to decide which participants in the study get the caffeine condition and which get the decaf condition. This is the term for that process.
What is random assignment?
In our class experiment, one group did the attention task with a TikTok video playing, while the other did the same task with their phone put away. The phone-away group is called this.
What is the control group/control condition?
Rate my Professor mostly hears from students who either loved or hated the class, which is an example of this bias.
What is non-response bias?
A researcher surveys the first 50 students she runs into at Dunkin because it's quick and cheap. This is the sampling method.
What is convenience sampling?
A study finds r=-.85 between hours spent scrolling social media and hours of sleep. Describe the direction and strength of that relationship.
What is a strong negative correlation?
This type of design involves each participant taking a memory test once after drinking coffee and once after drinking decaf.
What is a within-subjects design?
A researcher gives every group the same instructions, the same room, and the same time limit, so any difference in test scores can be attributed to the type of study music. This is the type of validity that makes that possible.
What is internal validity?
"Do you support lower tuition and free parking" breaks this item-writing rule because it asks about two things at once.
What is a double-barreled question?
A researcher randomly picks three dorms on campus and surveys every student living in them. This is the sampling method.
What is cluster sampling?
In a class survey, one student reports studying 10 hours a day, far from everyone else's score, which can distort the correlation. This is the term for that kind of extreme score.
What is an outlier?
A researcher gives a memory test but nearly everyone scores a 100, so the test cant show any differences between conditions. Because the test was too easy, this is the problem.
What is a ceiling effect?
A researcher gives one group of students a caffeine pill and another group a sugar pill. Deliberately changing the independent variable like this is the key feature of an experiment called this.
What is manipulation?
A researcher secretly watches how students behave in the dining hall, because people who know they're being watched may act differently. This is a term for that change in behavior.
What is reactivity?
To make sure freshmen, sophomores, juniors, and seniors are all included, a researcher divides students by class year and randomly picks students from each group. This is the sampling method.
What is stratified random sampling?
A little stress helps performance, but too much hurts it, making an upside down U on a scatterplot. Pearson's r can be misleading because the relationship has this shape.
What is curvilinear?
This is when a researcher does a trial run of the study on a small number of participants before running the full study with all of the participants.
What is pilot testing?
In a staged experiment, like Milgram's, a person who pretends to be a participant but is secretly working with the researcher is called this
What is a confederate?
Since 1900, the taller candidate has won 19 of 26 U.S. presidential elections, a pattern found by examining existing public records. This is the type of non-experimental design.
What is an archival study?
To find out how often college students exercise, a researcher surveys people only at the campus gym. Students who never go to the gym has no chance of being included, which is this type of bias.
What is selection bias?
A study finds that students who watch more TV tend to have lower GPAs, with r=-.58 and p=.04. Because p is below .05, this is how to describe the result.
What is a significant negative correlation?
This technique varies the order in which participants experience the conditions, so that practice, fatigue, and carryover effects don't favor one condition.
What is counterbalancing?
A participant figures out that the researcher expects caffeine to improve memory and tries harder on the test because of it. Features of a study that let participants guess the hypothesis are called this.
What are demand characteristics?