Terminology
Reasons for Nulls
Threats to Internal Validity
3 Rules of Causation
Name That Test
100
A variable whose value depends on that of another.
What is a dependent variable?
100
Dr. Phil is doing a study on financial intensive and test scores. He offers students 10 cents if they make an A on their PSY 303 final and compares the results with a class that wasn't offered any money for good scores. They were both taught by the same professor, but Dr. Phil did not find any significant results because of this.
What is weak manipulation? He didn't offer enough money to successfully manipulate participants.
100
Captain Crunch thinks his new crew is too rowdy. He takes sugar away from their diet, and in two weeks, they are not acting as rowdy. He thinks that sugar makes his crew more rowdy. What is a possible threat to his claim?
What is a maturation threat? Control group solves this.
100
This is the rule of causation which states asks if the researcher has considered alternative explanations for their results.
What is internal validity?
100
Different groups are placed in different levels of the IV. Afterwards, the DV is measured, and it's called this kind of test.
What is a between-subjects posttest study?
200
An experimenter's mistake in designing the independant variable
What is a design confound?
200
Judge Judy gives her defendants a test to see how smart they are. Out of her sample of 100 people, the average IQ was 20 with a SD of 2. She says her defendants are just stupid, but you say it's because of this effect.
What is floor effects? The questions were too difficult.
200
This effect is when you don't account for external events that could account for the change in the dependent variable.
What are history threats?
200
This rule asks if the changes in one variable accounts for the changes in another.
What is covariance?
200
One group goes through all levels of the IV. The DV is measured afterward, and it's this kind of test.
What is a within-subjects posttest design?
300
The effect when participants can choose groups or groups are separated by types of people (e.g. women, people who sign up for the study first, etc)
What is selection effect?
300
You want to see if your new spaghetti recipe is really spicy. You run a between-groups study, giving one group a control spaghetti and another group your Spicy Spaghetti. Then, you ask if it was spicy or not. No one in the control group says that it's spicy, but everyone in your Spicy Spaghetti group says, no, it's not spicy either. Your null result may have been reached because of this error.
What is a insensitive measure?
300
You are running a pretest/posttest design. The pretest has many extreme scores, but the posttest stores are much closer to the mean. You realize it's probably this threat to internal validity.
What is a regression threat? You add a control group to your study.
300
This rule of causation dictates that one causal variable must be measured before another.
What is temporal precedence?
300
Subjects are divided into two randomly assigned groups. The first group has their DV measured and then is subjected to IV1. Their DV is measured afterward. The second group has their DV measured, and then, they are subjected to IV2. Afterwards, their DV is measured. This is this design.
What is a between-subjects repeated measures test?
400
Not a confound, but when there is variability in experimental conditions. Also known as "noise."
What is unsystematic variability?
400
Obama funds a study to see how people think he did as president. His researchers distribute a questionnaire to a sample of 1000 people before and after his presidency. On a 5-point scale, Obama scored a good president average of 4.5 with a SD of .1 in his first year, and a mean of 4.6 in his eighth year with a SD of .2. A null result was reached because of this effect.
What is ceiling affect? The scale was unbalanced. 1. Horrific President 2. Awful President 3. Terrible President 4. Bad President 5. Amazing President
400
Dr. Dolittle wants to see if people run faster from dogs after 6 weeks of training. He runs a pretest on 1000 people, but only 120 people show up to the posttest 6 weeks later. Later, he realizes it's this threat to internal validity and decides to remove the scores of the people who don't show up.
What is an attrition threat?
400
Farmer MacDonald runs a study to see if horses run faster after drinking well water, bottled water, or tap water. The farmer makes them drink and then records how fast they run. He finds that horses run faster after tap water, so he thinks that tap water makes his horses faster. This is the rule of causation that Farmer MacDonald should consider.
What is internal validity?
400
Subjects are divided into two groups. This type of study goes like this: Group A -> DV -> IV1 -> DV -> IV2 -> DV Group B -> DV -> IV2 -> DV -> IV1 -> DV
What is a within-subjects repeated measures design?
500
A design in which each group is exposed to a different level of the independent variable. Also known as "independent-groups design."
What is between-groups design?
500
Statistically, the biggest reason for falsely confirming a null result.
What is lack of power?
500
Santa Claus recently ran a pretest/posttest between groups test. He wanted to see if naughty kids that got presents acted better next year than naughty kids who got no presents. However, he found that he judged the behavior of the kids who got no presents more leniently because he felt bad for them. He fell victim to this threat of internal validity.
What is an instrumentation threat?
500
Ronald McDonald wants to see if preferring his restaurant makes people happier. He finds that people who liked his restaurant were indeed happier with their lives on average according to a telephone survey administered to 5,000 people. He cannot establish causality because of this rule of causation.
What is temporal precedence?
500
You would use this type of test if you're measuring physiological effects.
What is a (pretest/postest) repeated measures design?
M
e
n
u