lecture 4 notes exam 1 Flashcards
What are the four big validities?
construct, external, statistical, and internal
construct validity
are the variables measuring what you think they are measuring?
external validity
do the results generalize to other people’s times or situations (a lot of the times its the general population)?
statistical validity
how well the numbers support a claim (confidence interval and point estimate)
internal validity
When a causal claim is made, have alternative explanations been ruled out? have we eliminated confounds?
steps to interrogate a frequency claim
- define the variable, e.g., degree of food insecurity
- Construct validity: ex: how did they measure food insecurity? Was it operationalized ina reasonable way?
- External validity: e.g., who did they survey? do the households they samples generalize to the full US population?
Statistical validity: What is the point estimate and the precision of that point estimate?
steps to interrogating causal claims (3 criteria)
- covariance (make sure A and B covary; it can be positive or negative).
- temporal precedence: the studys methods ensure that A comes before B
- internal validity: the studys methods ensure that there are no alternative explanations for the change in B
Do experiments support causal claims?
yes because one variable is manipulated and the other is measured
independent variable
the manipulated variable (the cause)
dependent variable
the measured variable (the effect)
random assignment
a method of assigning participants to levels of the independent variable such that each group is as similar as possible
random sampling
will increase external validity happens when you are choosing people to participate in the experiment
control variables
things that could vary but are kept constant for the purpose of the experiment (locations, activities, people)
“venting causes increased aggression slide.”
validities to interrogate causal claims and ranking their importance
internal (most)
construct
statistical
external (least)
Do all studies satisfy all validities?
no its impossible to find one that satisfies all validities at once
When is external validity essential?
frequency claims
ANOTHER SLIDE
Tuskegee Syphilis study explain it and say the 3 main violations
3 ethics were violated. Men were told they were going to receive medical treatment. No one was treated for their syphilis despite 2/3 of them being positive. They didn’t treat them with antibiotics to see what would happen if syphilis was left untreated. Many had major damage to their internal organs, and some even died.
1. they were not treated respectfully (no consent was given)
2. The patients were harmed: withholding treatment (because of this they got sicker, went through painful testing, and infected others)
3. only done on poor black men (anyone could get syphilis; they were targeted)
milgram obedience studies say the 3 main violations
Participants were told they were participating in research to help memory. When they got the answer incorrect, they got shocked, and each time they got it wrong, they upped the voltage. The authority figure (man in a white coat) urged the participants to keep shocking and potentially kill the person. They didn’t know that the person wasn’t actually being shocked and was an actor. (similar to people obeying hitler)
1. lied to the participants
2. participants had lasting effects that haunted them
Core ethical principles of the Belmont Report
- Principle of respect for persons (everyone deserves informed consent; children and people with disabilities need parental or other consent )
- Principle of beneficence (protect participants from harm)
- Principle of justice (the sample of participants should reflect the population….
Ethics violations in squid game
You need to give informed consent, and even if there is consent, you should be allowed to leave at any time and should not be coerced with money or large incentives. no physical or psychological damage; dont recruit someone in a vulnerable state (all of them were in debt and in a vulnerable state), everyone should benefit from the research even the participants in the study
ethical standards for research
NEED MORE INFO
