Chapter 3 Flashcards
3 types of claims
frequency
association
causal
frequency claims
describe a rate or degree of only ONE measured variable
-percentage, ratio, etc.
association claims
-one level of a variable is LIKELY to be associated with a level of another variable
-at least 2 measured variables
what type of study are scatterplots used for?
correlational
what is the slope of a zero association?
zero
which type of associations dont help with predictions?
zero associations
is “uncertain” language still okay? ex: “may” increase, etc…
yes, it can still indicate a causal relationship
3 criteria for causation
covariance
temporal precedence
internal validity
validity
appropriateness of a conclusion/decision
4 big validities
internal
external
construct
statistical
what do the 4 big validities do?
interrogate evidence and claims
construct validity
how well is the variable(s) operationalized/measured?
external validity
how well does the study represent the people outside of the study/how well would the results hold up if you included everyone/not just a sample?
statistical validity
extent to which the study’s stats are reasonable, precise and replicable
what is the point estimate in relation to the interval estimate?
point estimate is the middle value of the interval estimate.
what 3 validities are important for frequency claims?
construct, external and statistical
confidence interval
the range that the true population value falls in x% of the time
margin of error
range of values DEVIATING from the PI
what does statistical significance imply?
the result is probably not due to chance based on that sample
what 3 validities are important for association claims?
construct, external and statistical
covariance
A and B are related
temporal precedence
A comes before B in time
role of internal validity in causal claims
eliminate any alternative explanations
which validity should causal claims prioritize but which should they also refer to
internal & refer to all
which validities have a heavy trade off?
internal and external
what may stay the same in one study but could vary in another?
a constant
how many levels does a variable have?
at least 2
measured vs manipulated variables
measured: levels are observed/recorded
manipulated: controlled by researchers
in what type of study are variables manipulated?
causal/experiments.
2 reasons why some variables can only be measured and not manipulated
naturally occurring attributes (age, gender, etc.)
ethics
most of this type of variable can be both measured and manipulated
physiological
4 types of “roles” of variables
subject variables (self-esteem)
context variables (privacy)
stimulus variables (something presented that provokes a response)
response variables (test performance)
2 data types of variables
quantitative: levels differ in amount
qualitative: levels differ in quality/type
where do categorical, ordinal and nominal variables belong to?
categorical and nominal - qualitative
ordinal - quantitative
main difference between conceptual definition and operational definition of a variable
the operational definition defines the method that a variable is measured/manipulated; a conceptual definition is just an abstract/theoretical statement about a variable.
ex: hunger: number of hours of food deprivation (operational)
hunger: desire for food (conceptual)
3 reasons why operational definition are important
-replication
-forces researchers to clarify ideas
-objectivity and public verification
relation of variables in experimental studies vs non-experimental studies
experimental: IV is manipulated, DV is measured
non-experimental: predictor variable is assumed IV, criterion variable is assumed DV
difference in IV in experimental vs non-experimental studies
experimental: manipulated
non-experimental: measured