Jan 23 - Three Claim, Four validities 2 Flashcards
Explain what construct validity is.
- How well all the conceptual variables are operationalized
- ## how well does it measure/manipulate the variable of interest
Explain what internal validity is.
to what extent can we conclude it is the indepedent variable that acts on the dependent variable, rather than some other third variable that is responsible for changes in the IV?
Explain what external validity is.
the extent to which the results of your study is genralizable to some other population, time, context.
Explain what statistical validity is.
The extent to which the study’s conclusions are reasonable, precise, accurate, and replicable.
- how well do the numbers support the claim?
- how strong is the effect (association = strength of correlation (r), experiment = size of difference between conditions (d))
- is the finding significant (p-value < 0.05)
- how precise is the estimate?
what are the construct validity threats to frequency claims?
- Not enough items (e.g. are you shy?) more valid to ask about behaviour in specific situations
- poorly written or understood measure
- Low reliability (different scores if you repeate the measure)
Explain the relationship between conceptual and operational definitions in construct validity.
In order for the claim to be constructionally valid we need to have an operational definition included and we need to conclude the operational variables are a good approximation of the conceptional variables.
Compare validity and reliability of a measured variable.
validity - are you measuring what your supposed to? (accuracy)
reliability - is your measure consistent (stability)
describe margin of error and what it functionally means
It represents how far from the point estimate we think the true population value belongs; has an inverse relationship (negative correlation) with sample size.
explain independent and dependnt variables, experimental and control groups.
- Independent variables are manipulated and dependent variables are measured.
- experimental group is the thing we are manipulating and the control is the normal group.
what are the external validity threats in frequency claims?
Can you generalize your findings beyond the sample?
what are the statistical validity threats in frequency claims?
- Point estimate - a value that acts as an estimate of the true population value; usually in the form of a percentage.
- Margin of error - represents how far from the point estimate we think the true population value belongs; has an inverse relationship (negative correlation) with sample size.
- Confidence interval - a range of values that is likely to include the true population value
What are the internal validity treats in frequency claims?
Frequency claims are not asserting causality so internal validity is not relevant.
what are the construct validity threats in association claims?
Must assess validity of BOTH variables
what are the external validity threats in association claims?
can you generalize the association beyond the sample, context, time
what are the statistical validity threats in association claims?
How strong is the association (p<.05) How precise is it?
what are the internal validity threats in association claims?
association claims are not asserting causality, so internal validity is not relevant
what are the construct validity threats in causal claims?
How well has the researcher measured or manipulated the variables
what are the external validity threats in causal claims?
- can we generalize the causal claim beyond the sample, to other settings, times,
- how representative is the sample and the manipulations and measures?
what are the statistical validity threats in causal claims?
What is the estimated effect size? how large is the difference between groups? how precise is the estimate?
what are the internal validity threats in causal claims?
- Was the study an experiment
- does the study avoid internal validity threats?
- does the study achieve temporal precedence?
- does the study control for alternative explanations?
