Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
35 Cards in this Set
- Front
- Back
|
how well your operationlization REPRESENTS the constructs
|
TRANSLATION VALIDITY
|
|
|
HOW well a meaure APPEARS to measure what you want
|
face validity
|
|
|
how well a emasure covers the material included in a construct and avois material not cinlude idn a construct
|
content validity
|
|
|
validity of a measure based on its relationship to another IV measure as PREDICTED BY YOUR THEORY how hwo emasure hsould behave
|
criterion-related validity
|
|
|
how wella emasure predicts what you'd expect it to predict
|
predictive validity
|
|
|
how wella measure can distingiosh b/w ggroups that it should theoritically be able to distinguish
|
concurrent validity
ex: measure of intelligence should predict GPA as well |
|
|
how well a measure matches up with another measure of the same construct
|
convergent validity
ex: one measure of intelligevent should correlate highly with other accepted measures of intelligence |
|
|
how DIFFERENT your measure is from other measures that it should theoricatlly be different from them
|
discriminate validity
|
|
|
when you use only ONE VERSIONS OF A TREATMENT or program in your study
ex: you say studying affects grades, but you only look at one type of studying |
mono-operation bias
|
|
|
when you give some group a treatment, but they may also be gettingsome OTHER TREATMENTS THAT MAY AFFECT YOUR OUTCOME
|
interaction of different treatments
|
|
|
only ONE MEASURE fo a construct
ex: you're intersted in the effect of age on depression, but you only measure depression with one test |
mono-method biasq
|
|
|
this is when testing or measuring makes peoplemore/less sensitive to your treatment
|
interaction of testing and treatment
|
|
|
treatment has unintended consequences
|
restricted generalizbility accros construct
|
|
|
when you only look at one level of your construct but label it as if it was for ALL LEVELS
ex: saying studying doesn't improve grades, but you only measured grade improvement after 20 sec. of studying |
confounding constructs and levels of constructs
|
|
|
measurement error that is randomly distributed ACROSS measurements
- one way to reduce this is to use multiple ways of measurement |
random error
|
|
|
error in measurement that occurs in the same direction in most cases
- one way to reduce this is by being careful in the design of your study |
systematic error
|
|
|
what is the difference between an experiment and a quasi-experiment?
|
the participants are not randomly assigned to condition in a quasi-experiment
|
|
|
this type of varialbe averages out betwee/ groups with randoml selection.
ex: age, social status |
extranous variable
|
|
|
it is a variable that does not avergae out and causes the results of the experiment to be questionable
|
confounding variable
|
|
|
the double blind procedure is used to control _____________ bias
|
experimental
|
|
|
an experimental design where the sample is grouped into relatively homogenous subgroups and then members of each subgroup are randomly assigned to conditions.
|
randomized block design
|
|
|
t or f: the randomlized block design helps there be no other plausable altnernative explanation
|
TRUE
|
|
|
is a statistical method for reducing variance caused by extraneous variables, by factoring out pre-test measurements of the extraneous variable.
|
COVARIANCE DESIGN
|
|
|
Analysis of covariance is...
|
ANCOVA
|
|
|
A t-test has one IV with how many levels?
|
only 2
|
|
|
a single factor anova has one IV with how many levels?
|
only 3+
|
|
|
T OR F: you use an ANCOVA when you have pre-experimental data you want to account for as well
|
TRUE
|
|
|
________ Validity- the degree to which conclusions you reach about relationships in data are reasonable.
|
conclusion
|
|
|
what is a type I error?
-possibility of a this increases as your alpha level increases -instead bring the rejection region closer to the mean |
where you think there is a relationship, when it actually DOES NOT EXISTS
|
|
|
What is a type II error?
|
when you conlude there is no relationship, but it actualy
EXISTS |
|
|
t/f: -generally .80 or higher is acceptable power level for a study
|
true
|
|
|
t/f: -low power makes us conclude that no relationship exists, when it really does (Type II Error)
|
true
|
|
|
t/f: a correlation coefficient is a numerical index of the degree of relationship b/w 2 variable
|
true
|
|
|
________ reasoning: uses specific observsations to inder wider, more geneliable theories or conecept
bottom- up appraoch |
inductive
|
|
|
________reasoning: uses a broad theory to predict specific occurences
|
deductive
|
