final studying for stats

a question being simple and specific refer to what type of validity

face

a question being sufficient refers to

content validity

a construct is an

idea

a variable is an

item

variables can be ___ or -____

qualitative or quantitative

types of quantitative variables

discrete, continuous

types of quantitative variables

categorical, bianary, nominal, ordinal

a systematic issue w ur experiment results in

bias

types of validity 

study validity and measurement validity 

measurement validity 

measure what you think youre measuring 

study validity

study what you think youre studying

validity has to do with if ur study is

relavent or appropriate

reliability has to do with if your study is

precise/consistant

types of measurement validity 

content, discriminant, convergent 

content validity 

what you measured captures the whole construct 

discriminant validity

what you measured isnt somethng else

convergent validity

measured in different ways but all tell the same story — applewatch = EEG

internal validity

did the scientists do it right

external validity

is it generizable

validity is NOT about

precision

validity IS about

relavance

study validity is

studying what you think youre studying

face validity

makes sense to a pedestrian

_________ are the items that make up the construct that allow the researcher to test the idea 

variables 

variables are the _____ that make up the construct that allow the researcher to test the idea 

items 

variables are the items that make up the _____ that allow the researcher to test the idea 

construct 

variables are the items that make up the construct that allow the researcher to test the ______ 

idea 

constructs make the idea

measurable

items are found

top of columns

variables are found

in the cells

longitudinal measures

change

longitudinal is done

multiple times

can longitudinal experimental be retro

no

cross sectional is done

once

can cross sectional measure change

no

do you interfear with observational

no

BOTH cross sectional and longitudinal measure 

if things are different and if theyre related 

noise

data issues that dont affect how well a dataset reflects what youre studying 

bias 

data issues that do affect how well the dataset reflects what youre studying 

systematic errors have a ___ and introduce ___

pattern bias 

good questions are

simple specific sufficieent

prospective

data collected as experiment unfolds

ordinal

natural order to categories - letter grades

nominal

no natural order to categories - eye color

bianary/dichotomous 

only 2 possible groups 

discrete

counted answers,  not measured whole numbers, ex. number of courses youre taking

continuous 

measured answers, very precise every little number ex 1.1, 1.01, 1.001 …