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;
67 Cards in this Set
- Front
- Back
Indepentent Variable (IV)
|
affect or alter status on another V
|
|
Dependent Variable (DV)
|
depends on status from another V
|
|
Ptotocol Analysis
|
Content analysis of cognitive processes
|
|
Interval recording
|
observing a beh'r for a peroid of time
|
|
Event sampling
|
observing a beh'r each time occurs
|
|
Simple random sampling
|
Ea member of pop has equal chance of being selected
|
|
Stratefied random sampling
|
dividing pop into strata and randomly selecting subs
|
|
cluster sampling
|
selecting units of ind rather than inds and either including all inds or randomly selecting from units
|
|
extraneous Variable
|
(confounding)
source of systematic error. irrelevant V |
|
Techs to control effects of extraneous Variable
|
random assignment
holding ext v constant matching subs on ext v building ext into study (blocking) statistical control- Ancova |
|
internal validity
|
determine if a causal relat exists
|
|
maturation
|
extraneous V
biological or psychological change that occurs to sub during course of study as a function of time internal |
|
history
|
external event which affects subs
external |
|
testing
|
when exposure to test alters subs performance
|
|
instrumentation
|
accuracy of measures or procedures
|
|
interaction betw selection and treatment
|
subs characteristics make them respond in a certain way that can't be generalized to pop
|
|
reactivity
|
respond in a certain way when know observed
|
|
demand characteristics
|
Cues that inform subs of study purpose or what is expected
|
|
mutliple treatment interference
|
exposing subs to 2 or more levels on IV
|
|
counterbalance design
|
control treatment interference- sub receive levels in randon order
|
|
between groups designs
|
effects of dif levels of IV assessed by admin ea level to a dif grp of subs and then comparing the status
|
|
factorial design
|
2 or more IV
through info relat among V's analysis the main effect of IV and interaction between them |
|
main effect
|
effecton 1 IV onthe DV, disregardign effects of all other IV
|
|
interaction
|
effects of an IV differ at dif levels of another IV
|
|
within subs design
|
all levels of the IV are administered sequentailly to all subs
|
|
autocorrelation
|
subs performance on post-tests is likely correlated to performance on pretests.
can inflate the value therey increasing chance of Type I error |
|
mixed design
|
combo of between grp and within sub methologies
|
|
AB design
|
simple single sub
|
|
ABA, ABAB reversal designs
|
more than baseline and or more than one treatment
|
|
multiple baseline design
|
if reversal unethical
not require withdrawal of treatment |
|
nominal scale
|
unordered cats
eg) sex |
|
ordinal scale
|
divides into cats and gives info on order
eg ranks and Likert |
|
interval scale
|
order, equal intervals
eg) IQ tests |
|
ratio scale
|
order, equal intervals and absolute zero pt.
eg) Kelvin scale, no correct answers, reaction in seconds |
|
postively skewed
|
mo,md,m
|
|
negativley skewed
|
m, md, mo
|
|
range
|
subtract lowest score from highest
|
|
variance (meam square)
|
all of the scores
ss/n-1 |
|
standard deviation
|
measure of variablity
sq root ss/n-1 |
|
central limit theorem
|
as sample size increases, samplling distribution of the mean approaches normal distribution
the mean of the sampling dist = pop mean sd of the mean= pop SD divided by the sq rt of the sample size |
|
standard error of the mean
|
sd of a sampling dist of the mean
|
|
alpha
|
size of rejection region
.05 or 5% = rejection; 95% is retention rejection region always in tail |
|
type I error
|
reject true null
equals alpha |
|
type II error
|
retains a false null
equals beta |
|
parametric tests
|
pop mean, variance, other parameters
interval or ratio scale. when a study has more than 1 grp has homoscedasticity and value of interest in normally dist. in pop. |
|
nonparametric tests
|
analyze data collected on variables on a nominal or ordinal scale.
|
|
chi square
|
analyze frequency of observations of a nominal variable
|
|
mann-whitmey u test
|
Nonparametric
ordinal data study includes 2 independent grps and data on DV is reported in ranks |
|
wilcoxon matched-pairs signed ranks test
|
two correlated (matched) grps and the dif between the DV scores of subs who have been matched in pairs are converted to ranks
|
|
kruskal-wallis test
|
sim to mann-whitney
study included 2 or more indepentdent grps and the data to be analyzewd are ranks |
|
student t tes
|
eval dif between 2 means
|
|
analysis of variance
ANOVA |
compare 2 or more means
|
|
factorial ANOVA
|
study has 2 or more IV
|
|
randomized block factorial ANOVA
|
blocking has been used to control ext V
|
|
repeated measures ANOVA
|
within sub design in wch dif levels of the IV or combo of the levels of 2 or > IV are sequentially given to ea sub
|
|
analysis of covariance (ANCOVA)
|
combo of anova and regression analysis.
control ext v by statistically removing the portion of variability in the DV caused by the ext V |
|
mixed (split-polt) anova
|
mixed design.
at least one IV is a between-grp V and IV is a within-sub V |
|
trend analysis
|
1 or > quantitative IV and eval form of relat between IV and DV
|
|
mutivariate analysis of variance (MANOVA)
|
1 or > IV and 2 or > DV that are ea measured on an interval or ratio scale.
|
|
correlation coefficient
|
sums the degree of association bet V with a single number Pearson r
|
|
mutlivaraite techs
|
assess degree of association among 3 or > V and make predications that involve at least 2 predicators and 1 criterion
|
|
mutliple regression
|
2 or > continuous or discrete preds will be used to predict status on a continous criterion
|
|
canonical correlation
|
extensionof multiple regression
|
|
discriminant function analysis
|
2 or > continuous preds will be used to predict or est a person's status on a single nominal criterion
|
|
logistic regression
|
predict status on a single discrete criterion using 2 or > cont or discrete preds.
|
|
path analysis
|
extension of multiple regression
translating a theory about the causal relationship among a set of variables into a path diagram |
|
LISREL
|
path analysis
used when a causal model includes recursive (1 way) and non-recursive (2 way)paths. exams the relat between observed V and takes into account the latent traits those V are believed to measure and the effects of measurement error. |