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;
65 Cards in this Set
- Front
- Back
Reliability |
How consistent are the scores due to score and not error |
|
Types of Reliability |
Test-retest, interrater, alterne form, split half, internal consistency |
|
Spearman-Brown |
For split half reliability, estimates of # of observers needed, more than 2 observers |
|
Internal consistency |
Calculate reliability of existing scale scores |
|
Internal consistency statistics |
Coefficient alpha and KR-20 |
|
Interrater reliability statistics |
Kappa, % agreement, Pearson R, |
|
Pearson correlation |
1 observer session total |
|
Validity |
Does it measure what it's supposed to measure |
|
Norm reference |
Compare individual with others |
|
Criterion reference |
Compare individual to a standard |
|
Construct |
An abstraction, believed to unify or produce response measured on an assessment device |
|
Content validity |
Degree to which elements of an assessment instrument are relevant to representative of target construct for particular assessment purpose |
|
Types of validity |
Face,content, construct validity, , convergent, discriminant, discriminative, concurrent and predictive |
|
Content related validity |
Face and content |
|
Construct |
Converget, disriminant, discriminative, factorial (aka construct) |
|
Criterion related types |
Concurrent and predictive |
|
Converget validity |
Same construct measured by 2 tests and scores related to it |
|
Discriminant validity |
Unrelated measures =no correlation |
|
Discriminative validity |
Comp compare 2 groups that are known to differ in scores in my construct |
|
Concurrent validity |
Do our scores predict score on gold standard test given at the same time |
|
Predictive validity |
Do my scores predict score on gold standard test given at a later time |
|
Exploratory factor analysis |
Explore data, data reduction, test construction, model to data |
|
Confirmatory factor analysis |
Test of construct validity, confirm items to scales, data fit model, test measurement model part of substantive research |
|
Sensitivity |
Detection rate, % of true positives on test identified correctly |
|
Specificity |
% of true negatives identified correctly |
|
T-test assumptions |
DV is continious IV categorical Independence of observation - no relationship between subjects Homogeneity of variance |
|
ANOVA assumptions |
Homogeneity Normality Independence Linearity |
|
Factor analysis assumptions |
Linearity Normality No multicolliniarity-don't want perfect correlation |
|
Linear regression |
Normality Linearity Independence No multicolliniarity Homoscadesticity- even along line |
|
Logistic regression assumptions |
No specification errors Big sample DV is categorical November multicolliniarity Nonparetric- all variables invited |
|
ANOVA repeated measures Assumptions |
Normality Homogeneity Sphericity-all data moving together |
|
Type 1 error |
Says it's significant when it's not Rejects null when null is true |
|
Type 2 error |
Fail to reject and null was false |
|
Moderator |
Affects relationship; modifies |
|
Mediator |
It is the relationahip. If not present then there will be no significance |
|
T-test types |
1 sample- 1 group vs norm Independent t test - 2 different groups Paired t test- pre and post |
|
Positive and negative correlation |
+ both go up - one goes up other down |
|
T-test write up |
(M=?, SD+ ?), t(df)=t, p=2-tail sig, d= effect size |
|
correlation matrix write up |
r= correlation value, p <.01 |
|
coefficient alpha |
fish=.81 (good internal consistency) |
|
factor analysis write up |
KMO, Barlett, rotation used, # of factors and their names |
|
chi square write up |
X2(df) =chi square value, p <.05 |
|
logistic regression write up |
X2(df) =chi square value, p <.05u |
|
ANOVA write up |
F(dfb, dfw)=F, p <.01 |
|
ANOVA repeated measures write up |
F(dfb, dfw)=F, p <.01, np2=eta2 (partial eta2) |
|
ANOVA mixed write up |
F(dfb, dfw)=F, 0 <.01, np2=eta2 |
|
linear regression |
F(dfb-regression, dfw-residual)=F, 0 <.01, R2 in percent of variance and betaweights |
|
nominal variable |
aka categorical variables, no # value like gender or ethnicity |
|
ordinal variables |
categorical with an order like education level |
|
part |
semipartial, unique |
|
coefficient alpha |
measurement creating, reliability consistency, items correlate |
|
chi square |
looks at association between two categorical variables, group differences |
|
correlation matrix |
how 2 variables are related, for continuous data, r=1 if each variable is perfectly correlated |
|
t-test |
compares the difference between 2 groups by comparing the means, 2 unrelated group on same continuous, dependent variable |
|
independent variable |
changes is controlled |
|
dependent variable |
being measured |
|
effect size |
difference is large enough to be practically meaningful, standardized difference in SD units between the mean of a control group and mean of treatment group, how different are the distributions .2 small, .5 moderate .8 large |
|
paired t-test |
aka dependent t-test, when comparing time 1 and time 2, withing/repeated measures, 2 sets of scores that are related |
|
ANOVA |
analysis of variance- difference, expanded t-test compares 3 or more groups |
|
linear regression |
prediction |
|
logistic regression |
IV- any type of variable, DV- dichotomous (predicts yes or no) |
|
null hypothesis |
hypothesis that there is no sig difference between specified population, any observed difference due to sampling or experimental error |
|
multiple regression |
has clear predictor and clear outcome, generally try to do causal forward not backwards |
|
orthagonal |
quartimax and varimax |
|
oblique |
promax and direct oblmin |