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;
52 Cards in this Set
- Front
- Back
IV v. DV
|
IV: form the basis of the groups to be compared, can be manipulated or non-manipulated
DV: outcome measure |
|
true experimental v. quasi-experimental v. observational
|
true experimental: at least one IV is manipulated and subjects are randomly assigned
quasi-experimental: at least one IV is manipulated but there is non-random assignment observational: no intervention or manipulation |
|
group v. single subject design
|
group (nomothetic): can be b/w groups when comparing independent groups, or with/in subjects when comparing correlated or related groups (e.g., matched subjects or repeated measurement)
single subject (idiographic): one or very few subjects are studied and are measured many times during baseline and treatment (e.g., AB, ABAB, multiple baseline) |
|
AB v. ABAB v. multiple baseline v. simultaneous treatment v. changing criterion
|
AB: baseline condition followed by tx condition
ABAB: baseline, tx, baseline, tx multiple baseline: tx is applied sequentially or consecutively across subjects, situations, or bx's simultaneous treatment: two or more tx's alternated during tx phase changing criterion: attempt to change bx in increments to match a changing criterion |
|
simple random sampling v. stratified random sampling v. proportional sampling v. cluster sampling v. systematic sampling
|
simple random sampling: every member of the population has equal chance
stratified random sampling: population is divided into strata and then a random sample of equal size selected from each stratum proportional sampling: individuals are randomly selected in proportion to their representation in the general population cluster sampling: random selection of naturally occuring clusters systematic sampling: selecting every kth element after a random start |
|
interval recording v. event sampling
|
interval recording: also called time sampling, a type of bx measurement that involves breaking the period of interest into samller periods of time and observing whether the bx occurs the entire interval (whole-interval) or at all (momentary)
event sampling: tallying the number of times the target bx occured |
|
threats to internal validity
|
hx: specific incidents that intervene b/w measuring points
maturation: time passing testing: familiarity w/ testing instrumentation: changes in observers or calibration of equipment statistical regression: tendency for extreme scores to become less extreme selection: non-random assignment attrition: differential loss of subjects diffusion: when the non-tx group gets some tx |
|
threats to construct validity
|
attention and contact with clients: difficult to tell whether the change is due to the technique or contact w/ the therapist
experimenter expectancies (Rosenthal effect): cues or clues transmitted to subjects by experimenter demand characteristics: factors in the procedures that suggest how the subject should behave |
|
threats to external validity
|
sample characteristics: differences b/w the sample and the population
stimulus characteristics: artificiality of research arrangements contextual characteristics: the conditions in which the intervention is imbedded such as reactivity which involves the subjects behaving in a certain way b/w they are subjects (e.g., Hawthorne effect) |
|
threats to statistical conclusion validity
|
low power: a diminished ability to find significant results (e.g., small sample size, inadequate interventions)
unreliability of measures variability in procedures: inconsistency in tx procedures subject heterogeneity |
|
nominal v. ordinal v. interval v. ratio
|
nominal: categories with no inherent order
ordinal: ordered categories interval: equal intervals b/w scores, no absolute zero ratio: equal intervals and an absolute zero |
|
descriptive v. inferential statistics
|
descriptive: description of data
inferential: used to make inferences about the general population |
|
mean v. mode v. median
|
measures of central tendency
mean: average mode: most frequent score median: 50th percentile |
|
standard deviation v. range v. variance
|
measures of variability
standard deviation: average deviation from the mean range: the difference b/w the highest and lowest value variance: standard deviation squared |
|
criterion-referenced v. norm referenced scores
|
criterion-referenced: demonstrates how a person scored relative to a particular criterion (e.g., percentage correct)
norm-referenced: demonstrates how a person scored relative to a group (e.g., percentagile rank) |
|
z scores
|
a standard score that corresponds directly to standard deviation with a distribution identical to the raw score distribution
|
|
standard error of the mean and central limit theorum
|
SEM: if one were to plot the means of an infinate number of equal sized samples, the deviation in the means would be error
central limit theorum: if one were to plot the means of an infinite number of equal sized samples, a normally distributed distribution of means will result |
|
null hypothesis v. alternative hypothesis
|
null hypothesis: states that there are no differences b/w groups
alternative hypothesis: states that there are differences b/w groups |
|
rejection v. retention region
|
rejection region: region at the tail end of the curve, if the mean falls in this region, the researcher must reject the null b/c it is unlikely that this mean was obtained by chance
retention region: the null must be retained if the mean falls in this region |
|
alpha v. beta
|
alpha: the size of the rejection region, directly corresponds to the likelyhood of making a Type I error
beta: the probability of making a Type II error (Power = 1 - beta) |
|
Type I v. Type II error
|
Type I: the null is mistakenly rejected
Type II: the null is mistakenly retained |
|
power
|
1 - beta, the ability to correctly reject the null, increased when sample size is large, the magnitude of the intervention is large, random error is small, the statistical test is parametric, and the test is one-tailed
|
|
t-test v. ANOVA
|
t-test: interval or ratio data are collected for one or two groups of subjects
ANOVA: interval or ratio data are collected for more than two groups of subjects |
|
one-way ANOVA v. factorial ANOVA
|
one-way ANOVA: interval or ratio data are collected for more than two groups of subjects with only 1 IV
factorial ANOVA: |
|
split plot ANOVA v. randomized block ANOVA v. repeated measures ANOVA
|
split plot ANOVA:
randomized block ANOVA: repeated measures ANOVA: |
|
MANOVA v. ANCOVA
|
MANOVA: more than one DV
ANCOVA: |
|
main v. interaction effects
|
main effects: the effect of a single IV
interaction effects: the effect of an interaction b/w the two IVs |
|
trend analysis
|
when the IV is quantitative, the outcome frequently is nonlinear, trend analysis describes the trends of the data or the ups and downs
|
|
bivariate v. multivariate correlation
|
bivariate correlation: looks at the relationship b/w two variables where neither is an IV is the truest sense, x predicts y
multivariate correlation: involves several predictors(IVs) and one or more criterions (DVs) |
|
least squares criterion
|
A
|
|
Pearson r v. eta v. biserial correlation
|
Pearson r: a measure of how well a linear equation describes the relationship b/w X and Y when both X and Y are interval or ratio
eta: used to calculate the correlation b/w x and y when it is thought that they have a curvilinear relationship biserial correlation: a measure of how well a linear equation describes the relationship b/w X and Y when X is interval or ratio and Y is dichotomous |
|
zero order v. partial v. semipartial correlation
|
zero order: analyzes the relationship b/w X and Y when there are no extraneous variables affecting the relationship
partial: analyzes the relationship b/w X and Y with the effect of a third variable being removed semipartial correlation: analyzes the relationship b/w X and Y with the effect of a third variable being removed from only one of the origonal variables |
|
multiple R v. canonical v. discriminant v. loglinear correlation
|
multiple R: a correlation b/w two or more IVs and one DV where Y is always interval or ratio and at least one X is interval or ratio, squaring multiple R generates the coefficient of determination
canonical R: a correlation between two or more IVs and two or more DVs discriminant: two or more predictors and one criterion which is nominal loglinear: used to predict a categorical criterion based on categorical predictors |
|
multicollinearity
|
a problem that occurs in a multiple regression equation when the predictors are highly correlated with one another and therefore essentially redundant
|
|
correlation v. regression
|
correlation: measures the relationship b/w two variables where neither is an IV in the truest sense
regression: describes the relationship b/w two variables where neither is an IV in the truest sense by creating the line of best fit |
|
path analysis v. LISREL
|
path analysis: applies multiple regression techniques to testing a model that specifies causal links among variables
LISREL: enables researchers to make inferences about causation, can be used to test many different causal pathways involving multiple predictors and criterion variables |
|
orthogonal v. oblique
|
concepts related to factor analysis
orthogonal: perpendicular, results in no correlation oblique: non-perpendicular, results in correlation b/w factors |
|
factor analysis v. cluster analysis
|
factor analysis: extracts significant factors (dimensions) from the data
cluster analysis: statistically looking for naturally occuring subgroups in data collected on a variety of dependent variables |
|
reliability v. validity
|
reliability: amount of consistency, repeatability, and dependability in scores obtained on a given test
validity: meaningfulness, usefulness, or accuracy |
|
test-retest v. parallel form v. internal consistency v. interrater reliability
|
test-retest: correlating pairs of scores from the same sample of people who are administered the identical test at 2 points in time
parallel form: correlating the scores obtained by the same group of people on 2 roughly equivalent but not identical forms of the same test administered at 2 different points in time internal consistency: consistency of scores w/in the test, includes split-half and kuder-richardson or chronback's alpha (correlation of each item w/ every other item) interrater reliability: degree of agreement b/w 2 or more scorers when a test is subjectively scored |
|
Spearman-Brown prophecy formula
|
when calculating split-half reliability, use the Spearman-Brown to determine how much more reliable the test would be if it were longer
|
|
split-half v. coefficient alpha and Kuder-Richardson
|
types of internal consistency reliability
split-half: calculated by splitting the test in half and then correlating the scores obtained on each half by each person coefficient alpha and Kuder-Richardson: analyze the correlation of each item with every other item in the test |
|
standard error of measurement
|
the standard deviation of a theoretically normal distribution of test scores obtained by one individual on equivalent tests
|
|
calculating confidence intervals
|
the standard error of measurement is added and subtracted from the actual score once for 68% confidence, twice for 95% confidence, and three times for 99% confidence
|
|
content v. criterion-related v. construct validity
|
content: how adequately a test samples a particular content area
criterion-related: how adequately a test score can be used to infer, predict, or estimate criterion outcome (concurrent or predictive) construct validity: how adequately a new test measures a construct or trait (convergent or divergent) |
|
concurrent v. predictive validity
|
concurrent validity: the predictor and criterion are measured and correlated at about the same time
predictive validity: there is a delay b/w the measurement of the predictor and the criterion |
|
standard error of estimate
|
the standard deviation of a theoretically normal distribution of criterion scores obtained by one person measured repeatedly, maximum value is the SD of the criterion
|
|
Taylor-Russell tables
|
numerically describe the amount of improvement that occurs in selection decisions when a predictor test is introduced
base rate: the rate of selecting successful employees w/o using a predictor test selection ratio: the proportion of available openings to available applicants incremental validity: the amount of improvement in success rate that results from using a predictor test |
|
false positives v. true positives v. false negatives v. true negatives
|
false positives: those incorrectly identified as possessing what is being measured
true positives: those correctly identified as possessing what is being measured false negatives: those incorrectly identified as not possessing what is being measured true negatives: those correctly identified as not possessing what is being measured |
|
multi-trait multi-method matrix
|
a table with information about convergent and divergent validity, both of which are necessary to establish construct validity
|
|
convergent v. divergent (discriminant) validity
|
convergent validity: the correlation of scores on the new test with other available measures of the same trait, must be moderate to high
divergent (discriminant) validity: the correlation of scores on the new test with scores on another test that measures a different trait or construct, should be low |
|
classical test theory v. item response theory (item characteristics curve)
|
classical test theory: any obtained score is a combination of truth and error (reliablity concept)
item response theory (item characteristics curve): used to calculate to what extent a specific item on a test correlations with an underlying contstruct (validity concept) |