Term
|
Definition
**Defn of a concept or variable on the basis of the specific measures used in the RS project
1. Captures only a poriton of the conpet
2. May includ some irrelevant correlated features
3. Use of a single measure means the defn relies heavily upon the specific and unique features of the measure |
|
|
Term
|
Definition
Nonparametric test of statistical significance
> 2 Independent Samples
nonparametric, one-way ANOVA for rank-order data
Based on MEDIANS
Symbolized H |
|
|
Term
|
Definition
Nonparametric test of statistical significance
Ordinal data from correlated groups
> 2 groups
Nonparametric version of one-way, Repeated Measures ANOVA |
|
|
Term
|
Definition
Test of statistical significance of differences b/n 2 groups
Used when data for 2 samples are measured on an ordinal scale
Nonparametric equivalent to to t-test
Ordinal measures are used, but underlying continuous distribution of data is assumed
Used instead of t-test with interval-level data when researchers do not assume that the populations are normal |
|
|
Term
|
Definition
aka "signed-rank" or "rank-sum" test for ordinal data
nonparametric test of statistical significance for use with 2 correlated samples, such as the same subjects on a before-and-after measure
2 groups
Rank data
nonparametric equivalent for t-test for dependent samples |
|
|
Term
|
Definition
Nonparametric test of statistical significance for ordinal data
Used to determine whether 2 distributions differ and whether 2 samples may reasonably be assumed to come from the same population
Goodness-of-fit test |
|
|
Term
| Which tests are used for tests of differences b/n groups? (independent samples) |
|
Definition
1) Mann Whitney U test
2) Kolmogorov-Smirnov Test
3) Kruskal-Wallis Test |
|
|
Term
Name the:
1) # of groups
2) Type of data
3) Parametric test
For the following non-parametric tests:
a) Mann Whitney U test
b) Kolmogorov-Smirov test
c) Kruskal-Wallis Test
What are these tests used for? |
|
Definition
a) Mann Whitney U test
1) 2 groups
2) type of data: Rank
3) Parametric equiv: t-test for independent samples
b) Kolmogorov-Smirnov Test:
1) # groups: 2
2) type of data: rank
3) nonparametric equiv: t-test for independent samples
c) Kruskal-Wallis Test:
1) # groups: 3+
2) type of data: rank
3) parametric test: ANOVA/MANOVA
TESTS OF DIFFERENCES BETWEEN GROUPS (INDEPENDENT SAMPLES) |
|
|
Term
Name the:
1) # groups
2) type of data
3) parametric test
For the following nonparametric tests:
a) Wilcoxon Matched Pairs
b) Chi-Square
c) Friedman Test
WHAT ARE THEY USED FOR? |
|
Definition
1) Wilcoxon Matched Pairs:
a) # groups: 2
b) types of data: rank
c) parametric test equiv: t-test for dependent samples
2) Chi-Square
a)# groups: 2
b) type of data: dichotomous (categorical)
c) parametric test: t-test for dependent samples
3) Friedman Test
a) # groups: 3+
b) type of data: rank
c) parametric test equiv: Repeated Measure ANOVA
TESTS OF SIGNIFICANCE BETWEEN VARIABLES (BETWEEN SAMPLES) |
|
|
Term
Name the:
1) # groups
2) type of data
3) parametric test equivalent
For the following nonparametric tests:
a) Spearman
b) Chi-Square
What are these used to test? |
|
Definition
1) Spearman:
a) # groups: 2
b) type of data: rank
c) parametric test equiv: Correlation
2) Chi-Square
a) # groups: 2
b) type of data: dichotomous (categorical)
c) Chi-Square
TESTS RELATIONSHIPS BETWEEN VARIABLES |
|
|
Term
| What tests are used to test the homogeneity of variance (3)? |
|
Definition
a) Levene: Assess the differences of scores from each of their MEANS
b) Brown Forsythe: like Levene's but looks at the differences of scores from the MEDIAN
c) F Max: biggest variance/smallest variance --> If biggest variance is 4-10x greater, then you have a violation! |
|
|
Term
T-test are used for what?
What are types of t-tests? |
|
Definition
Difference between 2 group MEANS
Independent Samples T Test
Paired Samples T Test |
|
|
Term
| Describe the Flow chart to determine whether t Test is the correct statistic to use. |
|
Definition
1) Are you examining the difference between variables or between groups on 1+ variables?
2a) Relationships between variables
3a) How many variables?
4a) 2
5a) t-test for the significance of the correlation coefficient
4b) > 2 variables
5b) Regression, Factor analysis, Canonical analysis
2B) Differences b/n groups on 1+ variables
3B) Yes
4B) How many groups are you dealing with?
5B) 2 groups
6B) t-test for Dependent samples
5C) > 2 groups
6C) Related measures of ANOVA
3D) No
4D) How many groups are you dealing with?
5D) 2 groups
6D) t-test for Independent samples
5E) > 2 groups
6E) Simple ANOVA |
|
|
Term
|
Definition
Between groups: Sum of squared deviations of GROUPS MEANS from GRAND MEAN
Within groups: Sum of squared deviations of INDIVIDUAL SCORES from GROUP MEANS |
|
|
Term
| What are the 3 types of magnitude of effect? |
|
Definition
1) Omega squared: measure of strength of association; the proportion of variability in the dependent variable associated with the variability in the independent variable
-ranges from 0 to 1
--> when it is 0, knowing X tells us nothing about Y; when it is 1, knowing X lets us predict Y exactly
2) R2: coefficient of multiple determination b/n a DV and 2+ IVs
-measure of goodness-of-fit of a linear model
3) Eta squared: a measure of how much the variance in a DV (measured at the interval level) can be explained by a categorical (nominal, discrete) IV |
|
|
Term
| Which statistic do you use to test the independence of observation for ANCOVA? |
|
Definition
|
|
Term
| Mahalanobi’s (statistic for measuring outliers) |
|
Definition
gives you a number for each subject telling you how far away they are from
the theoretical center of data |
|
|
Term
| Leverage (statistic for measuring outliers) |
|
Definition
gives you a number for each subject telling you how far they are from all other
subjects |
|
|
Term
| DfBeta (statistic for measuring outliers) |
|
Definition
gives you a number for each subject on every predictor telling you how much that
subject influenced each predictor’s regression weight |
|
|
Term
| Cook's (statistic for measuring outliers) |
|
Definition
gives you a number for each subject telling you how much they influenced the
regression equation as a whole |
|
|
