Univariate Bivariate Multivariate - Youngstown State University

[Pages:4]Overview

Analysis Strategy

Univariate (One variable)

Bivariate (Two variables)

Multivariate (> 2 variables)

1

Overview

Univariate (One variable)

Continuous Variable

Categorical Variable

Central Tendancy

Variation

Distribution

Plots

Frequencies

Plots

Mean

(C.I., t-test, Signed Rank test)

Median

(Signed Rank test)

Mode

Variance

Standard Deviation

Range

Percentiles

Interquartile Range

Normal (Normality test) Uniform

Histogram Box Plot

Exponential

...

(Goodness of fit test for testing distribution)

Stem Plot

Dot Plot

Line Chart

Time Series Plot

Counts

Bar Graph

Frequencies Pie Graph

Odds

Pareto

Graph

Percentages

(C.I.,

z-test for proportion,

Goodness of Fit test,

Binomial test)

For paired sample design, t-test and signed rank test can be used to test for the mean of the paired differences. In this case, the one variable is the paired difference.

2

Overview

Categorical Y Categorical X

Bivariate (Two variables X & Y)

Continuous Y Categorical X

Continuous Y Continuous X

Y-2 Categories X-2 Categories

Pearson's Chi-square

Fisher's Exact

McNemar's Test

Y or X are > 2 categories

Y-Normal X-2 Categories

Y-Normal X>2 Categories

Pearson's Chi-square

MantelHaenszel

Y-Non-normal X-2 Categories

Y-Non-normal X>2 Categories

2 Independent Samples t-test

Paired-Sample Test (related samples)

Wilcoxon One-way Rank-sum ANOVA

Signed-rank Test (related samples)

KruskalWallis Test

Y and X Normal

Y or X Non-normal

Scatter plot

Simple Linear Regression

Pearson's Correlation

Spearman's Correlation

Y = Dependent, Outcome, or Response Variable; X = Independent variable, Explanatory variable

3

Overview

Multivariate (More than two variables)

Continuous Y

Dichotomous Y

Nominal Y > 2 Categories

Multiple Regression

Analysis of Variance

Analysis of Covariance

Logistic Regression

Discriminant Analysis

Multinomial Logistic

Ordinal Y

Ordinal Logistic

Y is "Time" Survival Analysis

Multivariate Y

Life Table

Cox Proportional Hazards Model

Repeated Measures

MANOVA

Factor Analysis

Y = Dependent, Outcome, or Response Variable; X = Independent variable, Explanatory variable

4

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download