Advanced Statistics Session: Which tool do I use? Institute on Teaching and Mentoring Lewis VanBrackle, Ph.D. Professor of Statistics October 30, 2009.

Advanced Statistics Session: Which tool do I use?Advanced Statistics Session: Which tool do I use?

Institute on Teaching and MentoringInstitute on Teaching and Mentoring

Lewis VanBrackle, Ph.D.Lewis VanBrackle, Ph.D.

Professor of StatisticsProfessor of Statistics

October 30, 2009October 30, 2009

Which Tool Do I Use?Which Tool Do I Use?

In this session, we will be discussing In this session, we will be discussing the core tools in the statistical toolbox the core tools in the statistical toolbox that you will need to defend your that you will need to defend your hypotheses. hypotheses.

Prior to thinking about your statistical Prior to thinking about your statistical toolbox, you should have completed toolbox, you should have completed the following:the following:

1.1.Identified your research question(s)Identified your research question(s)

2.2.Developed a theoretical Developed a theoretical representation of your modelrepresentation of your model

3.3.Identified the data to be collected…Identified the data to be collected…

Which Tool Do I Use?Which Tool Do I Use?

Independent T-testIndependent T-test

Mann-Whitney TestMann-Whitney Test

Paired T-testPaired T-test

Wilcoxon Matched PairsWilcoxon Matched Pairs

One Way ANOVAOne Way ANOVA

Kruskall Wallis TestKruskall Wallis Test

Repeated Measures ANOVARepeated Measures ANOVA

Friedman’s ANOVAFriedman’s ANOVA

Pearson Correlation or RegressionPearson Correlation or Regression

Spearman Correlation or Kendall’s TauSpearman Correlation or Kendall’s Tau

Ind. Factorial ANOVA or RegressionInd. Factorial ANOVA or Regression

Factorial Repeated Measures ANOVAFactorial Repeated Measures ANOVA

Factorial Mixed ANOVAFactorial Mixed ANOVA

Multiple RegressionMultiple Regression

Multiple Regression/ANCOVAMultiple Regression/ANCOVA

Pearson Chi-Square or Likelihood RatioPearson Chi-Square or Likelihood Ratio

Logistic RegressionLogistic Regression

Loglinear AnalysisLoglinear Analysis

MANOVAMANOVA

Factorial MANOVAFactorial MANOVA

MANCOVAMANCOVA

YesYes

NoNo

NoNo

YesYes

NoNo

YesYes

YesYes

YesYes

YesYes

YesYes

YesYes

YesYes

NoNo

YesYes

NoNo

YesYes

NoNo

YesYes

DifferentDifferent

SameSame

SameSame

BothBoth

DifferentDifferent

DifferentDifferent

DifferentDifferent

SameSame

DifferentDifferent

YesYes

DifferentDifferent

TwoTwo

Three +Three +

CategoricalCategorical

ContinuousContinuous

BothBoth

CategoricalCategorical

CategoricalCategorical

BothBoth

BothBoth

ContinuousContinuous

CategoricalCategorical

ContinuousContinuous

CategoricalCategorical

CategoricalCategorical

ContinuousContinuous

OneOne

Two +Two +

OneOne

Two +Two +

OneOne

Two +Two +

ContinuousContinuous

CategoricalCategorical

ContinuousContinuousTwo +Two +

How Many How Many DependentDependentVariables?Variables?

OneOne

What TypeWhat TypeOf DependentOf DependentOutcome?Outcome?

How ManyHow ManyPredictors?Predictors?

What type What type Of predictors? Of predictors?

If Categorical If Categorical Predictor,Predictor,How many How many Categories? Categories?

If Categorical If Categorical Predictor,Predictor,Same Participants Same Participants or Different in each or Different in each category? category?

Does Data Does Data MeetMeetParametric Parametric Assumptions? Assumptions?

ANALYSIS TOOL ANALYSIS TOOL

Logistic Regression/DiscriminantLogistic Regression/Discriminant