David Normando A PowerPoint®-based guide to assist in choosing the suitable statistical test. NOTE:...

Post on 30-Mar-2015

227 views 1 download

Tags:

Transcript of David Normando A PowerPoint®-based guide to assist in choosing the suitable statistical test. NOTE:...

David Normando

A PowerPoint®-based guide to assist in choosing the suitable statistical test.

NOTE: This presentation has the main purpose to assist researchers and students in choosing the

appropriate statistical test for studies that examine one variable (Univariate). Some multivariates analyses are also included.

Please proceed to the next page ...

If you have any suggestion, criticism, please contact the author by e-mail: davidnor@amazon.com.br

David Normando

What do you want to do? For an answer, click on the button

1) I want to assess whether my data have a Normal distribution

2) I want to compare groups (Looking for differences between samples)

3) I want to make correlation or regression analysis between variables.

4) I want to check the replicability of data (analysis of random and systematic error)

5) I would choose the appropriate graph to my data.

David Normando

Tests for Data Analysis Distribution- NormalityNormal distribution is requested when using continuos data and n<30Normal distribution is requested when using continuos data and n<30

You may choose the test according to sample size.

Use D’Agostino, Use D’Agostino, ifif n n≥≥1010

Use D’Agostino-Pearson, Use D’Agostino-Pearson, ifif nn≥≥2020

Use Lilliefors or Shapiro-Wilk, Use Lilliefors or Shapiro-Wilk, for anyfor any n value n value

Back to beginningBack to beginning

David Normando

What kind of data you have? (Click on the button).

Nominal

Ordinal

NUMERICALNUMERICALContinuousContinuous

NonparametricCategorical data

Parametric(mean)

Ex: height / length / weight (Assuming a normal distribution on n>30)

Ex: Frequency: Yes / No Race Gender

Ex: Middle (1) / Moderate(2) Severe (3)

Comparing groups (samples)

Back to beginningBack to beginning

How to check Normality ?How to check Normality ?

David Normando

Numerical Data (parametric) If the distribution is not Normal,

skip to "Abnormal" ABNORMAL

How many groups (samples) do you have?

1

2

>2

David Normando

Are your samples paired or dependent?

No

Yes

Not sure? Not sure? Dependent Samples mean: Dependent Samples mean:

Before X After Before X After Left Side X Right Side Left Side X Right Side

T1 x T 2 x T3T1 x T 2 x T3

Numerical Data (parametric) If the distribution is not Normal, skip to "Abnormal"

ABNORMAL

David Normando

Answer: one sample t test

Back to beginningBack to beginning

Numerical Data (parametric) If the distribution is not Normal,

skip to "Abnormal" ABNORMAL

David Normando

Answer: Independent t testor ANOVA.

Back to beginningBack to beginning

Numerical Data (parametric) If the distribution is not Normal,

skip to "Abnormal" ABNORMAL

David Normando

Answer: Paired t test or ANOVA for repeated measurements..

Back to beginningBack to beginning

Numerical Data (parametric) If the distribution is not Normal,

skip to "Abnormal" ABNORMAL

David Normando

Answer: Analysis of Variance (ANOVA)or MANOVA (Multiple Analysis of Variance), if you have >1 variable.

Numerical Data (parametric) If the distribution is not Normal,

skip to "Abnormal" ABNORMAL

Back to beginningBack to beginning

David Normando

Ordinal Categorical Data (Nonparametric)

2

>2

How many groups (samples) do you have ?

David Normando

No

Yes

Ordinal Categorical Data (Nonparametric)

Not sure? Not sure? Dependent Samples mean: Dependent Samples mean:

Before X After Before X After Left Side X Right Side Left Side X Right Side

T1 x T 2 x T3T1 x T 2 x T3

Are your samples paired or dependent?

David Normando

No

Yes

Ordinal Categorical Data (Nonparametric)

Not sure? Not sure? Dependent Samples mean: Dependent Samples mean:

Before X After Before X After Left Side X Right Side Left Side X Right Side

T1 x T 2 x T3T1 x T 2 x T3

Are your samples paired or dependent?

David Normando

Answer: Mann-Whitney test

Back to beginningBack to beginning

Ordinal Categorical Data (Nonparametric)

David Normando

Answer: Wilcoxon (signed rank test) or Signal test.

Back to beginningBack to beginning

Ordinal Categorical Data (Nonparametric)

David Normando

Answer: Kruskal-Wallis’ Test

Back to beginningBack to beginning

Ordinal Categorical Data (Nonparametric)

David Normando

Answer: Friedman’s Test

Back to beginningBack to beginning

Ordinal Categorical Data (Nonparametric)

David Normando

Nominal Categorical Data (Nonparametric)

2

>2

How many groups (samples) do you have ?

David Normando

No

Yes

Nominal Categorical Data (Nonparametric)

Not sure? Not sure? Dependent Samples mean: Dependent Samples mean:

Before X After Before X After Left Side X Right Side Left Side X Right Side

T1 x T 2 x T3T1 x T 2 x T3

Are your samples paired or dependent?

David Normando

No

Yes

Nominal Categorical Data (Nonparametric)

Not sure? If some of the cells in the contingency table give

values (expected) lower than 5.

Is there any expected value <5 ?

David Normando

No

Yes

Nominal Categorical Data (Nonparametric)

Not sure? Not sure? Dependent Samples mean: Dependent Samples mean:

Before X After Before X After Left Side X Right Side Left Side X Right Side

T1 x T 2 x T3T1 x T 2 x T3

Are your samples paired or dependent?

David Normando

Answer: Chi-square (x²) test or Binomial Test, if using 2 samples and proportion (%)

Back to beginningBack to beginning

Nominal Categorical Data (Nonparametric)

David Normando

Answer: Cochran’s test (absolute or relative frequence: %)

Nominal Categorical Data (Nonparametric)

Back to beginningBack to beginning

David Normando

Answer: McNemar’s test

Back to beginningBack to beginning

Nominal Categorical Data (Nonparametric)

David Normando

Answer: Exact Fisher’s test

Nominal Categorical Data (Nonparametric)

Back to beginningBack to beginning

David Normando

Correlation or Regression Analysis

Ordinals

Numerical

Nominal

What kind of data you have? (Click on the button).

Ex: height / length / weight (Assuming a normal distribution)

Ex: Frequency: Yes / No Race Gender

Ex: Middle (1) / Moderate(2) Severe (3)

How to check Normality ?How to check Normality ?

NonparametricCategorical data

Parametric(mean)

David Normando

Numerical Data (parametric) If the distribution is not Normal,

skip to "Abnormal" ABNORMAL

How many variables do you have?

2

>2

David Normando

Answer: Pearson’s Correlation

Simple Linear Regression

Back to beginningBack to beginning

ABNORMAL

Correlation tests or regression analysis to Continuos data

If the distribution is not Normal, skip to "Abnormal"

David Normando

Answer: Pearson’s Correlation (parcial) or Canonical Correlation

Multiple Linear Regression

Back to beginningBack to beginning

ABNORMAL

NOTE: For Correlation all variables examined must have a Normal Distribution.For Linear Regression dependent variable must have a Normal Distribution

How to check Normality ?How to check Normality ?

Correlation tests or regression analysis to Continuos data

If the distribution is not Normal, skip to "Abnormal"

David Normando

Answer : Spearman or Kendal Correlation

Back to beginningBack to beginning

Correlation test to Ordinal data

(nonparametric)

David Normando

How many variables do you have?

2

>2

Correlation and Regression Analysisto Nominal data (nonparametric)

David Normando

Answer: Contingency coefficient C

Simple Logistic Regression

Correlation test to Nominal data

(nonparametric)

Back to beginningBack to beginning

David Normando

Answer: Contingency coefficient C

Multiple Logistic Regression

Correlation test to Nominal data

(nonparametric)

Back to beginningBack to beginning

David Normando

Replicability or Reproducibility (Systematic error)

OrdinalOrdinal

NumericalNumerical

NominalNominal

What kind of data you have? (Click on the button).

Ex: height / length / weight (Assuming a normal distribution)

Ex: Frequency: Yes / No Race Gender

Ex: Middle (1) / Moderate(2) Severe (3)

NonparametricCategorical data

Parametric(mean)

How to check Normality ?How to check Normality ?

David Normando

Answer: Parametric test for dependent data

Note: Intraclass correlation can be used, if you would like to check the association between 2 or more measurements.

Replicability or Reproducibility (Systematic error for numerical data)

2 samples2 samples >2 samples>2 samples

ABNORMAL

Back to beginningBack to beginning

For random or casual error , you may use TEM (technical error measurement):

D= difference between repeated measures n=number of individuals

David Normando

Answer: Weighted Kappa

Replicability or Reproducibility (Systematic error for ordinal data)

2 sample > 2 samples

NOTE: in case of an ordinal variable, nonparametric tests for paired or dependent data can also be used

Back to beginningBack to beginning

David Normando

Answer: Kappa

Back to beginningBack to beginning

Replicability or Reproducibility (Systematic error for Nominal data)

David Normando

Nominal

Ordinal

NumericalParametric(mean)

Graph Selection

What kind of data you have? (Click on the button).

Ex: height / length / weight (Assuming a normal distribution)

Ex: Frequency: Yes / No Race Gender

Ex: Middle (1) / Moderate(2) Severe (3)

Back to beginningBack to beginning

NonparametricCategorical data

David Normando

Comparing Independent Samples

Making Data Correlation or regression

Comparing Dependent Samples (paired)

Graph Selection

Not sure? Not sure? Dependent Samples mean: Dependent Samples mean:

Before X After Before X After Left Side X Right Side Left Side X Right Side

T1 x T 2 x T3T1 x T 2 x T3

David Normando

BOX-PLOT - Comparing Groups.Continuous or Ordinal Data (Score)

This chart describes the measure of central tendency (MEAN for continuos data or MEDIAN for Ordinal This chart describes the measure of central tendency (MEAN for continuos data or MEDIAN for Ordinal data), measures of dispersion (Standard deviation for parametric data or interquartiles deviation for data), measures of dispersion (Standard deviation for parametric data or interquartiles deviation for

Ordinal data) and the whiskers (maximum and minimum values )Ordinal data) and the whiskers (maximum and minimum values )

BackBack

Outlier: an observation that Outlier: an observation that is numerically distantis numerically distantfrom the rest of the data. from the rest of the data.

David Normando

Line graph for longitudinal data

This chart describes the measure of central tendency (mean for Continuos data or median This chart describes the measure of central tendency (mean for Continuos data or median for Ordinal data) longitudinallyfor Ordinal data) longitudinally

BackBack

David Normando

Graphic for Correlation Tests or Regression Analysis

BackBack

David Normando

Bar/Column GraphicNominal data (frequency)

BackBack