Scribd
Upload
97 views1 page

Which Statistical Tests To Use

This document provides an overview of common statistical tests and their appropriate uses based on the level of measurement and data type. It lists statistical tests for nominal, ordinal, interval, and ratio level data, whether the data is discrete or continuous. For each test, it gives the requirements or limitations and a brief statement of purpose. The document serves as a reference for selecting the proper statistical test based on the nature of the data.

Uploaded by

Pam Fajardo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
97 views1 page

Which Statistical Tests To Use

This document provides an overview of common statistical tests and their appropriate uses based on the level of measurement and data type. It lists statistical tests for nominal, ordinal, interval, and ratio level data, whether the data is discrete or continuous. For each test, it gives the requirements or limitations and a brief statement of purpose. The document serves as a reference for selecting the proper statistical test based on the nature of the data.

Uploaded by

Pam Fajardo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 1

Which Statistical Tests to Use?

Level of
Measurement
& Data Type
Level
Nominal
(Non
Parametric)
Type
Discrete
Grouped

Level
Ordinal (NON
Parametric)
Type
Data arranged
on interval
scales which
have been given
ranks

Level
Ordinal (Non
Parametric)
Type
Data arranged
on interval
scales which
have been given
ranks
Level
Interval
(Parametric)
Type
Continuous
Individual

Level
Ratio
(Parametric)
Type
Continuous
Individual

Statistical Test
Chi square tests for
one, two sample of k
independent samples
Binomial test for one
sample
Fisher Exact
Probability Test
Contingency
Coefficient
Mann Whitney U
Test for two
independent samples.
Almost as powerful as
t- test.
Wilcoxon test for two
related samples.
Median Test.
Signs test.
Kolmogorov- Smimov
two sample test.

Requirements/ Limitation(s)
80% cells need 5 or more expected
cases

To test if observed frequencies differ from


expected frequencies

Can use medians and quartiles. No


frequency requirements. Use when
chi square requirements not met.
Use when small N instead of chi
square.
As for chi square.

Tests yes/no data for nonrandomness of


small, one sample cases. Useful test for
small subpopulation.
As for chi square.

Use when t-test requirements not


met. Use for small N but not for
many tied values. Two samples.
Needs continuously distributed data.

As for U-test but for paired samples.

Randomness; independent sets of


data. Data less normal than
Wilcoxon.
As for median test. Needs relatively
continuous data.

Tests for differences and significance between


two sets of data. Before/ after type situations.

Independent random sample (a) two


samples, (b) k samples

Binomial (proportions)
test
Spearmans rank
correlation coefficient
Kendalls rank
correlation

Data in percentages or other


proportions
Cannot handle too many paired
values.
Can handle many paired values.

Assumes normal distribution

Approximately normal distribution

Standard Error of
means
Regression
Analysis of Variance
(ANOVA)
Pearsons Product
Moment Correlation

To test degrees of association (correlation)


between two variables.
Tests for central tendency, and whether two
independent samples from the same population.
Tests for differences and similarities.

Needs relatively continuous data.


As for U test but for paired values.
Use when U test has too many tied
values

(a) Friedman test


(b) Kruskal Wallis
test
Runs Test

t-test for two


independent samples,
and matched pairs
Standard, Z scores

Purpose

Approximately normal distribution.

As for U-test. Uses when U-test requirements


not met.

Tests similarities of distributions. Alternative


to chi square where small sample and many
tied values.
Analysis of Variance
To test number of runs in a sequence. To test
number of runs which may be considered
random. Tests for differences in dispersion.
As for z scores
To test correlation between variables having
paired values.
To test correlation between variables having
paired values.
To test match means between samples or
groups. Only slightly more powerful than Utest.
To test probability of an occurrence, or to
compare places of individuals on two
distributions.
To determine confidence limits for samples.
To show graphically, relationship between
variables, and likely trends or prediction
To test significance of differences between
means.
To test correlation between variables. Very
Powerful.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy