1

MODULE 8

STATISTICAL TEST AND THEIR ISSUES

I. Introduction

This module for independent study in Educational Statistics aims to provide the graduate
students with reading materials relative to the different statistical tests and their issues.

II. Learning Concept

Statistical Tests and Their Issues

General types of Statistical Tests: the parametric and the non-parametric.

III. Learning Objectives

After working this module, the graduate students are expected to:

1. Gain understanding on the different statistical tests and their issues.

2. Analyze how the Z-Test of one Sample Mean is computed.

3. Explain the uses Z-Test for Dependent Sample Mean.

4. Discuss T-Test for Dependent Sample Mean differs from T-Test for

Independent Sample Mean.

5. Analyze the different types of tests and be able to identify their uses.

IV. Learning Activities/Methodologies

Concept Reading

Independent Study

Organizing Ideas or Key Ideas

Answering the Guide Questions

V. Concept Map

Topics/Concepts Objectives Means Evaluating Study

Guide
Statistical tests and Analyze how other -Independent Reading How do you know
their issues Statistics tests are -Concept Reading what kind of test to
used. use?
1. The Z-Test of One Analyze the Z-Test of -Answering the guide How do we compute
Sample Mean one sample mean questions for Z-Test of one
is composed. sample mean? Give
an example
 2

2. T-Test for Gain knowledge about -Concept Reading Discuss the features
Independent Sample T-Test for independent of the T-Test for
Mean sample means independent sample
mean
3. T-Test for Differentiate T-Test Independent reading, How does T-Test for
Dependent Sample for Dependent Sample Independent study, dependent sample
Mean Means from T-Test for Computing a given means differ from T-
Independent Sample data Test for Independent
Mean Sample Mean?
4. One-Way Analysis Analyze how the one- Independent Study Give the salient
of Variance way analysis of Concept reading features of One-Way
variance and the two- Solving Problems Analysis of Variance
5. Two-Way way analysis of and two-way
Analysis of variance are Analysis of
Variance computed. Variance.
6. The Person Analyze/determine Computing the When do we use the
Product-Moment whether there is a Person Product- Person Product-
Correlation correlation of variable Moment Correlation Moment
to the other Concept Reading Correlation?
7. The Chi-Square Gain knowledge about Concept Reading How do we compute
the Chi-Square test. For the Chi-Square
Test?
8. The Chi-Square Analyze and determine Independent Reading How do we compute
Goodness-of-Fit Test if an observed Concept reading for the Chi-Square
frequency distribution Computing for the Goodness-of-Fit
or a variable differs Chi-Square Goodness Test?
significantly from an of-Fit Test.
expected theoretical
distribution of
frequencies.
9. The Chi-Square Analyzing/determining Concept reading Compute for the
Test of Association whether or not two Computing for the Computing for the
variables are Chi-Square Test of Chi-Square Test of
associated for Association. Association.
dependent or
correlated with each
other.

VI. Learning Context (Key Concepts)

CHUNK 1
Some Statistical Tests and Their Issues
There are two general types of statistical tests: the parametric and the non-
parametric. Parametric tests are usually used for data that are of the interval or ratio levels
of measurement. In addition, the within group scatter of scores of each of the groups to be
compared should be equal, and that the population distributions from where the samples
were obtained must be bell-shaped, normal or symmetrical. Although these two
additional quantities are still mentioned as assumptions that must be satisfied before a
parametric test is run, these requirements are overly emphasized. Present studies show
that they are not
 3

important at all. Non parametric tests are sometimes called distribution-free statistics
since their issues do not always depend on some specific type of score distribution like a
normal curve. The following section presents some selected parametric and non-
parametric tests that are frequently used in behavioral research.

CHUNK 2

TYPES OF STATISTICAL TESTS

Now that you have looked at the distribution of your data and perhaps conducted
some descriptive statistics to find out a mean, median or mode, it is time to make some
inferences about the data. As previously covered in the module inferential statistics are
the set of statistical tests we use to make inferences about data. These statistical tests
allow us to make inferences because they can tell us if the pattern we are observing is real
or just due to chance.

How do you know what kind of test to use?

Types of Statistical Tests: There are a wide range of statistical tests. The decision
of which statistical test to use depends on the research design, the distribution of the data
is normally distributed, you will choose from parametric tests. If the data is non-normal,
you choose from the set of non-parametric tests. Below is the table listing just a few
common statistical tests and their issues.

Types of Tests: Use:

Correlational These tests look for an association
between variables
Person Correlation Tests for the strength of the association
between two variables.
Spearman Correlation Tests for the strength of the association
between two ordinal variables (does not
rely on the assumption of normal
distributed data)
Chi-Square Tests for the strength of the association
between two categories variables.
Comparison of Means: look for the difference between the means of variables.
Paired T-Test Tests for the difference between two
related variables.
Independent T-Test Tests for the difference between two
independent variables.
ANOVA Tests the difference between group means
after any other variance in the outcome
variables is accounted for.
Regression: assess if change in one variable predicts change in another variable.
Simple Regression Tests how change in the predictor variable
predicts the level of change in the
outcome variable.
4
 Multiple Regression Tests how change in the combination of two
or more predictor variables predict the level
of change in the outcome variable.
Non-parametric: are used when the data does not meet assumptions required for parametric
tests.
Wilcoxon rank-sum test Tests for difference between two related
variables – takes into account magnitude and
direction of difference.
Sign Test Tests if two related variables are different –
ignores magnitude of change, only takes into
account direction.

VII. Synthesis/Summary

Inferential statistics are the set of statistical test allow us to make inferences about data.
These statistical tests allow us to make inferences because they can tell us if the pattern we are
observing is real or just due to chance.

There are a wide range of statistical tests. The decision of which statistical test to use
depends on the research design, the distribution of data, the type of variable. In general, if the
data is normally distributed you will choose from parametric tests. If the data is non-normal you
choose from the set of non-parametric tests.

1. There are two general types of statistical tests: parametric and non-parametric

2. The Z-Test of one sample mean is used to determine if an obtained sample mean or average of
scores or values is but a random sample from a population with a given or hypothesized or
expected population mean.

3. The T-Test for independent sample mean is used to determine if an observed difference
between the averages of two of independent groups is statistically significant.

4. The T-Test for dependent sample mean is used to determine if there is a significant difference
between two groups of correlated scores in terms of their means.

5. The one-way analysis of variance is used in order to determine if there are difference among
means of three or more groups.

6. The two-way analysis of variance, also called factorial analysis of variance, is employed in
order to determine the main and interaction effects of two independent factors.

7. When a significant value yielded by the analysis of variance test (also called F-value) occurs,
there is need to do a posteriori tests in order to determine which means are different.

8. The Pearson Product – Moment Correlation is employed when there are two sets of scores and
you would like to determine if the two sets are correlated.
 5

9. The Chi-Square Goodness-of-Fit test tells if an observed frequency distribution a variable

differs significantly from an expected or theoretical distribution of frequencies. The computation
calls for data on either the nominal or ordinal.

10. The Chi-Square test of association is used to determine whether or not two variables are
associated with each other.

VII. Study Guide

1. What statistical analysis should you use when we are conducting a research?

2. How do we compute for the Z-Test of One Sample Mean? Give an example.

3. How do you know what kind of test to use?