College of Education and Behavioral

Studies

Department Of EDPM

The Basic Assumptions of Parametric

Tests

TM

Daher M. (GSR/9233/2016) December, 2023

 Presentation Outline

Concept Of Parametric Test Statistical Analysis of Parametric Testsof Parametric Tests

Models
 Concept of Parametric Test

Background Commonly Used Parametric Tests

1. Pearson Product Correlation Coefficient
A parameter can be termed as a value that o Tells us how well 2 continuous variables from the same
provides information about the population subject correlate to each other
Parametric Test 2. Student t-Test
 A test whose model specifies certain o The most widely used parametric test & Uses Mean, SD
conditions about the parameters of the o Whether a difference occurs between the means of 2 similar
population from which the research data sets
sample was drawn o The critical value for the Student t-test is 1.96 for an alpha of
 The tests depend heavily upon the 0.05,
parameters obtained and estimated 3. The z-Test
from the data o The variance of the standard population is used to obtain the
 A parametric statistical test is one z-test statistic. The p-value is less than 0.05 and statistical
that makes assumptions about the significance is achieved
parameters 4. ANOVA
 Assumptions of Parametric Tests

Normality
 Data Have A Normal Distribution (Or At Least Is Symmetric)

Homogeneity of variances

 Data From Multiple Groups Have The Same Variance

Linearity

 Data Have A Linear Relationship

Independence
 Data Are Independent

Collinearity
 Refers to a situation where two or more predictor variables are closely related to
one another.
 Linearity of Normal Distribution

For example, t-tests, F-tests, and regression analyses all require in some sense that
the ne numeric variables are approximately normally distributed.
 Parametric Procedure Analysis

Analysis Type Example Parametric

Procedure
Compare means between two Is the mean annual temperature of extreme southern Two-sample t-test
distinct/independent groups Jimma different from the mean annual temperature of
extreme northern Jimma?
Compare two quantitative Was there a significant change in soil fertility between a Paired t-test
measurements taken from the soil which inorganic fertilizers was applied and the same
same individual soil which group which organic manure was applied after
one
year?
Compare means between three If our experiment had three rock types (e.g., igneous, Analysis of
or more distinct/independent sedimentary and metamorphic), we might want to know variance (ANOVA)
groups whether the mean mineral content at baseline differed
among the
three groups?
Estimate the degree of association Is excessive deforestation associated with soil erosion? Pearson coefficient
between two quantitative variables of
Correlation
 Difference Between Parametric And Non-
parametric
PARAMETRIC TESTS NON-PARAMETRIC TESTS
 They do not make numerous
 They make numerous or stringent assumptions about
Assumptions about Parameters
Parameters

 Difference, and equivalent non-parametric test Data are changed from scores to ranks or signs
focuses on the difference between medians.

 Must have the same Population Variance  Variable under study has underlying continuity

 Focuses on the mean Parametric  Focuses On order/Ranking

 Parametric statistical procedures rely on assumptions  Nonparametric statistical procedures rely on no or

about the shape of the distribution (i.e., assume a few assumptions
normal distribution) (i.e., means and standard
deviations) of the assumed distribution.
 Advantages and Limitations of Parametric Tests

Applications/Limitations Advantage
 Parametric tests are often used in a variety of different
 Parametric tests are preferred because non-
applications, including medical research, market research, and
parametric tests tend to be less sensitive at
social sciences.
detecting differences between samples or an
This test is mainly useful;
effect of the independent variable on the
 This test is used when the given data is quantitative and
dependent variable.
continuous.
 The power efficiency of the nonparametric test
 When the data is of normal distribution.
is lower than its parametric counterpart.
 The parametric tests are helpful when the data is estimated
 A larger sample size is required for the non-
on the approximate ratio or interval scales of measurement.
parametric test than the parametric test
(Robson, 1994).
 For example, when comparing the means of
two independent samples, the variances of the
two distributions should be approximately
 Statistical Analysis of Parametric Tests
1. T-test (Test of Significance)

 ‘t’ test or test of significance of the

There are many types of t test :
difference between means for large  The one-sample t-test, used to compare the mean of a
independent samples is used to
population with a theoretical value.
compare the means between any two  The unpaired two sample t-test, used to compare the
groups on any of the variables.
mean of two independent samples.
 If the ‘t’ value is below a cut-off point
 The paired t-test, used to compare the means between
(depending on the degrees of
two related groups of samples.
freedom), the differences in means is 
considered not significant, and the
null hypothesis is accepted.
 When the ‘t’ value exceeds a cut-off
point, the difference is said to be
significant and the null hypothesis is
 Statistical Analysis of Parametric Tests

2. ANCOVA (Analysis of Co-efficient of Variation) (Correlated Variance

 Analysis of co-variance is useful for experimental psychologists where for variance reasons it is impossible or difficult to

equate experimental and control groups at the start a situation which often obtains in actual situations or in experiments.

 Defined as the function of two correlated factors and their analysis into corresponding parts”.

 Practically analysis of co-variance is technique to adjust the initial scores to final scores, so that net effect can be

analyzed.

 The analysis of variance technique is to analyse and test the significance difference among final scores or initial scores.
 Statistical Analysis of Parametric Tests
3. F-ratio (ANOVA-Single Factor)

 Useful technique, for testing the difference The F-test is an effective way to determine whether the
between the means of multiple independent
means of more than two samples are too different to
samples.
attribute to sampling error. It contains of following
 Test the differences among the means of the
operations.
samples by examining the amount of
 The sum of scores and the sum of squares of
variation between the samples relative to
the scores are obtained.
the amount of variation between the
 The variances of the score of one composite
samples.

 If the ‘F’ value worked out is equal or to one composite group are known as the

exceeds the ‘F’ limit value (from tables) it total group variance
 Statistical Analysis of Parametric Tests
4. Multivariate Analysis of Variance (MANOVA)

 Is an extension of common analysis of Assumptions and Limitations

variance (ANOVA).
The following assumptions are made when using a
 The number of response variables is
MANOVA.
increased to two or more.
 The response variables are continuous.
 Has one or more factors (each with two or
 The residuals follow the multivariate-normal
more levels) and two or more dependent
probability distribution with means equal to zero.
variables.
 The variance-covariance matrices of each group

of residuals are equal.

 The individuals are independent.

 Statistical Analysis of Parametric Tests

5. Regression Test
) variable
 Predict or explain the variation in one
KEY TAKE AWAYS
based on another variable.
 A regression is a statistical technique that relates a
 Determine the strength and character of the
dependent variable to one or more independent
relationship between one dependent variable
(explanatory) variables.
(usually denoted by Y) and a series of other
 A regression model is able to show whether changes
variables (known as independent variables).
observed in the dependent variable are associated
 Linear regression is the most common form of
with changes in one or more of the explanatory
this technique.
variables.
 Linear regression establishes the
 It does this by essentially fitting a best-fit line and
linear relationship between two variables based
seeing how the data is dispersed around this line.
Regression Analysis Types

 The most extensively used modelling technique is linear

regression, which assumes a linear connection between a
dependent variable (Y) and an independent variable (X).
 The linear regression model can be simple (with only one
dependent and one independent variable) or complex (with
numerous dependent and independent variables) (with one
dependent variable and more than one independent variable).
 Statistical Analysis of Parametric Tests

6. Correlation Test
)
 Measures two variables and assesses the  The first is that they do not believe that the statistical

statistical relationship (i.e., the correlation) relationship is a causal one or are not interested in causal
between them with little or no effort to relationships.
control extraneous variables.  Used to describe the strength and direction of the relationship
 There are many reasons that researchers
between two variables
interested in statistical relationships  The strength of a correlation between quantitative variables is typically

between variables would choose to conduct measured using a statistic called Pearson’s Correlation Coefficient (or
a correlational study rather than an Pearson’s r).
experiment.  Pearson’s r is a good measure only for linear relationships, in

which the points are best approximated by a straight line. It is

 Statistical Analysis of Parametric Tests

6.1. Types of Correlation Test

)
Three types of correlational research have been Correlational Research Has Three Main Characteristics.

identified:  Non-experimental: Correlational study is non-

1. Positive correlation: Is when an experimental. It means that researchers need not

increase in one variable leads to a rise in the other manipulate variables with a scientific methodology to

variable. A decrease in one variable will see a either agree or disagree with a hypothesis.

reduction in the other variable.  Backward-looking: Correlational research only looks

2. Negative correlation: If there is an back at historical data and observes events in the

increase in one variable, the second variable will past.

show a decrease and vice versa.  Dynamic: The patterns between two variables from

3. No correlation: There is no correlation correlational research are never constant and are

between the two variables in this third type. A always changing

 Models of Parametric Tests

1 Importance of COST MODEL SCOPE

DETERMINATION

The first step in developing a parametric model is to establish its scope.

This includes defining the end use of the model, the physical characteristics
of the model, the cost basis of the model, and its critical components and
cost drivers.
 COST MODEL

1 DATA COLLECTION
 Both cost and scope information must be identified

DATA NORMALIZATION and collected

2
• Parametric model is to normalize the data
before the data analysis stage. 3 DATA APPLICATION
 involves establishing the user interface and presenta
form for the parametric cost model.
 Mathematical, statistical (the various inputs to the cost mode
identified, and an interface)
DATA ANALYSIS 4
• There are many diverse methods and
5 TESTING
techniques that can be used in data
analysis • Is to test its accuracy and validity
5 DOCUMENTATION
• Must be documented thoroughly