Internal Quality Assurance Cell (IQAC) &

Shasun Knowledge Centre (SKC)

Organise
Two days Faculty Development Program on
STATISTICAL APPLICATION USING SPSS
Resource Person
Dr.R.Ravanan
Joint Director of Collegiate Education, Chennai Region,
Former Principal, Presidency College, Chennai,
Former Head, Department of Statistics, Presidency College,
Secretary, Indian Society for Probability and Statistics(ISPS), India,
General Secretary, Tamil Nadu Statistical Association (TNSA), Tamil Nadu
16.05.2020 & 17.05.2020
Study Materials
 Resource Person

Dr. R. RAVANAN
Joint Director of Collegiate Education, Chennai Region
Former Principal, Presidency College, Chennai
Former Head, Dept of Statistics, Presidency College

E-mail: [email protected]
Mobile: 98403 75672 / 94442 21627
 What is SPSS?
Statistical Package for Social Science
General Purpose Statistical Software
Consists of three components
Data Window - data entry and database
(.sav)
Output Window - all output from any SPSS
session (.lst)
Syntax Window - commands lines (.sps)
Data Entry & Preparation
Data entry
New or Recalled (SPSS or non-SPSS)

Data Definition

Data Manipulation and Variable

Development
 Data Definition
Purpose:
Give meanings to the numbers for ease of
reading the output
Involves
Data Format
Variable Name
Value Labels
Missing Values
Command: Data  Data Definition
 Data Manipulation
Recoding
To give new values to old values (especially
reversing negatively worded questions)
To form nominal variable from continuous
data
Variable Development
To form new variables combinations of old
ones or functions of old ones
Command: Transform  Recode/ Compute
Data Analysis - Descriptive
Purpose:
To describe each variable - What is the
current level of the variable of interest?
Command
Frequency
Means, Minimum, Maximum, Standard
Deviation, Quartiles, Standard
Deviation
Analyze  Frequencies /Descriptives
Data Analysis - Descriptive
Frequencies for two or more nominal
variables
Analyze  Summarize  Crosstabulation

Means of variables by subgroups defined by

one or more nominal variables
Analyze  Compare Means  Means (Use
of Levels)
 Parametric
Test of Differences
When
dependent continuous variable and we
want to test differences across groups
Command
Analyze  Compare Means 
Independent t-test/ Paired t-test/ one-
way ANOVA
 Non-Parametric Test of
Differences
When
dependent variable ordinal or normal
assumption not met
Command
Analyze  Non-parametric  2
Independent/ 2 related samples/ k
independent samples/ k related
samples
Parametric Two-Way ANOVA
When
continuous dependent variable and
related groups
Command
Analyze  General Linear Model 
Simple
Note: Fixed Factor Effect
 Bivariate Relationship
When
Covariation between two variables
Correlation:
When both are continuous or ordinal
Command
Analyze  Correlate  Bivariate (with
option for Spearman if both ordinal)
 Reliability Analysis
When
Before forming composite index to a variable
from a number of items
Command
Analyze  Scale  Reliability Analysis (with
option for Descriptives item, scale, scale if
item deleted)
Interpretation
alpha value greater than 0.7 is good; more
than 0.5 is acceptable; delete some items if
necessary
 Measures of Reliability
Internal Consistency: (of items in a scale):

1. Average inter-item correlation If average inter-item

correlation > 0.7, then standardize items and add them
together as an index.
2. Cronbach's alpha , which measures " internal consistency
of items in a scale" Garson ,G.D.(1999) and is
 Exercise 1

Test whether opinion regarding

Comfortable with Physical Work
Environment are above average level
 t test for Single Mean

1. It is a parametric test

2. It is Univariate Analysis

3. Data should be ordinal or scale

 t test for Single Mean

Null Hypothesis: Opinion regarding Comfortable with Physical

Work Environment are equal to average level

Alternate Hypothesis: Opinion regarding Physical Work

Environment of are not equal to average
level

Test Statistic: t test for single mean is

which follows Student’s t distribution with n-1 degrees of freedom

 Exercise: 2

Test whether significant difference between

Gender with regard to Factors of Job
Description of Employees
 t -TEST FOR DIFFERENCE OF TWO MEANS
or
INDEPENDENT SAMPLE t test

1. It is a Parametric test

2. It is a Bivariate Analysis

3. It is used for comparing two group means

Null Hypothesis: There is no significant difference between male and
female with regard to Factors of Job Description of Employees

Alternate Hypothesis: There is significant difference between male and female

with regard to Factors of Job Description of Employees

Test Statistic: t test for difference of two means is

which follows Student’s t distribution with n1 + n2 -2 degrees of freedom (df)

 Exercise: 3

Test whether significant difference among

Age Group with regard to Factors of Job
Description of Employees
 One way ANOVA
followed by
Duncan Multiple Range Test (DMRT)
1. It is Parametric test

2. It is Bivariate Analysis

3. It is used for comparing More than Two groups means

4. After applying ANOVA, then use Post Hoc test

Null Hypothesis: There is no significant difference among
age group with respect to Factors of Job
Description of Employees

Alternate Hypothesis: There is significant difference

among age group with respect to Factors
of Job Description of Employees

Procedure for One way ANOVA:

1. Find Correction Factor (CF)
2. Find Total Sum of Square (TSS)
3. Find Between Sum of Square (BSS)
4. Find Error Sum of Square (ESS)
5. Form the ANOVA table
 Exercise 6

Test whether level of Job Description of Employees are

equal to moderate level
 CHI SQUARE TEST FOR GOODNESS OF FIT

1. It is Non-Parametric test

2. It is Univariate Analysis

3. Data should be Nominal

 CHI SQUARE TEST FOR GOODNESS OF FIT

Null Hypothesis: Levels of Job Description of Employees are

equally distributed .

Alternate Hypothesis: Levels of Job Description of Employees not

equally distributed

Test Statistic: Chi-square test for goodness of fit is

 Exercise 7

Test whether association between

Gender and Level of Job Description of
Employees
 CHI SQUARE TEST FOR INDEPENDENCE OF
ATTRIBUTES

1. It is Non-Parametric Test

2. It is Bi-variate analysis

3. Both variables are Nominal

Solution:

1. Null Hypothesis: There is no association between Gender and

Level of Job Description of Employees

2. Alternate Hypothesis: There is association between Gender and

Level of Job Description of Employees

3. Test Statistic: Chi-square test for independence of attributes is

Solution:

1. Null Hypothesis: There is no association between sex and levels

of role description.

2. Alternate Hypothesis: There is association between sex and

levels of role description

3. Test Statistic: Chi-square test for independence of attributes is

Solution:

1. Null Hypothesis: There is no association between sex and levels

of role description.

2. Alternate Hypothesis: There is association between sex and

levels of role description

3. Test Statistic: Chi-square test for independence of attributes is

Solution:

1. Null Hypothesis: There is no association between sex and levels

of role description.

2. Alternate Hypothesis: There is association between sex and

levels of role description

3. Test Statistic: Chi-square test for independence of attributes is

 Exercise 8

Test whether significant relationship

between Factors of Job Description of
Employees
 TEST FOR SIGNIFICANCE OF CORRELATION
COEFFICIENT

1. It is Parametric test

2. It is based on bivariate analysis

3. Both variables are scale

 Coefficient of Correlation ( -1 ≤ r ≤ +1)

1. Simple correlation (r) : Two variables

2. Multiple Correlation (R) : More than Two variables

Types of Correlation:

1. Positive or Direct Correlation ( 0 < r < +1)

2. Negative or Inverse Correlation ( -1< r < 0)

3. Perfect Correlation (r = ± 1)

4. Uncorrelation or No Correlation ( r = 0 )
Solution:
First find the coefficient of correlation by using the formula

1. Null Hypothesis: There is no relationship between Factors of Job

Description of Employees
2. Alternate Hypothesis: There is relationship between Factors of Job
Description of Employees
3. Test Statistic: t test for coefficient of correlation is
 Non-Parametric Test
One sample test:
– Binomial Test
– Chi-Square test for goodness of fit
– Kolmogorov-Smirnov one sample test

Two Independent sample:

– Fisher Exact test
– Chi-Square test for intendance of attributes
– Median test
– Mann-Whitney U test
– Kolmogorov-Smirnov Two sample test
 Non-Parametric Test
Two dependent sample
– McNemar test
– Sign test
– Wilcoxon Matched-Pairs signed rank test
– Walsh test

More than two independent samples

– Krushkal_Wallis one-way analysis
– Chi-square test for k impendent sample
– Extention of Median test

More than two dependent samples

– Friedman Two way analysis
– Cochran Q test
 Exercise : 4

Test whether significant difference between

Nature of Job with regard to Factors of Job
Description of Employees
 Mann-Whitney U test
1. It is Non-Parametric test
2. It is equal to Independent sample t test in parametric test
3. It is used for comparing two groups mean rank

Null Hypothesis: There is no significant difference between

mean rank of Full time and Part time job with
regard to Factors of Job Description of Employees
Alternate Hypothesis: There is significant difference between
mean rank of Full time and Part time job with
regard to Factors of Job Description of
Employees
 Conditions to apply Non-Parametric test
1. Data not follows Normal Distribution
2. If the given data is ranking data, then apply Non-Parametric test
3. If SD is more than Mean, then apply Non-Parametric test

TEST FOR NORMALITY

Analysis, Non-parametric test, Legacy dialog, One sample K-S test

Null Hypothesis: The data follows Normal Distribution is good

1. If P value is greater than 0.05, then the data follows Normal
Distribution
2. If P value is less than or equal to 0.05, then the data not follows
Normal Distribution
Mann-Whitney U test is

which follows Standard Normal Distribution

where
 Exercise: 5

Test whether significant difference among

Experience in year with regard to Factors of
Job Description of Employees
 Krushkal-Wallis Test
1. It is Non-Parametric test
2. It is equal to One Way ANOVA in parametric test
3. It is used for comparing more than two groups mean rank

Null Hypothesis: There is no significant difference among

mean rank of Experience in year with regard to
Factors of Job Description of Employees

Alternate Hypothesis: There is significant difference among

mean rank of Experience in year with
regard to Factors of Job Description of
Employees
Krushkal - Wallis test is

which follows Chi-square distribution with k-1 degrees

of freedom

where R = Sum of rank of each group

N = Total number of observations

n = Number of observation in each group

k = Number of groups
 Exercise: 9

Test whether significant difference among

mean rank towards Factors of Job
Description of Employees
 Friedman Test
1. It is Non-Parametric test
2. It is equal to ANOVA with repeated measures in parametric test
3. It is used for comparing more than two related groups of
mean rank
4. All the variables are equal weightage. If it is unequal weightage,
then converted into equal weightage by taking average
5. Any ranking data, Friedman test is most suitable test

Null Hypothesis: There is no significant difference among mean

rank towards Factors of Job Description of
Employees
Alternate Hypothesis: There is significant difference among mean
rank towards Factors of Job Description of
Employees
Friedman test is

which follows Chi-square distribution with k-1 degrees

of freedom

where R = Sum of rank of each items

N = Total number of observations

k = Number of items