FACTOR ANALYSIS
OUTLINE
1. WHAT IS FACTOR ANALYSIS?
2. BASIC ASSUMPTION
3. HISTORY AND GOAL OF FACTOR ANALYSIS
4. OBJECTIVE,LIMITATION AND APPLICATION
5. STEPS
6. UNDERSTANDING OF FACTOR ANALSIS THROUGH DATA
2
Factor Analysis
• Factor analysis is a statistical procedure used
to identify a small number of factors that can
be used to represent relationships among sets
of interrelated variables
3
COMPUTER USE BY
TEACHERS is a
broad construct that
can have a number
of FACTORS
4
• Data reduction and summarization.
• Exploratory research
• Reducing the number of variables
• Explorative technique.
5
BASIC ASSUMPTION
• Underlying dimensions – or factors
– can be used to explain complex
events or trends.
6
HISTORY OF FACTOR ANALYSIS
• Psychologist Charles Spearman
• Mathematical Skill, Vocabulary,
Other Verbal Skills, Artistic Skills,
Logical Reasoning Ability
• Explained by one underlying
"factor" of general intelligence that
he called “g”
7
GOAL
• To identify otherwise not-directly-
observable factors on the basis of a
set of observable variables.
8
OBJECTIVE OF FACTOR ANALYSIS
• SIMPLIFYING THE DATA
• ANALYZING THE
INTERDEPENDENCE
9
BENEFITS OF FACTOR ANALYSIS
1.It brings out the hidden dimensions
2. find out relationships
3. when the data is large
4. simplified and condensed.
10
LIMITATION OF FACTOR ANALYSIS
• Complicated tool.
• Reliability of the results
• Suitability
11
FACTOR ANALYSIS IN MARKETING RESEARCH
• EXPLORATORY
• CONFIRMATORY
Factor analysis in market research
• Customer satisfaction studies.
13
FOUR STEPS:
1. Compute a correlation matrix for all
variables.
2. Determine the number of factors
necessary to represent the data and the method
of calculating them (factor extraction):.
3. Transform the factors to make them
interpretable (rotation)
4. Compute scores for each factor
14
STEP-I COMPUTE A CORRELATION MATRIX
FOR ALL VARIABLES.
BARTLETT’S TEST OF SPHERICITY -
used to test the hypothesis that the
correlation matrix is an identity matrix
• Level of SIGNIFICANCE has to be less
than .05
• All items are perfectly correlated with
themselves (one), and have some level of
correlation with the other items
15
STEP 1-COMPUTE A CORRELATION
MATRIX FOR ALL VARIABLES
KAISER-MEYER-OLKIN
• Measure of sampling adequacy
• Large KMO values are good
• KMO is below .5, don’t do a factor
analysis.
16
Step 2 Determine the number of factors necessary to
represent the data and the method of calculating
them (factor extraction):.
• The ‘Eigenvalue’ is the total variance
explained by each factor. Any ‘factor’
that has an Eigenvalue of less than
one does not have enough total
variance explained to represent a
unique factor, and is therefore
disregarded.
17
Screen Plot
18
Interpretation through Factor Analysis
19
Should Factor Analysis be Used
Variable Value Given Particular Rule/Action
KAISER-
MEYER-OLKIN
0.698 Large KMO values
are good because
correlations
between pairs of
variables
Rule: If the KMO is below .5, don’t do
a factor analysis
Action: As the Value is 0.698 i.e
above 0.5 hence Factor Analysis can
be done
Barlet's Test of
Sphericity
Look for
Significance Value
Value=0.000
is used to test the
hypothesis that
the correlation
matrix is an
identity matrix
Rule: Looking for SIGNIFICANCE (less
than .05) because if less than 0.05
the variables are correlated
. This value, 0.000 is less than 0.05,
and is an indication you can continue
with the Factor Analysis.
20
Calculate the Eigen Value
21
Total Variance Calculation and Communalities
22
23
You Tube Link
• https://youtu.be/FqTZrH0KRs0
24

Factor analysis

  • 1.
  • 2.
    OUTLINE 1. WHAT ISFACTOR ANALYSIS? 2. BASIC ASSUMPTION 3. HISTORY AND GOAL OF FACTOR ANALYSIS 4. OBJECTIVE,LIMITATION AND APPLICATION 5. STEPS 6. UNDERSTANDING OF FACTOR ANALSIS THROUGH DATA 2
  • 3.
    Factor Analysis • Factoranalysis is a statistical procedure used to identify a small number of factors that can be used to represent relationships among sets of interrelated variables 3
  • 4.
    COMPUTER USE BY TEACHERSis a broad construct that can have a number of FACTORS 4
  • 5.
    • Data reductionand summarization. • Exploratory research • Reducing the number of variables • Explorative technique. 5
  • 6.
    BASIC ASSUMPTION • Underlyingdimensions – or factors – can be used to explain complex events or trends. 6
  • 7.
    HISTORY OF FACTORANALYSIS • Psychologist Charles Spearman • Mathematical Skill, Vocabulary, Other Verbal Skills, Artistic Skills, Logical Reasoning Ability • Explained by one underlying "factor" of general intelligence that he called “g” 7
  • 8.
    GOAL • To identifyotherwise not-directly- observable factors on the basis of a set of observable variables. 8
  • 9.
    OBJECTIVE OF FACTORANALYSIS • SIMPLIFYING THE DATA • ANALYZING THE INTERDEPENDENCE 9
  • 10.
    BENEFITS OF FACTORANALYSIS 1.It brings out the hidden dimensions 2. find out relationships 3. when the data is large 4. simplified and condensed. 10
  • 11.
    LIMITATION OF FACTORANALYSIS • Complicated tool. • Reliability of the results • Suitability 11
  • 12.
    FACTOR ANALYSIS INMARKETING RESEARCH • EXPLORATORY • CONFIRMATORY
  • 13.
    Factor analysis inmarket research • Customer satisfaction studies. 13
  • 14.
    FOUR STEPS: 1. Computea correlation matrix for all variables. 2. Determine the number of factors necessary to represent the data and the method of calculating them (factor extraction):. 3. Transform the factors to make them interpretable (rotation) 4. Compute scores for each factor 14
  • 15.
    STEP-I COMPUTE ACORRELATION MATRIX FOR ALL VARIABLES. BARTLETT’S TEST OF SPHERICITY - used to test the hypothesis that the correlation matrix is an identity matrix • Level of SIGNIFICANCE has to be less than .05 • All items are perfectly correlated with themselves (one), and have some level of correlation with the other items 15
  • 16.
    STEP 1-COMPUTE ACORRELATION MATRIX FOR ALL VARIABLES KAISER-MEYER-OLKIN • Measure of sampling adequacy • Large KMO values are good • KMO is below .5, don’t do a factor analysis. 16
  • 17.
    Step 2 Determinethe number of factors necessary to represent the data and the method of calculating them (factor extraction):. • The ‘Eigenvalue’ is the total variance explained by each factor. Any ‘factor’ that has an Eigenvalue of less than one does not have enough total variance explained to represent a unique factor, and is therefore disregarded. 17
  • 18.
  • 19.
  • 20.
    Should Factor Analysisbe Used Variable Value Given Particular Rule/Action KAISER- MEYER-OLKIN 0.698 Large KMO values are good because correlations between pairs of variables Rule: If the KMO is below .5, don’t do a factor analysis Action: As the Value is 0.698 i.e above 0.5 hence Factor Analysis can be done Barlet's Test of Sphericity Look for Significance Value Value=0.000 is used to test the hypothesis that the correlation matrix is an identity matrix Rule: Looking for SIGNIFICANCE (less than .05) because if less than 0.05 the variables are correlated . This value, 0.000 is less than 0.05, and is an indication you can continue with the Factor Analysis. 20
  • 21.
  • 22.
    Total Variance Calculationand Communalities 22
  • 23.
  • 24.
    You Tube Link •https://youtu.be/FqTZrH0KRs0 24