Chapter 3

Exploratory Factor Analysis

3-1
 Correlation Matrix for Store Image Elements

V1 V2 V3 V4 V5 V6 V7 V8 V9
V1 Price Level 1.00
V2 Store Personnel .427 1.00
V3 Return Policy .302 .771 1.00
V4 Product Availability .470 .497 .427 1.00
V5 Product Quality .765 .406 .307 .472 1.00
V6 Assortment Depth .281 .445 .423 .713 .325 1.00
V7 Assortment Width .354 .490 .471 .719 .378 .724 1.00
V8 In-Store Service .242 .719 .733 .428 .240 .311 .435 1.00
V9 Store Atmosphere .372 .737 .774 .479 .326 .429 .466 .710 1.00

3-2
 Correlation Matrix of Variables After
Grouping Using Factor Analysis

V3 V8 V9 V2 V6 V7 V4 V1 V5
V3 Return Policy 1.00
V8 In-store Service .733 1.00
V9 Store Atmosphere .774 .710 1.00
V2 Store Personnel .741 .719 .787 1.00
V6 Assortment Depth .423 .311 .429 .445 1.00
V7 Assortment Width .471 .435 .468 .490 .724 1.00
V4 Product Availability .427 .428 .479 .497 .713 .719 1.00
V1 Price Level .302 .242 .372 .427 .281 .354 .470 1. 00
V5 Product Quality .307 .240 .326 .406 .325 .378 .472 .765 1.00

Shaded areas represent variables likely to be grouped together by factor analysis.

3-3
 Exploratory Factor Analysis
Defined

Exploratory factor analysis . . . is an

interdependence technique whose primary
purpose is to define the underlying structure
among the variables in the analysis.

3-4
 What is Exploratory Factor Analysis?

Exploratory Factor Analysis . . .

• Examines the interrelationships among a large number
of variables and then attempts to explain them in terms
of their common underlying dimensions.
• These common underlying dimensions are referred to
as factors.
• A summarization and data reduction technique that
does not have independent and dependent variables,
but is an interdependence technique in which all
variables are considered simultaneously.

3-5
 Disciples that use Factor Analysis
• A variety
• Psychology
• HRM
• Marketing
• Finance
• Economics
 Factor Analysis Decision Process

Stage 1: Objectives of Factor Analysis

Stage 2: Designing a Factor Analysis
Stage 3: Assumptions in Factor Analysis
Stage 4: Deriving Factors and Assessing Overall Fit
Stage 5: Interpreting the Factors
Stage 6: Validation of Factor Analysis
Stage 7: Additional uses of Factor Analysis Results

3-7
 Application of Factor Analysis
to a Fast-Food Restaurant:
Data Summarization and Data Reduction
Variables Factors
Waiting Time

Cleanliness Service Quality

Friendly Employees

Taste

Temperature Food Quality

Freshness

3-8
Factor Analysis has three primary objectives

• Identification of the structure of relationships among either

variables or respondents.

• Identification of representative variables from a much

larger set of variables for use in subsequent multivariate
analyses.

• Creation of an entirely new set of variables, which are much

smaller in number in order to partially or completely replace
the original set of variables for inclusion in subsequent
multivariate techniques.
 R Type Factor Analysis Vs. Q type Factor
Analysis
• There are two approaches to calculate the correlation
matrix that determine the type of factor analysis
performed:
– R-type factor analysis: input data matrix is computed from
correlations between variables.

– Q-type factor analysis: input data matrix is computed from

correlations between individual respondents. Similar to cluster
analysis. But there is a difference- Cluster analysis defines
similarities in terms of euclidean distance, whereas factor
analysis defines similarities based on similar covariance
structure pg 101 fig 3.3
Exploratory Vs. Confirmatory Factor Analysis

1. Exploratory Factor Analysis (EFA) = is used to

discover the factor structure of a construct
and examine its reliability. It is data driven.

2. Confirmatory Factor Analysis (CFA) = is used

to confirm the fit of the hypothesized factor
structure to the observed (sample) data. It is
theory driven.
 STAGE 2: DESIGNING A FACTOR ANALYSIS

Variables in factor analysis are generally metric. Dummy variables may

be used in special circumstances with a special analysis (Try to avoid
this)

The researcher should minimize the number of variables included in

the analysis, but include a sufficient number of variables to represent
each proposed factor (i.e. five or more).

Sample size is an important consideration in factor analyses.

– The sample size should be 100 or larger. Sample sizes between 50 and 100
may be analyzed but with extreme caution.
– The ratio of observations to variables should be at least 20 to 1 in order to
provide the most stable results.
 STAGE 3: ASSESSING ASSUMPTIONS
Conceptual rather than Statistical

Conceptual : Statistical :
1. Underlining structure 1. Multi co linearity
1. Correlation matrix >.3
2. Partial Correlation Matrix
3. Bartlett test of sphericity -
p< .05
2. Homogeneous sample 4. KMO/ MSA > .5

2. Statistical assumptions like normality

is only desired to the extent that
other statistically techniques will be
performed after the factor analysis
 STAGE 4: DERIVING FACTORS AND
ASSESSING OVERALL FIT
• Step 1: Decide on the method/model to extract the
factors: Principal Component Analysis or Common
Factor Analysis

• Step 2: Decide on the number of factors

 Stage 4- Decision 1: Extraction Method
Types of Variance Carried into the Factor Matrix

Diagonal Value Variance

Unity (1)
Total Variance

Communality Common Specific and Error

Variance extracted

Variance not used

3-15
Stage 4- decision 2: Decide on the number of factors

• When a large set of variables are factored, the

method first extract the combination of variables
explaining the greatest amount of variance and
then proceeds to combination that accounts for
smaller and smaller amounts of variance.:

• The stopping criteria for the number of factors to extract :

» Latent root criterion
» A priori criterion
» % of variance extracted
» Scree Test criterion
Latent Root Criterion Priori Criterion
• % of Variance explained
Total Variance Explained

Initial Eigenvalues Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings
% of Cumulative % of Cumulative % of Cumulative
Component Total Variance % Total Variance % Total Variance %
1 3.949 49.358 49.358 3.949 49.358 49.358 3.461 43.268 43.268
2 1.376 17.201 66.559 1.376 17.201 66.559 1.863 23.291 66.559
3 .716 8.948 75.507
4 .615 7.687 83.194
5 .430 5.369 88.563
6 .398 4.980 93.543
7 .305 3.816 97.359
8 .211 2.641 100.000
Extraction Method: Principal Component Analysis.
• Scree Test 5
Scree Plot

2
Eigenvalue

0
1 2 3 4 5 6 7 8

Component Number
STAGE 5: INTERPRETING THE FACTORS

• Investigate Factor Loading: Rotated and Un

rotated
• Comment on practical and statistical
significance of factor loading
• Interpret the factor Matrix
• Re specification of the matrix should be
considered
 Stage 5- decision 1: Investigate Factor Loading

• Investigate the unrotated factor loadings

Component Matrixa

Component
1 2
AC1 -.537 .778
AC2 -.586 .734
GT1 .798 .223
GT2 .666 .171
US1 .762 .113
US2 .768 .268
GV1 .606 .195
GV2 .835 .174
Extraction Method: Principal Component Analysis.
a. 2 components extracted.

• Rotate factor Solution

1. Orthogonal = axes are maintained at 90 degrees.

2. Oblique = axes are not maintained at 90 degrees.
 Oblique Factor Rotation
Unrotated
Factor II Orthogonal Rotation:
+1.0 Factor II

V1 Oblique Rotation:
Factor II
+.50 V2

Unrotated
Factor I
-1.0 -.50 0 +.50 +1.0
V3
V4 Oblique
-.50 Rotation:
V5 Factor I
Orthogonal
Rotation: Factor I
-1.0
3-22
 Choosing Factor Rotation Methods
• Orthogonal rotation methods . . .
o are the most widely used rotational methods.
o are The preferred method when the research goal is
data reduction to either a smaller number of variables
or a set of uncorrelated measures for subsequent use
in other multivariate techniques.
• Oblique rotation methods . . .
o best suited to the goal of obtaining several
theoretically meaningful factors or constructs
because, realistically, very few constructs in the “real
world” are uncorrelated.
3-23
Orthogonal Rotation Methods

• Quartimax (simplify rows)

• Varimax (simplify columns)

• Equimax (combination)

3-24
 Stage 5- decision 2: Investigate Practical and
Statistical significance

• Magnitude for practical significance: Factor

loadings can be classified based on their
magnitude:
– Greater than + .30 -- minimum consideration
level.
– + .40 -- more important
– + .50 -- practically significant (the factor accounts
for 25% of variance in the variable).
 Stage 5- decision 2: Investigate Practical and
Statistical significance

• Power and statistical significance: Given the

sample size, the researcher may determine
the level of factor loadings necessary to be
significant at a predetermined level of power.
For example, in a sample of 100 at an 80%
power level, factor loadings of .55 and above
are significant.
 Guidelines for Identifying Significant
Factor Loadings Based on Sample Size
Factor Loading Sample Size Needed
for Significance*

.30 350
.35 250
.40 200
.45 150
.50 120
.55 100
.60 85
.65 70
.70 60
.75 50

*
Significance is based on a .05 significance level (a), a power level of 80 percent, and standard errors
assumed to be twice those of conventional correlation coefficients.

3-27
 Stage 5- decision 3: Interpreting the Factor Matrix

·An optimal structure exists when all variables have high loadings only on a single
factor.

·Variables that cross-load (load highly on two or more factors) are usually
deleted unless theoretically justified or the objective is strictly data reduction.

·Variables should generally have communalities of greater than .50 to be retained

in the analysis.

·Naming the factor is based on an interpretation of the factor loadings.

– Significant loadings: The variables that most significantly load on each factor should be
used in naming the factors. The variables' magnitude and strength provide meaning to
the factors. Subjected to the researcher preference
 Stage 5- decision 4: Re specify the Model

Respecification should always be considered.

Some methods are:
• Deletion of a variable(s) from the analysis
• Employing a different rotational method for
interpretation
• Extraction of a different number of factors
•Employing a different extraction method
 STAGE 6: VALIDATING THE RESULTS

Validation assesses
1) the degree of generalizability of the findings
2) the degree to which the results are influenced by individual cases.

Results should be replicable. Confirmatory factor analysis is the most

commonly used replication technique. Analyses can be run using a split
sample or another new data set.

The factor structure should be stable across additional analyses. Stability is

highly dependent on sample size and the number of observations per
variable.

The impact of outliers should be determined by running the factor model with
and without the influential observations.
STAGE 7: ADDITIONAL USES OF THE FACTOR ANALYSIS RESULTS

• Beyond the interpretation and understanding of

the relationship among the variables, the
researcher may wish to use the factor analysis
results in subsequent analysis. Factor analysis
may be used to reduce the data for further use by
either
(1) the selection of a surrogate variable
(2) creation of a new variable with a summated scale
(3) replacement of the factor with a factor score.
 The single surrogate variable:
• A surrogate variable that is representative of
the factor may be selected as the variable
with the highest loading.
 Advantages: simple to administer and interpret.
 Disadvantages:
1) does not represent all “facets” of a factor
2) prone to measurement error.
 Factor scores:

• A factor Score is computed using all variables

loading on a factor, may also be used as a
composite replacement for the original variable.
 Advantages:
1) represents all variables loading on the factor,
2) best method for complete data reduction.
3) Are by default orthogonal and can avoid complications caused
by multicollinearity.
 Disadvantages:
1) interpretation more difficult since all variables contribute
through loadings
2) Difficult to replicate across studies.
 Summated scales:
• All the variables loading highly on a factor may be
combined (the sum or the average) to form a
replacement variable
 Advantages:
1) compromise between the surrogate variable and factor score options.
2) reduces measurement error.
3) represents multiple facets of a concept.
4) easily replicated across studies.
 Disadvantages:
1) includes only the variables that load highly on the factor and excludes
those having little or marginal impact.
2) not necessarily orthogonal.
3) Require extensive analysis of reliability and validity issues.
 Basic Issues of Scale Construction:
A conceptual definition is the starting point for creating a scale. The scale must appropriately
measure what it purports to measure to assure content or face validity.
A scale must be unidimensional, meaning that all items are strongly associated with each other
and represent a single concept.

 Reliability of the scale is essential. Reliability is the degree of consistency between multiple
measurements of a variable.
Test-retest reliability is one form of reliability.
 Another form of reliability is the internal consistency of the items in a scale. Measures of internal consistency
include item-to-total correlation, inter-item correlation, and the reliability coefficient.

Once content or face validity, unidimensionality, and reliability are established other forms of
scale validity should be assessed.

Discriminant validity is the extent that two measures of similar but different concepts are distinct.

Nomological validity refers to the degree that the scale makes accurate predictions of other
concepts.