BRM ASSIGNMENT
Factor Analysis using SPSS
N. VIDYA NEELIMA
21BSP0434
INTERPRETATION
CORRELATION MATRIX:
A correlation matrix is simply a rectangular array of numbers that gives the correlation
coefficients between a single variable and every other variable
With respect to the correlation matrix if any pair of variables has a value less than 0.5, consider
dropping one of them from the analysis. For this factor, analysis needs to be reperformed with
the exclusion of pair of variables with less than 0.5 value. The off-diagonal elements (The values
on the left and right sides of the diagonal in the table below) should all be very small (close to
zero) in a good model .
Correlation Matrix
Resp_
X1 X5 X6 X7 X8 X10 X11 X12 X13 X16 X18 X17 X15 X14 no
C X1 1.000 .102 .218 .316 -.121 -.100 .224 .158 -.119 .013 -.054 .140 .154 -.035 .141
or
X5 .102 1.000 .128 .212 .025 .028 .036 .027 .079 -.090 .161 .029 .480 .018 .067
re
la X6 .218 .128 1.000 .424 -.006 .081 .073 -.007 .163 .128 .015 .205 .291 .235 -.096
ti X7 .316 .212 .424 1.000 .116 .044 .163 .005 .078 -.024 .022 .062 .226 .308 -.045
o
X8 -.121 .025 -.006 .116 1.000 .007 -.020 -.013 .015 -.076 -.097 -.146 .103 -.094 -.098
n
X10 -.100 .028 .081 .044 .007 1.000 .007 .008 .423 .331 .392 .168 .151 .146 .144
X11 .224 .036 .073 .163 -.020 .007 1.000 -.019 .020 .156 .038 .175 .062 .081 .172
X12 .158 .027 -.007 .005 -.013 .008 -.019 1.000 .016 .001 .280 .032 .222 .008 -.109
X13 -.119 .079 .163 .078 .015 .423 .020 .016 1.000 .500 .445 .208 .219 .386 .171
X16 .013 -.090 .128 -.024 -.076 .331 .156 .001 .500 1.000 .382 .290 .050 .283 .144
X18 -.054 .161 .015 .022 -.097 .392 .038 .280 .445 .382 1.000 .214 .314 .263 .013
X17 .140 .029 .205 .062 -.146 .168 .175 .032 .208 .290 .214 1.000 .095 .381 -.061
X15 .154 .480 .291 .226 .103 .151 .062 .222 .219 .050 .314 .095 1.000 .185 -.052
X14 -.035 .018 .235 .308 -.094 .146 .081 .008 .386 .283 .263 .381 .185 1.000 -.095
Resp_no .141 .067 -.096 -.045 -.098 .144 .172 -.109 .171 .144 .013 -.061 -.052 -.095 1.000
Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .646
Bartlett's Test of Sphericity Approx. Chi-Square 204.692
Df 105
Sig. .000
KMO
The KMO measures the sampling adequacy (which determines if the responses given with the
sample are adequate or not) which should be close to 0.5 for satisfactory factor analysis to
proceed. Kaiser recommend 0.5 (value for KMO) as a minimum (barely accepted), values
between 0.7-0.8 acceptable, and values above 0.9 are superb. Looking at the table above, the
KMO measure is 0.646, which is close to 0.7 and therefore can be accepted.
Bartlett's test is another indicator of the strength of the relationship between variables. This
experiment tests the null hypothesis that the correlation matrix is an identity matrix. All diagonal
members of an identity matrix are 1 and all off-diagonal elements (as specified above) are near to
0.
From the same table, we can see that Bartlett’s Test Of Sphericity is significant (0.105). That is,
the significance is less than 0.05. In fact, it is actually 0.0105, i.e. the significance level is small
enough to reject the null hypothesis. This means that the correlation matrix is not an identity
matrix.
Communalities
Initial Extraction
X1 1.000 .665
X5 1.000 .569
X6 1.000 .552
X7 1.000 .638
X8 1.000 .427
X10 1.000 .501
X11 1.000 .423
X12 1.000 .651
X13 1.000 .684
X16 1.000 .599
X18 1.000 .694
X17 1.000 .523
X15 1.000 .696
X14 1.000 .607
Resp_no 1.000 .732
Extraction Method: Principal Component Analysis.
COMMONALITIES:
The next item in the output is a table of commonalities, which shows how much of the variation is
shared by all of the variables (i.e. the communality value, which should be more than 0.5 to be
considered for further analysis). Otherwise, these variables would be removed from the following
factor analysis processes) have been accounted for by the recovered factors in the variables. From
the above table X11 and X18 are less than 0.5 so it can be from further analysis.
Total Variance Explained
Initial Eigenvalues Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings
Componen % of Cumulative % of Cumulative % of Cumulative
t Total Variance % Total Variance % Total Variance %
1 2.998 19.984 19.984 2.998 19.984 19.984 2.509 16.725 16.725
2 1.926 12.839 32.823 1.926 12.839 32.823 1.915 12.768 29.492
3 1.484 9.890 42.713 1.484 9.890 42.713 1.710 11.402 40.894
4 1.325 8.832 51.545 1.325 8.832 51.545 1.500 10.002 50.897
5 1.228 8.188 59.733 1.228 8.188 59.733 1.325 8.836 59.733
6 .976 6.508 66.241
7 .875 5.832 72.072
8 .736 4.906 76.978
9 .666 4.442 81.421
10 .646 4.305 85.726
11 .556 3.707 89.433
12 .468 3.122 92.555
13 .410 2.733 95.288
14 .392 2.611 97.899
15 .315 2.101 100.000
Extraction Method: Principal Component Analysis.
TOTAL VARIANCE:
The eigenvalue is the number of retrieved components whose sum should equal the number of
items subjected to factor analysis. All of the components that may be retrieved from the analysis,
as well as their eigenvalues, are listed in the following item.
The Eigenvalues table is divided into three sections:
At the start, the eigenvalues
Sums of Loadings Squared (Extracted)
Sums of Loadings Squared Rotation
For analysis and interpretation, we only care about Initial Eigenvalues and Extracted Sums of
Squared Loadings. For calculating the number of components or factors expressed by selected
variables, the presence of eigenvalues higher than one is essential. The first component has a value
of 2.998 > 1, the second component has a value of 1.926 > 1, the third component has a value of
1.484 > 1, and the fourth component has a value of 1.325 > 1, according to the table. As a
consequence, three components may be seen in the eight variables and twelve observations given
above. Furthermore, according to the extracted sum of squared holding percent of the variance,
the first component accounts for 19.984 percent of the variance characteristics from the given
data, while the second factor accounts for 12.839 percent., and the third factor accounts for 9.890
percent.
SCREE PLOT:
The scree plot is a graph of the eigenvalues against all the factors. The graph is useful for
determining how many factors to retain. The point of interest is where the curve starts to
flatten. It can be seen that the curve begins to flatten between factors 5 and 6. Note also that
factor 7 onwards have an eigenvalue of less than 1, so only six factors have been retained.
Component Matrix
Component
1 2 3 4 5
X13 .700 -.363 .232
X18 .652 -.276 -.312 .175 -.253
X14 .609 .138 -.464
X16 .601 -.424 .239
X10 .533 -.372 -.129 .121 .216
X15 .521 .423 -.462 .180
X17 .513 .337 -.238 -.298
X7 .368 .622 .123 -.191 .253
X1 .155 .560 .400 .331 -.239
X6 .439 .473 .140 -.308 .144
X5 .283 .441 -.384 .344 .169
X11 .230 .191 .499 .289
Resp_no -.171 .329 .687 .338
X12 .177 .130 -.346 .194 -.668
X8 .133 -.375 -.149 .490
Extraction Method: Principal Component Analysis.
a. 5 components extracted.
Component Matrix:
The larger the absolute value of the loading, the more the component contributes to
the variable. From the eighteen questions, we chose five variables and divided them
into five groups based on the most important items with comparable responses in
components 1, 2,3,4 and 5. The table's empty regions represent loadings of less than
0.5, making it easier to read. All loadings of less than 0.5 were turned off.
Component
1 2 3 4 5
X13 .801 .158 .110
X16 .706 .152 -.200 .187
X10 .693 .125
X18 .658 .222 .455
X6 .691 .230 .109
X14 .401 .650 -.102
X7 .644 .377 .197 -.184
X17 .304 .501 -.247 .173 .298
X5 .744 .109
X15 .186 .226 .743 .240
X1 -.263 .245 .178 .662 .257
Resp_no .287 -.400 .642 -.267
X11 .156 .628
X12 .234 .767
X8 .358 -.335 -.430
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
a. Rotation converged in 5 iterations.
Rotated Component Matrix:
If the value for one of the components is less than 0.5 or the stated limit (which might be 0.6
depending on the researcher's requirement to include the intended factor loading), that variable
might be worth investigating further. This variable, however, is inefficient in assessing a single
category since it represents two components when more than 0.5 (or 0.6) loading is present in
more than one component. As a result, it has to be eliminated. They can't be included for further
study since familiarity with the product and product quality measure more than one component. As
a result, there is a greater need for processing.
Component Transformation Matrix
Component 1 2 3 4 5
1 .757 .512 .287 .182 .220
2 -.614 .482 .568 .256 .038
3 -.070 .273 -.617 .712 -.181
4 .107 -.656 .379 .627 .144
5 .181 -.014 .265 .012 -.947
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.