# 4.3. the variables and the factors to which

4.3.  Factor
Analysis is used in order to skillfully meet the  objective
of  the  research,
EFA  has  been
used  to  reduce
the  number  of
variables to  a  minor
set  of  factors
(Hair  et  al.,
2006;  Field,  2013)
and  to  abbreviate the  issue
of multicollinearity
(Malhotra  and  Dash,
2011).  The motive is to find the degree
of dependency of these individual variables on the dependent variable.  The whole set of data has been accumulated in
SPSS  21. The execution  of
the  project commenced with  the
grouping  of  the
variables  using  factor
analysis,  and  then the
relationship  between  the
dependent  variable  and
the  predictors  has
been  fixed in a firm position
using  multiple regression  analysis.
The  output  in
the  form  of
regression  equation  has
been  analyzed  at
the  last  of  the
topic.  The  variables
recognized  in  the
research  and  the
data  sample  collected
on  these  variables
are  input to  EFA.
Through  EFA,  the
number  of  variables
is  reduced  to
a  fewer  set
of  factors  using
varimax rotation.  The  EFA
technique  is  widely
used  for  the
reduction  of the  problem
of  multicollinearity  among the
variables.  The  criteria
for  EFA  for
finalizing  the  factor
structure  are  as
follows  (Hair  et
al.,  2006): 1  Correlation
matrix  and  Anti-Image
Matrix  results  show
how  appropriate  the
data  is  for
the  EFA technique.  These
have  been  used
as  a  tool
to  filtrate  out
the  variables  not competant for  factor
analysis (Field,  2013). 2  Kieser-Meyer-Olkin  (KMO)
value  and  Bartlett’s
test  for  sphericity
are  used  as
measures  of  sampling adequacy.  To accomodate
the  criteria,  the
KMO  value  must
be  more  than
0.8  With  the
KMO  value  of
0.922, and  Bartlett’s  significance
value  of  0.000,
the  sample  collected
is  extremely competant  (Hair
et  al.,  2006; Field,
2013). 3  Further  the
the  amount  of  equivalence  between
the  variables  and
the  factors to  which
they  belong.  The
can  vary  from
–1  to  +1
and  square  of  any  of
these numbers  defines  the
amount  of  variability
accounted  for  this
factor.  Lower  factor
stowing values,  i.e. less  than
0.5  and  lower
communality  variables  are
filtered  from  the
data  set  (Hair
et  al.,  2006;
Field, 2013). 4  The  Eigen
value  amounts  the
total  variance  explained
by  the  factors.
It  is  also
a  kind  of clarifying method,  as
the  factors  with  Eigen
values  greater  than
1  are  considered.
As  we  can
see  in  the
table below,  there  are
in  all  6
factors  which  have
Eigen  value  greater
than  1  (Hair
et  al.,  2006). 5
The  percentage  variance
defines  the  percentage
of  total  variance
being  explained  by
that  particular factor.