The FACTOR
procedure
is used to conduct EFA in SAS. You must be careful when conducting
an EFA in SAS because the FACTOR
procedure
is also used to conduct PCA—if you do not specify an extraction
method or if you just select one of the extraction options, you could
end up with a PCA and not an EFA! Below is a brief summary of the
syntax and the basic options available. We will delve into further
details and options in the chapters to come.
PROC FACTOR DATA = dataset-name
NFACTORS = number-of-factors-to-retain
METHOD = factor-extraction-method
ROTATE = rotation-method;
VAR variables-to-include;
RUN;
In the above syntax, you would specify
all of the arguments that are highlighted. You would identify the
data set to use for the analysis, tell SAS the number of factors to
extract (we will discuss the multi-step procedure for this in Chapter
3), specify one of the seven EFA extraction methods (note there are
other options, but they are not EFA extraction methods), specify one
of the 25 rotation methods, and list the variables to be factored.
If you do not specify the NFACTORS
, METHOD
,
and ROTATE
options along with the VAR
statement,
then the analysis will still run, but SAS will use its default options
of retaining the number of factors identified by the Kaiser Criterion,
performing PCA extraction, not conducting rotation, and using all
of the variables in the data set. Thus, it is best practice to get
into the habit of specifying all of the options, even if you choose
to use one of the methods that is a default, so that you do not accidentally
overlook a key component of your analysis.
Let’s
take a minute to quickly review the structure of the syntax above.
Note that there is a semicolon on the end of some lines and not others.
The placement of the semicolon is of insurmountable importance in
SAS—if it is in the wrong place your code will not run and
you will get errors galore! The semicolon signals the end of a statement and
tells SAS how to parse what we are requesting. In the above syntax,
we have a PROC FACTOR
statement, a VAR
statement,
and a RUN
statement. You will notice that
there are some other key terms in the statement that allow users to
specify the inputs for our analysis (e.g., DATA =
, NFACTORS
=
). These are referred to as options within
the statement. Finally, notice that the syntax starts with PROC
FACTOR
. This tells SAS that we are using the FACTOR
procedure,
and the following statements and options will specify the details
of the analysis we would like to conduct. We will use this terminology
throughout this book.
We hope most of what
was just described was a review. If it was not, we recommend you first
review some other excellent introductory texts to using SAS (e.g.,
Cody, 2007; Delwiche & Slaughter, 2012) before attempting to follow
the syntax in this book. It is expected that readers have an understanding
of basic data manipulation, the DATA
step,
and at least a few basic procedures (e.g., MEANS
, SORT
, FREQ
).
In addition, readers are encouraged to refer to the SAS documentation
about PROC FACTOR
for additional information
throughout this book (SAS Institute Inc., 2015).