Exporatory principal components analysis
In this example, the principal components analysis will be conducted on dataset/dataframe called dat
containing ten variables that represent data collected with ten questions. These ten variables are called item1
, item2
, item3
, item4
, item5
, item6
, item7
, item8
, item9
, and item10
.
SPSS
To perform the principal components analysis, use:
FACTOR
/VARIABLES item1 item2 item3 item4 item5
item6 item7 item8 item9 item10
/MISSING LISTWISE
/ANALYSIS item1 item2 item3 item4 item5
item6 item7 item8 item9 item10
/PRINT INITIAL ROTATION
/PLOT EIGEN
/CRITERIA EIGEN(1) ITERATE(25)
/EXTRACTION PCA
/CRITERIA ITERATE(25)
/ROTATION VARIMAX
/METHOD=CORRELATION.
To extract a specific number of components, for example three components, replace EIGEN(1)
by FACTORS(3)
.
To suppress the printing of factor loadings lower than a specific value, for example .3, use:
FACTOR
/VARIABLES item1 item2 item3 item4 item5
item6 item7 item8 item9 item10
/MISSING LISTWISE
/ANALYSIS item1 item2 item3 item4 item5
item6 item7 item8 item9 item10
/PRINT INITIAL ROTATION
/PLOT EIGEN
/CRITERIA EIGEN(1) ITERATE(25)
/EXTRACTION PCA
/CRITERIA ITERATE(25)
/ROTATION VARIMAX
/METHOD=CORRELATION.
/FORMAT=SORT BLANK(.3).
R
To perform the principal components analysis, use:
facComAnalysis(dat,
items=c('item1', 'item2', 'item3',
'item4', 'item5', 'item6',
'item7', 'item8', 'item9',
'item10'),
fm='pca',
rotate='varimax');
To extract a specific number of components, for example three components, use:
facComAnalysis(dat,
items=c('item1', 'item2', 'item3',
'item4', 'item5', 'item6',
'item7', 'item8', 'item9',
'item10'),
fm='pca',
rotate='varimax',
nfactors = 3);
To suppress the printing of factor loadings lower than a specific value, for example .3, use:
facComAnalysis(dat,
items=c('item1', 'item2', 'item3',
'item4', 'item5', 'item6',
'item7', 'item8', 'item9',
'item10'),
fm='pca',
rotate='varimax',
maskLoadingsUnder = .3);