have really fulfilled their function in measuring the
intended factors.
Factor analysis test begins with looking at the
value of Keizer-Meyers-Olkin (KMO), which
measures whether the sample is sufficient.
Furthermore, construct validity is seen based on the
factor weight value (loading factor) which shows the
magnitude of the correlation between the initial
variables and the factors formed. The construct
validity is said to be good if the loading factor was
greater than 0.5 (Santoso, 2002). The data analysis
method used to test the hypothesis in this study was
multiple regression analysis.
3.3 Happiness at Work Measurement
Instrument
The instrument of measuring happiness at workwas a
scale that was arranged based on the dimensions of
happiness according to Diener, Oishi & Lucas (2003),
which consists of the cognitive dimension and the
affective dimension. The Happiness at workScale
used five answer choices, namely Very Suitable (VS),
Suitable (S), Neutral (N), Not Suitable (NS), and
Very Unsuitable (VU). The rating for favorable items
is a score of 1 for the Very Unsuitable (VU) choice, a
score of 2 for the Not Suitable (NS), a score of 3 for
the Neutral (N), a score of 4 for the Suitable (S) and
a score of 5 for the Very Suitable(VS). As for the
unfavorable item score 1 for the choice of Very
Suitable (VS), score 2 for the choice of Suitable (S),
score 3 for the choice of Neutral (N), score 4 for the
option Not Suitable (NS) and score 5 for the choice
Very Unsuitable (VU).
The results of factor analysis of the happiness at
workscale showed that the cognitive dimension
obtained KMO values of 0.783 and MSA values
ranging from 0.732 to 0.867. Meanwhile the loading
factor value ranges from 0.642 to 0.843. Furthermore,
in the affective dimension, the KMO value is 0.746
and the MSA value ranges from 0.721 to 0.771.
Meanwhile the loading factor value ranges from
0.670 to 0.870. Based on the results of factor analysis
obtained 9 valid items with an alpha coefficient of
0.844.
3.4 Job Satisfaction Measurement
Instrument
Job satisfaction measurement instrument in the form
of a scale arranged based on aspects of job
satisfaction according to Luthans (2005), which
consists of five aspects, namely the work itself,
salary, promotion, supervision and co-workers. This
scale uses a Likert model with five answer choices
which are Very Suitable (VS), Suitable (S), Neutral
(N), Not Suitable (NS), and Very Unsuitable (VU).
The rating for favorable items is a score of 1 for the
Very Unsuitable (VU), a score of 2 for the Not
Suitable (NS), a score of 3 for the Neutral (N) choice,
a score of 4 for the Suitable (S) and a score of 5 for
the Very Suitable choice. (VS). As for the
unfavorable item score 1 for the choice of Very
Suitable (VS), score 2 for the choice of Suitable (S),
score 3 for the choice of Neutral (N), score 4 for the
option Not Suitable (NS) and score 5 for the choice
Very Unsuitable (VU).
The results of the job satisfaction scale analysis
showed that in the aspect of the work itself obtained
an alpha coefficient of 0.637. The KMO value is
0.687, then the MSA value moves from 0.576 to
0.802. Meanwhile the loading factor value moves
from 0.670 to 0.851. Furthermore, in the aspect of
salary, the alpha coefficient value is 0.731. The KMO
value is 0.653, then the MSA value moves from 0.630
to 0.666. Meanwhile the loading factor value moves
from 0.690 to 0.764. In the aspect of promotion.
Alpha coefficient value of 0.735. The KMO value is
0.689, then the MSA value moves from 0.663 to
0.693. Meanwhile the loading factor value moves
from 0.603 to 0.774. Furthermore, in the aspect of
supervision, the alpha coefficient is 0.804. The KMO
value is 0.754, then the MSA value moves from 0.702
to 0.832. Meanwhile the value of the loading factor
moves from 0.624 to 0.829, and finally on the scale
of job satisfaction shows the aspect of co-workers, the
alpha coefficient is 0.626. The KMO value is 0.698,
then the MSA value moves from 0.659 to 0.750.
Meanwhile the loading factor value moves from
0.603 to 0.778. Based on factor analysis, 21 valid
items are obtained.
3.5 Perceived Organizational Support
Measurement Instrument
The instrument for measuring perceived
organizational support was in the form of a scale
compiled based on aspects of perceived
organizational support according to Rhoades &
Eisenberger (2002) which consists of three aspects:
fairness, supervisor support and organizational
reward and job conditions. This scale also used a type
of Likert scale with five answer choices, namely Very
Suitable (VS), Suitable (S), Neutral (N), Not Suitable
(NS), and Very Unsuitable (VU). The rating for
favorable items is a score of 1 for the Very Unsuitable
(VU), a score of 2 for the Not Suitable (NS), a score
of 3 for the Neutral (N) choice, a score of 4 for the