![](bg6.png)
intelligence. The consequences must be kept in a
professional manner, so as not to be biased and the
quality changes, (Izard: 1977).
Related to the bias of the test, according to Hays
(2013: 62-63) there are two categories of bias sources,
namely (1) the bias associated with the test content is
referring to "inappropriate selection of test item or
general content coverage, and (2) response process
that refers to "situations when item elicit responses
not intended by the test, called a response set. One of
the causes of the bias that comes from the test itself is
the difficulty level of item difficulty (Crocker and
Algina, 1986: 388). Problems that are too difficult or
too easy so that everyone is wrong or everyone is
right, does not give any benefit to decision making
(Subino, 1987).
APM is a non-verbal test that is free from
impaired ability to use verbal language. Ravens
(2011: 8) says that "... APM is a nonverbal assessment
tool is designed to measure an individual's ability to
perceive and think clearly, make meaning out of
confusion and formulate new concepts when faced
with novel information". So this test is structured to
measure the ability to think clearly, which by Subino
(1984: 9) is said as the efficiency of intellectual work,
which will determine the success or failure of a
person in learning; which is demonstrated by learning
easily, quickly, and appropriately. Because doing this
test requires intellectual work efficiently, then people
who are carelessly clear results will be bad.
On the other hand this test is said to be free of
cultural influences, language-free; because it is in the
form of drawing design, in any region or region of
people reading or interpreting the image, so according
to the language in itself (Anastasi, 1988; Naglieri
2009). However Matsumoto (2008: 135-136) says
there is no free test from cultural influence
(intelligence tests were biased and did not accurately
measure the mental ability of people from different
cultures). It is further said that: "... There are ethnic
group differences in measured intelligence (although
the ethnic groups scoring low on the standard tests
change across time). The average score of some
minority groups in the United State are 12 to 15
percentage point lower than the average for European
Americans”. It does not mean that no one is good at
the group, but generally lower. The low achievement
of scores in minority groups may not be due to their
lack of potential but due to other factors that make
their potential un actualized. As in Indonesia, in some
schools students do not want to be invited to think
higher or lazy to face a more difficult problem, yet
nothing likes to say "dizzy". This is a culture that can
make scores in tests low, distractors become ugly, due
to the influence of poor ways of thinking. Therefore,
in the 2013 curriculum it is said that the very thing
that needs to be changed from the teachers is the
wrong mindset.
Based on Subino's research (1984: 241-242) the
questions on APM 90.7% contain the "g" factor,
which Freeman (1965) measured the logical
relationship of non-verbally stated things. To work on
these APM problems, a person is required to be able
to think in an integrated, abstract, and comparative
analysis. Such a pattern of thought is a show of
intellectual acts that must be done as efficiently as
possible (within a limited time). That's why APM is a
test that measures the efficiency of intellectual work.
The rest (9.3%) measures the "s" factor derived from
learning and experience. This means the model
questions that measure the factor "s" is, is a problem
found by the test in learning and everyday life.
To obtain complete information, what factors
influence the achievement of one's APM score,
Matsumoto and Juang (2008) suggests linking it to
school qualifications, age, ideals, home conditions
(family), sex, shelter, (number of siblings), and the
way of learning.
5 CONCLUSIONS
The results of the Ravens Advanced Progressive
Metrices (RAPM) intelligence measurement tool
have met the requirements of good tests. The results
showed that from the difficulty index, the problems
developed had a balanced percentage of items among
easy, moderate, and difficult questions; each question
has a good homogeneity index, the discrimination
index of each question is generally capable of
distinguishing between smart and unintelligent
people, and the choices of answers have functioned
well (no choice is never unselected). In the
discrimination index, RAPM still does not have a
perfect progressive differentiator.
After the requirements of a good test (metric) are
met (validity, differentiation, difficulty, and analysis
are analyzed) then the data obtained will be assured
of its validity, and the test can be used for decision
making. But on the other hand, the thing that will
make the scores obtained a "bias" in the interpretation
is a matter of norm. Therefore the next study is the
adjustment of norms with the sample groups, age, and
sex. As suggested by the Age, Colom, Rebollo, and
Escorial (2004).
REFERENCES
Abad, F., Colom, R., Rebollo, I., Escorial, S., 2004. Sex
Differential Item Functioning in Raven’s Advanced
Validation of Advanced Progressive Matrices as a Instrument Intelligence Test in Indonesian Cultural Perspective
485