2. PSYCHOMETRICS
Psychometry is a branch of psychology that focuses on the field of measurement of psychological
constructs.
However, the measurement of psychological construct does not only occur in psychology, but also
in education.
Therefore, psychometrics can be said as a field of science related to testing, measurement,
assessment etc. which is in the area of psychology and education.
The aspects measured usually include knowledge, skills, intelligence, personality, attitudes, abilities
and achievements.
3. Psikometri
Regionally, the history of the development of psychometrics is divided into two, namely Victorian
Stream and Germanian Stream..
Victorian Stream was originally initiated by Sir Francis Galton, who was actually inspired by his
teacher, Charles Darwin (author, "The Origin of Species"). Galton then wrote a book entitled
"Hereditary Genius" which contains about individual characteristics and how these characteristics fit
(fit) with each other. Galton is referred to as "Father of Psychometrics". Galton's best-known student
is Karl Pearson.
Germanian Stream was pioneered by Herbart who researched the "Structure of Human
Consciousness" through scientific methods. He is able to make mathematical modeling of the
structure of the human mind. E.H Weber and G.T Fechner are the successors of Herbart, until then
Wilhelm Wundt had an experimental laboratory in the field of psychology.
4. Psychometrics Theory
Psychometricians have developed two main theories in the field of measurement,
namely classical test theory (CTT) and item response theory (IRT).
Both have different approaches to scoring psychological constructs, especially on
the scale used and the features they have.
Psychometrician also developed the covariance / correlation method into a
method called factor analysis.
In addition, in the context of more complex correlations, the psychometrician
developed a method called structural equation modeling and path analysis.
6. Some well-known Psychometricians
Francis Galton
Charles Spearman
Raymond Cattell
Alfred Binet
Karl Joreskog
Bengt Muthen
Kenneth Bollen
Peter Bentler
Barbara Byrne
Sumadi Suryabrata
Jahja Umar
Saifuddin Azwar
Urip Purwono
Bahrul Hayat
Bastari
7. Whatever exists at all, exists in some amount. Anything that
exists in amount can be measured (Thorndike, 1918).
What is test theory? Why is it needed?
Did test theory originate with psychology, or it is a new field of study?
How is test theory related to research design and statistics?
8. Fundamental Concepts of Measurement
Weitzenhoffer (1951) defines measurement as "a certain procedure that is inherent
in the world of physics and is carried out by an observer.
Stevens (1946) said that measurement is giving a number / quantification of an
object according to a certain rule.
Furthermore, Lord & Novick (1968) extended Stevens's definition, to be a valid
measurement of the property of an object rather than the object itself.
9. Fundamental Concepts of Measurement
Measurement is the procedure of giving norms (quantification) of an attribute or
variable in a continuum.
Or, measurement is the activity of giving norms in the form of numbers on
variables or attributes that are to be or have been measured.
Norms are a measure of explicit quantification of attributes.
Therefore, in measurement, norms must be clearly defined.
10. Fundamental Concepts of Measurement
For example:
when someone wants to measure temperature, then one must first set the norm to be
used in measuring temperature, whether using Celsius, Fahrenhait or Kelvin, so that we
will get a real measurement of the temperature, for example 23.8⁰ Celsius.
The norm used in measurement must be the right norm for the attribute to be
measured.
11. Psychological variables are generally latent variables
Measurements can only be indirect
Procedures:
Fundamental Concepts of Measurement
stimulus respon Data Skala
12. Fundamental Concepts of Measurement
Measuring = Determine the Position on a Scale.
The scale must have (1) origin and (2) units.
Measurements are for continuum variables.
For categorical variables = Counting.
13. Fundamental Concepts of Measurements
Before measuring a variable, a rule about theoretical construct with observable
behavior is needed.
This process is referred to as an operational definition.
Usually, psychologists or psychometricians, make a "tool" to get a sample of the
behavior.
This tool in psychology is known as "TEST".
14. Fundamental Concepts of Measurements
According to Crocker & Algina (1986) the test is defined as a standard procedure
for obtaining a sample of behavior from a specified domain.
Based on behaviors, the test is divided into two, namely:
1. Optimal performance, which is to measure a person's talent or ability where he is asked
to show his best abilities,
2. Typical performance, which is a questionnaire or inventory in which a person is asked to
judge about his feelings, attitudes, interests and reactions to certain situations.
15. Fundamental Concepts of Measurements
Why Psychologists and Educators try to measure constructs that cannot be directly
observed?
Discussed in a group!
In the field of Psychology and Education, a construct provides the easiest method for
labeling the same behavior.
Example: intelligence
Without a construct, the measurement of human behavior becomes very confusing.
16. Problems in Measurement
1. No single approach to the measurement of any construct is universally accepted.
2. Psychological measurements are usually based on limited sample of behavior.
3. The measurement obtained is always subject to error.
4. The lack of well-defined units on the measurement scales poses still another
problem.
5. Psychological constructs cannot be defined only in terms of operational
definitions, but must also have demonstrated relationships to other constructs or
observable phenomena.
17. No single approach to the measurement of any construct is
universally accepted.
Because measurements of a psychological construct are always indirect, based on
behaviors that are perceived as relevant to the construct under study, there is
always the possibility that two theorists who talk about the same construct may
select very different types of behavior to define that construct operationally.
Obviously different measurement procedures would result from these different
operational definitions and might well lead to different conclusions about the
student’s levels of knowledge.
18. Psychological measurements are usually based on
limited sample of behavior.
It is impossible to ask students to solve all the problems in long division that they
might be expected to solve.
Therefore, any procedure to measure their construct must involve only a sample of
all such problems.
Determining the items and the domains are necessary to provide adequate
sample of the behavioral domains.
19. The measurement obtained is always subject to
error
Since measurement in psychological fields are based on limited sample and usually are
taken at only point in time, it could have different results.
Perhaps, it is because of fatigue, boredom, forgetfulness, guessing or careless.
Such inconsistencies in individuals scores due to sampling tasks or occasions must be
regarded as errors.
Thus persistent problem in psychological measurement is how to estimate the degree
of error in a given set of observations.
20. The lack of well-defined units on the measurement
scales poses still another problem
If a student cannot answer all the items in a test, does it mean he or she has “zero”
mastery of a skill?.
In psychology or in social sciences, there is no absolute value of zero.
Defining the properties of the measurement scale, labelling the units, and
interpreting the values derived are complex issues.
21. Psychological constructs cannot be defined only in terms of operational
definitions, but must also have demonstrated relationships to other
constructs or observable phenomena
A psychological measurement, even though it is based on observable responses,
would have little meaning or usefulness unless it could be interpreted in light of
the underlying theoretical construct.
Psychological measurements can be defined on two levels (Lord & Novick, 1968):
The construct must be defined in terms of observable behavior
The construct must be defined in terms of its logical or mathematical relationship to
other constructs within the theoretical system
23. References
Crocker , L., & Algina, J. (1986). Introduction to Classical and Modern Test Theory.
Harcourt Publishers: New York.
Lord, F.M., & Novick, M.R. (1968). Statistical Theories of Mental Test Scores.
Reading, Mass: Addison-Wesley.