International Journal of Pedagogics
150
https://theusajournals.com/index.php/ijp
VOLUME
Vol.05 Issue03 2025
PAGE NO.
150-154
10.37547/ijp/Volume05Issue03-42
Analysis of test systems used in classical and modern test
theories for assessing students' knowledge
Rasulov Ulugbek Murodulloyevich
Assistant at the Uzbekistan-Finland Pedagogical Institute, Uzbekistan
Received:
27 January 2025;
Accepted:
25 February 2025;
Published:
23 March 2025
Abstract:
The article presents an analysis of test systems used in classical and modern testing theory in the world.
Based on classical theory and widely used systems, the objectives of the iSpring, Google Forms, SAT, GRE, GMAT,
LSAT, MCAT tests, the level of academic orientation, structure, cases of computer adaptation, duration of testing,
information about the coverage of subjects in the test are presented. The main purpose of the article is to show
how important it is to create test systems based on modern testing theory. For this purpose, modern testing
systems StartExam, Proaction, TOEFL, and IRT tests were analyzed. It is shown that it is advisable to use the
modern IRT test system in higher education institutions.
Keywords:
Test, system, task, result, testing, assessment, modern, model, theory.
Introduction:
Currently, test technologies in the higher
education system are widely used to monitor the
academic performance of students during the learning
process. Therefore, let us examine the test systems
based on classical and modern test theories used in
foreign and Commonwealth countries.
Test systems based on Classical Test Theory include
iSpring, GRE, SAT, GMAT, TOEFL, and others. Let us
review these test systems.
The iSpring test system typically operates based on
classical (traditional assessment) methods and is aimed
at automating the learning and assessment processes.
iSpring functions as part of a platform for creating e-
learning and educational materials. It provides
convenient tools for creating, evaluating, and analyzing
various types of tests and surveys. iSpring is mainly
used for conducting online tests. It simplifies the
creation of tests, adding questions, and assessing them.
This system can be used on computers or mobile
devices. iSpring automatically grades tests and analyzes
the results in real time. The results and statistics from
the tests include correct and incorrect answers, time
tracking, and final scores. iSpring also enables the
creation of interactive learning materials and allows
the integration of video and audio. This helps engage
students and make educational content more
interesting. The system also offers adaptive testing
capabilities, which means questions can be adjusted
during the test based on students’ responses. This
helps better identify the individual knowledge level of
each student. iSpring is a convenient platform for
creating, evaluating, and analyzing modern and
interactive tests. It offers various test types, statistical
tools, and real-time assessment capabilities. Adaptive
testing and computer-based administration make
iSpring a modern test system. This system is used for
creating tests and assessment systems in schools,
universities, and online courses, for evaluating staff
knowledge, conducting training and certification
processes, and measuring knowledge level or work
efficiency through testing and analysis. iSpring is a
modern and interactive platform for creating,
assessing, and analyzing tests, providing various test
types, statistical data, and real-time evaluation
features.
The GRE (Graduate Record Examination) is a test
designed to assess analytical, mathematical, and verbal
skills acquired during undergraduate studies and is
considered one of the classical tests. The test is divided
into two types: General (for business schools) and
Subject (for scientific and technical fields). The GRE
Gene
ral test is a general exam that assesses a student’s
knowledge of algebra, mathematics, geometry, and
English. This version of the test is required for
admission to graduate programs abroad. The GRE
International Journal of Pedagogics
151
https://theusajournals.com/index.php/ijp
International Journal of Pedagogics (ISSN: 2771-2281)
Subject test evaluates a student's knowledge in a
specific subject area. The GRE assesses verbal and
mathematical skills, written expression, and critical
thinking abilities. This comprehensive and rigorous
exam not only evaluates a student’s analytical abilities
but also determines their readiness for graduate-level
studies. Globally, the GRE is available in two formats: a
traditional paper-based standardized exam and an
online version conducted via computer. In the latter,
the difficulty of the questions depends on the test-
taker’s previous answers. Thus, the a
ccuracy of a
student’s responses in the verbal and math sections
determines the complexity of the questions in
subsequent sections. The GRE is not a fully traditional
test because it is considered an adaptive test system.
However, it is not fully adaptive either, as it adapts only
at the section level, not at the individual item level.
Therefore, the GRE is considered a partially adaptive
test. The GRE is based on the Item Response Theory
(IRT) and adapts not at the individual item level but at
the section level.
The SAT (Scholastic Assessment Test) is a standardized
test used in the United States to assess the knowledge
of applicants and high school students. In order to be
admitted to prestigious U.S. universities, one must
achieve a high score on the SAT, and in most cases,
American students take this test before graduating
from high school. Every student who dreams of
studying abroad is advised to take the SAT test.
Since its introduction in 1926, the SAT has undergone
several changes, primarily in its language and scoring
system, as well as in its name
—
eventually being
renamed the Scholastic Assessment Test. Today, SAT is
recognized as an independent brand name and is
widely used.
The SAT test consists of two components:
- SAT Reasoning Test (Thinking Test): assesses
mathematical skills, reading comprehension, and
writing ability.
- SAT Subject Test: includes 20 tests divided into 5 key
blocks
—
sciences, mathematics, world and U.S. history,
English literature, and foreign languages.
Although the SAT has recently transitioned into a
partially adaptive format, it still remains closer to a
classical testing approach. Starting in 2023, the SAT
moved to a digital format and became a semi-adaptive
test. This means the SAT is no longer a fully traditional
test but, like the GRE, it adapts to a test-
taker’s skill
level in a limited way.
GMAT (Graduate Management Admission Test) is a
computer-adaptive test based on the principle of
sequential analysis. In such tests, the selection of the
next question depends on the test-
taker’s response to
the previous one. The GMAT is required for admission
to business schools and is used worldwide. It assesses
students' analytical, verbal, and mathematical skills, as
well as their writing and reading abilities. GMAT results
are valid for up to five years.
The GMAT consists of four parts:
1. Analytical Writing Assessment
–
a written task with a
30-minute limit, where the test-taker must write an
essay based on analysis and reasoned critique.
2. Integrated Reasoning Section
–
includes 12 questions
with a 30-minute time limit.
3. Quantitative Section
–
consists of 37 questions with
a 75-minute limit.
4. Verbal Section
–
includes 41 questions, also with a
75-minute time limit.
Unlike standardized tests, the GMAT is a computer-
adaptive test. This means the system adjusts the
difficulty and content of the questions based on the
student’s abilities. The test begins with medium
-
difficulty questions. If the test-taker answers correctly,
more difficult questions follow; if the answers are
incorrect, the computer adapts by offering easier
questions. This process continues until the system
accurately determines the test-
taker’s ability level.
Another globally recognized test is the LSAT (Law
School Admission Test), which is used for admission to
law schools. This test evaluates reading skills and verbal
logical reasoning. In the U.S., Canada, and many other
countries, LSAT results are considered a key indicator
for law school admissions.
The LSAT consists of five sections:
- Logical reasoning
- Reading comprehension
- Analytical reasoning
- Writing sample
- Experimental section (this section is unscored)
Each section is allotted 35 minutes. As mentioned, only
four of the five sections contribute to the final score;
the experimental section is used to pre-test new
questions and formats. Each section includes three
types of questions: reading, analytical, and logical.
Thus, the LSAT assesses whether a student is ready for
legal education.
The LSAT is scored on a scale from 120 to 180, with each
question carrying equal weight. To apply to law schools,
applicants must submit their total score across all test
sections. The LSAT can be taken up to three times per
year.
Another globally recognized test is the MCAT (Medical
International Journal of Pedagogics
152
https://theusajournals.com/index.php/ijp
International Journal of Pedagogics (ISSN: 2771-2281)
College Admission Test), which is used for admission to
medical universities. The MCAT is extremely important
for future doctors and veterinarians. Primarily, this test
is necessary to evaluate the knowledge level of those
applying to educational institutions in the medical field.
In other words, the goal of the MCAT is to assess the
student’s knowledge and analytical thinking skills to
determine their readiness for studying in medical
institutions.
The MCAT consists of four sections: reasoning, physics,
biology, and essay writing. The full duration of the test
is 7 hours. A student can earn up to 15 points in each
section, making the total maximum score 60 points. A
distinctive feature of this test is that no points are
deducted for incorrect answers.
The validity of the MCAT certificate and scores depends
on the specific medical school the student is applying
to. Typically, the certificate is valid for 5 years. If a
student fails to take the MCAT twice, they must request
special permission to take it a third time. Although the
MCAT is considered a classical test, it has been adapted
to meet the demands of modern medical education. In
its current form, it can be classified among modern
tests.
In addition to the classical tests discussed above,
modern test theory plays a particularly important role.
In 2023, four popular test tools were used for staff
assessments: Google Forms, iSpring, StartExam, and
Proaction. Even during global economic crises, many
companies did not cut back on employee training
budgets. Research has shown that even small
businesses value their staff. For companies, it is crucial
to select candidates not only based on professional
qualifications but also on how well they align with the
company culture.
To achieve this, many organizations found it
appropriate to use tests in the hiring process. There are
several reasons for this:
1.
Expert-based design: Unlike other selection
methods, tests are well-studied tools developed by
sociologists, psychologists, and specialists. Employers
can set the number of questions, testing time, result
presentation format, and more based on expert
recommendations.
2.
Transparency: Testing ensures a higher level of
transparency
—
all employees receive the same set of
questions with the same difficulty level, reducing the
influence of human bias.
3.
Proven validity: While tests are not the most
precise tools, their simplicity and cost-effectiveness
have led to their widespread use. Moreover, they are
easy to conduct
—
tests can be created in spreadsheets
or on paper. When the number of test-takers is large,
automated computer-based test systems are used.
Now let’s explore in detail modern tests based on IRT
(Item Response Theory), which are widely used in
higher education.
The Proaction testing system is one of the modern
systems built on Item Response Theory (IRT),
distinguishing it from classical test theory. Proaction
uses IRT methodology to evaluate tests, allowing for
individual analysis of each question and each student.
IRT models make tests more adaptive and precise.
Proaction applies a personalized approach in test
analysis and focuses on the individual evaluation of
every student. In this system, each test response, its
difficulty level, and the reliability and validity of the
administered
tests
are
assessed
using
IRT
methodology. The system is developed using advanced
technologies and modern methodologies.
Compared to classical test theory, Proaction offers a
more nuanced and individualized approach. It
evaluates each stude
nt’s responses and knowledge
level separately, and thus provides personalized test
results for each individual.
Adaptive Testing System
An adaptive testing system is an intelligent test system
that adjusts the test questions according to the test-
taker’s l
evel of knowledge. Unlike traditional tests, this
system changes the difficulty of each subsequent
question based on the response given to the previous
one. The adaptive testing system operates on the basis
of Computerized Adaptive Testing (CAT) technology.
This technology analyzes the test-
taker’s responses and
automatically selects questions that match their level
of knowledge during the test process. It allows for the
individual assessment of students’ knowledge and
provides them with questions tailored to their abilities.
In higher education, adaptive testing systems can be
highly effective in assessing students’ knowledge,
enhancing personalized learning, and optimizing the
educational process. The benefits of adaptive testing in
higher education include:
Individualized approach
–
Questions are tailored to
each student’s knowledge level.
Accurate assessment
–
Results reflect the student’s
actual proficiency more precisely.
Time efficiency
–
Reliable outcomes can be achieved
with fewer questions.
Identification of strengths and weaknesses
–
Helps
determine areas where the student excels or struggles.
Optimization of learning
–
Enables instructors to
identify which topics need to be explained in greater
International Journal of Pedagogics
153
https://theusajournals.com/index.php/ijp
International Journal of Pedagogics (ISSN: 2771-2281)
depth.
Compared to traditional tests, adaptive testing
provides more accurate and efficient assessments
while allowing for personalized instruction. Though
traditional tests are simpler and more familiar to most,
adaptive tests are more effective in today’s modern
education systems.
TOEFL (Test of English as a Foreign Language) is a
standardized academic English test, primarily taken by
those applying to universities in the United States. In
addition to the U.S., TOEFL scores are recognized in
many other countries as proof of English language
proficiency. The TOEFL includes several formats, with
the most common being TOEFL iBT and TOEFL
Essentials.
The TOEFL iBT (Internet-Based Test) uses Item
Response Theory (IRT) to adapt the difficulty of
questions according to the test-
taker’s responses,
allowing for more precise measurement of English
proficiency. Other TOEFL versions include TOEFL ITP
(Institutional Testing Program), TOEFL Junior, and
TOEFL Primary, which are designed for school-level
students
The TOEFL iBT test consists of four sections: Reading,
Listening, Speaking, and Writing. It lasts approximately
3 hours and is intended to assess a student’s readiness
to study in an English-speaking academic environment.
Students typically prepare for the TOEFL iBT over 3 to 6
months. In contrast, the TOEFL Essentials test is a
simpler and more affordable version with a lower
academic focus.
TOEFL is considered a modern test system because it
primarily uses the IRT model. Its adaptive version
changes question difficulty and type based on the test-
taker’s responses, wh
ich requires the use of IRT.
IRT (Item Response Theory), often referred to as the
modern theory of testing, theory of responses to items,
or parameterization and modeling theory of
pedagogical tests, is a collection of methods that
estimate the probability of a test-taker answering
various difficulty-level items correctly.
IRT is used to eliminate uninformative questions from
questionnaires, assess the relationship between latent
traits and observable variables, and optimize
recommended tasks for each test-taker. In the IRT
model, neither the test items nor the test itself are
evaluated in isolation
—
instead, the interaction
between the respondent and the item is modeled.
In psychometrics, IRT is considered a foundational
framework for analyzing and evaluating tests,
questionnaires, and measurement tools. It assumes a
statistical relationship between responses and the
underlying traits being measured. Various statistical
models are used to estimate both task and respondent
parameters. Unlike traditional tests where all questions
are considered to have equal difficulty, IRT recognizes
that each item has unique properties that must be
included in the model.
This differs from methods like Likert scaling, where all
tasks are assumed to be interchangeable replications.
In contrast, IRT treats each item as a data point with
unique parameters that must be incorporated into the
model.
CONCLUSION
In summary, IRT models the probability of a test-taker
responding correctly to each item. A core characteristic
of modern testing theory is the separation of
parameters for both test-takers and items
—
meaning
that the probability of a correct response depends on
the interaction between the respondent’s and the
item’s latent traits. The specific form of this interaction
is determin
ed by the researcher’s assumptions and is
described through precise mathematical functions,
forming the IRT model.
Modern IRT models draw from techniques such as
factor analysis, generalized linear models for mixed
effects, network models from statistical physics (e.g.,
Markov fields and the Ising model), and various
methods from data science (e.g., collaborative filtering
models and restricted Boltzmann machines).
Today, IRT is regarded as one of the most advanced and
theoretically grounded testing approaches that
accounts for human nature during the test process and
is widely used in data-driven educational and
psychological assessments.
REFERENCES
Miller, K. (2011). The TOEFL Test: A Study Guide for
English Language Learners. McGraw-Hill
The College Board (2016). The Official SAT Study Guide.
College Board
Educational Testing Service (ETS) (2018). The Official
Guide to the GRE General Test. ETS
Kaplan (2019). Kaplan GMAT Complete 2020. Kaplan
Publishing
Kelley, K. (2017). The LSAT Trainer: A Comprehensive
Self-Study Guide for LSAT Preparation. LSATMax
Kaplan (2019). Kaplan MCAT Complete 7-Book Subject
Review. Kaplan Publishing
Hambleton, R. K., Swaminathan, H., & Rogers, H. J.
(1991) Fundamentals of Item Response Theory
Baker, F. B., & Kim, S.-H. (2017)The Basics of Item
Response Theory Using R
International Journal of Pedagogics
154
https://theusajournals.com/index.php/ijp
International Journal of Pedagogics (ISSN: 2771-2281)
Wainer, H., & Dorans, N. J. (2000). Computerized
Adaptive Testing: A Primer. Lawrence Erlbaum
Associates
Davies, R. (Ed.) (2006). Computerized Testing and the
Internet. Cambridge University Press.
Pashley, P. J., & Jodoin, M. G. (2014). Handbook of
Automated Essay Evaluation. Information Age
Publishing.
