Methods of Estimating Reliability

The size of the reliability coefficient is related to the methods of

estimating reliability in the following way:

Split half method: Largest reliability coefficients reported for a

given tests. Estimate inflated by factors such as speed is another
sub type of internal consistency reliability. The process of obtaining
split- half reliability is begun by splitting in half all items of a test
that are intended to probe the same area of knowledge ( e.g, World
War ll) in order to form two sets of items. The entire test is
administered to a group of individuals, the total score for each set
is computed, and finally the split-half reliability is obtained by
determining the correlation between the two totals set scores.
Test- retest method: medium to large reliability coefficients reported
for a given test and it become smaller as time interval between test
is increased. It is a measure of reliability obtained by administering
the same test twice over a period of time to a group of individuals.
The scores from Time 1 and Time 2 can then be correlated in order
to evaluate the test for stability over time.
Equivalent forms method: (without time interval) medium to large
reliability coefficient reported for a given test.

Equivalent forms method: (with time interval) smallest reliability

coefficient reported for a given test becomes smaller as time interval
between forms is increased.

Types of Reliability
1. Scorer reliability
2. Content reliability
3. Temporal reliability

Scorer reliability: concerns itself with the degree of arrangement between

two scorer of the same test answer. It deals both the degree of
consistency in grading the same test answer by the same scorer on two
different occasion and this is also called inter scorer reliability.

Content reliability: it deals with the ability of all the items of a test to
measures competencies in the same general content area.

Temporal reliability: it is the third major dimension of reliability and as the

name suggests it concerns itself with the stability off the result of attest
over time, for attest to be reliable it must yield consistent result overtime.
Factors Influencing Reliability
1. Data collection method.
2. Interval between testing occasion.
3. Test length, higher length-more reliable.
4. Speed of the method.
5. Group homogeneity.
6. Difficulty of the items.
7. Ambiguous wording.
8. Inconsistency in test administration.
9. Objectivity of scoring is more reliable than subjectivity of scoring.

Measures to Improve Reliability

1. Limiting subjectivity of all kind.
2. Ensure that questions are clear.
3. Ensure that expected answers are definite and objective.
4. Checking to make sure the time limits are adequate.
5. Giving simple, clear and unambiguous instruction.
6. Keeping choice with in a test paper to minimum .
7. Conducting test under identical and ideal examination conditions.
8. When using less reliable methods increase the number of question
observation or examination time.

Relationship between Validity and Reliability

A. Validity is truth-fullness, while reliability is truth worthiness. These
both are the two aspect of efficiency.
B. The test cannot be valid unless it is reliable.
C. Validity includes reliability.

OBJECTIVITY
Objectivity in measurement helps to increase test validity and reliability,
this is the extent to which independent and competent examiners agree
on what constitutes a good answer for each of the items of a measuring
instrument. Most standardized test of aptitude and achievement are high
in objectivity.

Advantages of Objectivity
Objective test are more reliable.
They enable to more extensive survey in a given time than can be
obtained by any other type of test.
 The time element is reduced.
A greater number of items can be tested.
Greater the validity of test.

Disadvantages of Objectivity
More time is required to prepare good objective question.
It provides little or no opportunities for the measurement of the
students ability.

USABILITY
It is important that a test is practical for its purpose. This implies such as
factors as the time taken to conduct the test, the cost of using it and
practicability for everyday use. It includes the practical consideration such
as cost, mode of administration, scoring procedures availability of test and
test score interpretation.

Errors of Measurement

There are two types of errors that are concerned in measurement random
error or chance error and systematic or constant error. Validity is
concerned with random error. The three components of measurement
score are true score (T) observed score (O) and the error score (E).

NON STANDARDISED TESTS

Introduction

Non-standardized assessment looks at an individuals performance, and

does not produce scores that allow us to compare that performance to
anothers. It allows us to obtain specific information about the student,
and this can be in different formats.

Non Standardized Testing Is Informal Testing

Another term for non-standardized testing is informal testing. These tests

are classroom tests and are usually developed by the teacher as opposed
to some group of outside testers. These classroom tests assess students
learning over a period of time or after a particular unit of study. A score of
80% on a multiple choice test after reading a short story is a non-
standardized score because it does not tell us how the student did in
relation to his peers.
Criterion Referenced Measurement

This is also a form of non-standardized testing. The desires level of

achievement is the criterion . The criterion referenced measurements are
used to measure learning for a particular student. This way the teacher
can document whether or not learning is taking place. Evidence of
learning , or not, is readily apparent, and the focus here is on the
performance of an individual student as opposed to the norm-referenced
tests.

Domain-referenced tests are similar to criterion-referenced. Performance

is measured against a well-defined body (domain) of knowledge or tasks,
and the focus is on the individual.

Forms of Non-Standardized Testing

Forms include portfolios, interviews, informal questioning, group

discussions, oral tests, quick pop quizzes, and exhibitions of work, projects
and performance exams.

With portfolios the student gathers his work over a period of time, and the
teacher will evaluate the work based on a scoring guideline. The student
is encouraged to reflect on his work, which enhances the learning process.
Performance exams are tests given to all students and are based on a
students performing some task, like writing an essay, or giving an oral
presentation. These tasks are created by the teachers who teach the
students, and so the exams drive the curriculum. It makes more sense for
those doing the teaching to create the tests.

Accountability and Non-Standardized Tests

Parents and the community have a right to know how students are doing:
therefore, non-standardized tests need to show how well schools and
students are doing. Teachers are constantly assessing their students, and
by doing so they are constantly adjusting and changing their teaching to
meet individual students needs. There can still be accountability with
non-standardized assessment that provides parents, local officials, and
state officials with the information needed. Teachers can be in constant
touch with parents through the internet, by calling, by parent conferences
and by sending home progress reports and samples of work.

Success with Non-Standardized Testing

The key questions to ask with any kind of assessment is, What is the
purpose of this assessment ? and Is this purpose meaningful and
worthwhile? If these questions are constantly referred to and constantly
addressed then the assessment in itself is important, and this helps
teachers address what is important to learn. Its a kind of backwards
design. Ultimately the goal is to help students to learn, and to help them
to learn the information and the skills that are important.