Test Construction and Development conditions and scales to use, of developing

administration and form. an alternative

Perform the steps of test development through the following:
procedures. form.
• Test Conceptualization

• Test Construction What special What type of How will

training will be responses will meaning be
• Test Tryout required of test be required of attributed to
• Item Analysis users to use the the test taker? the scores on
tests? this test?
• Test Revisions Discusses about Identifies the Will it be based
the responses that on norms or
Test Development
competencies of will be given by criterion?
A term used for all the processes that goes into the method of test users. test takers.
creating a psychometrically sound test.

1. Test Conceptualization Is there any potential Who benefits from the

harm as the result of an administration of the
2. Tests Construction
administration of this test?
3. Test Tryout test?
Identifies the potential Specifies population that
4. Item Analysis harm for test takers may benefit from the test
5. Test Revisions development.

Test Conceptualization Test Construction

The initial process of test development. Developing a scale, writing its items, the scaling procedures,
setting scoring rules, and designing & building a test.
• Stage where the idea for a test is conceived.
Scaling
Questions to Ask
The process of setting rules for assigning numbers of
What is the test What is the Is there a need measurement.
designed to objective of the for this test?
measure? test? • Age-Based Scaling – if the performance is tested as a
Identifying the Details the things Explains the need function of age wherein interpretations of the scores obtained
construct that will that the test will to develop a new may differ based in age-range.
be measured. do to the chosen test.
• Grade-Based Scaling - if the performance is tested as a
construct.
function of grade wherein interpretations of the scores
obtained may differ based on the current grade level of the
Who will use Who will take What content responded.
this test? the test? will the tests
• Unidimensional Scaling - if the performance is tested as a
cover? function of a single construct.
People who can Details the Provides an
administer the target overview of the • Multidimensional Scaling – if the performance is tested as a
test respondents. potential words, function of multiple constructs.
images, and
• Comparative – if the performance is tested as a function of
other aspects of
comparison.
the test.
• Categorical Scaling – if the performance is tested and is
How will the What is the Should more interpreted in a nominal way.
test be ideal format of than one form
administered? the test? of the test be • Rating Scale – a test that group words, statements, or
developed? symbols wherein the strength of a trait is indicated by the test
Specifies the Discusses about Discusses about taker.
testing the item format, the possibility
• Summative Scale – wherein final score is obtained by
summing the ratings across items. Likert Scale is a summative
scale that has numerous alternative responses.

• Method of Paired Comparison – wherein takers were

presented in pairs where they are asked to compare.

• Sorting Method – provides an ordinal information through

either comparative or categorical scaling.

• Guttman Scale – wherein items are arranged sequentially

from weaker to stronger expression of attitudes, beliefs, or
feelings that are being measured.

Writing Items Test tryout

What range of content – how much of the construct being Tryout on target population
measured will be included in the test.
Administration to a representative sample of testtakers under
Type of Item Format conditions that simulate the conditions that the final version of
Selected Response the test will be administered under

• Multiple Choice A stage wherein items should be tried out on people who are
similar in critical demographic aspects to the identified target
• Matching Type population.
• Binary Choice

• Likert Type

Constructed Response
Pilot Work
• Completion
• Also known as Pilot Study.
• Short Answer
• Wherein it includes preliminary research about the
• Essay creation of the prototype of the test.
• It involves literature review, experimentation, creations,
How Many Items revisions, and deletion of preliminary test items.
usually 2-3x the desired number of items. Item Analysis
For Computer Administration Testing the Test.
• Item Bank – large pool of items. Statistical procedures employed to assist in making judgments
about which items are good as they are, which items need to
• Computer Adaptive testing or Item Branching – tailored
be revised, and which items should be discarded.
based on the test taker performance.
Is set of statistical scrutiny that the test data should undergo.
Scoring Items
• Item Difficulty Index – obtained by calculating the
The process of assigning numbers to the responses.
proportion of the total number of test takers who answered the
• Cumulative Scoring – summing scores obtained from each test correctly.
items designated for each or a construct.
• Item Reliability Index – indication of an internal
• Class Scoring – when scores categorize people on a certain consistency of the test.
class or type.
• Item Validity Index – the degree to which a test is
• Ipsative Scoring – a scoring procedure that results into a measuring what it purports to measure.
total score coming from a variety of items from the test.
• Item Discrimination Index – how an item separates or
discriminates between high and low scorers.

Other analytical techniques.

• Analysis of items alternatives – analysis if the effect of Do’s & Don’ts in Test Construction (Item Writing)
foils or distractors.

• Guessing Analysis – no generally acceptable method but

may be through consistent responding based on thresholds
established.

• Item Fairness – the degree, if any test item is biased.

• Speed Test Analysis – analyzes the amount of time spent

but may yield misleading results.

Qualitative Analysis

Data generation and analysis that analyses primarily verb

responses.

• Thinking Aloud Test Administration.

• Expert Panel Interviews.

Test Revisions

Continuous modifications and improvement.

Reasons for revisions:

• Verbal content is not understood.

• Reliability and validity measures can be improved.

• Norms are no longer adequate, hence it needs to be

updated.

• Population and culture changed.

• Material is considered as outdated.

• Compared to original, it can be improved.

Validation Processes

Cross-validation

The process of revalidation on a different sample.

Co-validation

Using two or more tests on the chosen sample.