Validty and Reliability Handout

Download as pdf or txt
Download as pdf or txt
You are on page 1of 39

Validity and Reliability

Objective of the Session:

This session aims to help the participants


show understanding of the concepts,
purposes, principles and processes in
establishing validity and reliability of
measures/tests.
Outline of the Session

1.What is measure/test validity?


2.What are the different types of test validity?
3.How do we perform statistical test of validity?
4. What is reliability?
5. What are the different types of test reliability?
6. How do we perform statistical test of
reliability?
What is Test Validity?

Validity indicates whether an assessment tool is


measuring what it intends to measure. Validity
estimates indicate whether the latent variable
shared by items in a test is in fact the target variable
of the test developer.
What is Test Validity?

Validity also refers to the ability of a scale or test to


predict events, relationship with other measures,
and representativeness of item content.
Different Types of Test
Validity:

1. Content Validity
2. Criterion-Prediction Validity
3. Construct Validity
4. Convergent Validity
5. Divergent Validity
Content Validity:

NATURE USE PROCEDURE STATISTICS


Systematic More appropriate Items are based Typically no
examination of for on statistics
the test content cognitive instructional
to determine measures (e.g. objectives,
required
if it covers a achievement course syllabi &
representative tests) textbooks
sample of the Consultation
behavior domain with
to be experts
measured  Making tests
specifications
Criterion – Prediction
Validity:
NATURE USE PROCEDURE STATISTICS
Prediction Hiring job Test scores are Pearson
from the test applicants, correlated correlation
to selecting with other
any criterion students for criterion Regression
situation over admission measures
time interval
Criterion – Prediction
Validity:

GCAT GWA

Practice Final Exam

Self-esteem Intention to use drugs

Competencies Treatment Success


Criterion – Prediction
Validity:

Theoretical Exam Practical Performance

Post-test Work Performance


Remember….
Construct Validity:
NATURE USE PROCEDURE STATISTICS
The extent to Used for Correlate a new Pearson
which the test personality test correlation
with a similar
may be said to tests. earlier test
measure a Measures that as measured Factor
theoretical are approximately Analysis
construct or multidimensio the same general
trait. nal behavior
 Factor analysis
 Correlate
subtest with
the entire test
Convergent Validity:
NATURE USE PROCEDURE STATISTICS
The test Commonly Correlate a test with a Pearson
should for second test that correlation
correlate personality measures a construct
positively measures that should Regression
from theoretically be
variables it positively related to
is related to the construct
measured by the test
Divergent Validity:
NATURE USE PROCEDURE STATISTICS
The test Commonly Correlate a test with aPearson
should not for second test that correlation
correlate personality measures a construct
positively measures that should Regression
from which it theoretically be
should differ unrelated/negatively
related to the
construct measured by
the test
Convergent and Divergent
Validity:

Let us try this...

Hope and... Hopelessness


Faith
Depression
Optimism
Hair Color
Construct Validity:

Sample and Exercise….


Convergent and Divergent
Validity:

Let us try this...

Decision-making skills and...


Hopelessness
Intelligence
Numerical Skills
Weight
Age
Self-efficacy
Review: Determine the type of
validity

1. A scale measuring motivation was


correlated on a scale measuring
laziness, a negative coefficient was
expected.
Review: Determine the type of
validity

2. The 16 PF that measures 16


personality factors were
intercorrelated with the 12 factors of
the Edwards Personality Preference
Schedule (EPPS). Both instruments are
measures of personality but contain
different factors.
Review: Determine the type of
validity

3. The scores of Mike’s mental ability


taken during fourth year high school
were used in order to determine
whether he will be qualified to enter in
the college he wants to study.
Review: Determine the type of
validity

4. Mrs. Ocampo a math teacher before


preparing her test constructs a table of
specifications and after making the
items it was checked by her subject
area coordinator.
Review: Determine the type of
validity

5. The scores on the depression


diagnostic scale were correlated with
the Minnesota Multiphasic Personality
Inventory (MMPI). It was found that
clients who are diagnosed to be
depressive have high scores on the
factors of MMPI.
Validity:

Sample and Exercise….


What is Test Reliability?

Reliability is the consistency of scores across


the conditions of time, forms, test, items
and raters.
Different Types of Reliability?

1. Test-retest
2. Alternate/Parallel Form
3. Inter-scorer/Inter-rater Reliability
4. Internal Consistency
4.1. Split-Half
4.2. Inter-item
4.3. Coefficient Alpha
4.4. Kuder-Richardson
Test – Retest Reliability
Test – Retest Reliability
Alternate/Parallel Form
Reliability
Inter-scorer/Inter-rater
Reliability
Split – half Reliability
Split – half Reliability
Inter – item Reliability
Coefficient Alpha Reliability
Kuder-Richardson Reliability
Review: Determine the type of
reliability

1. The Work Values Inventory (WVI) was


separated into 2 forms and two set of
scores were generated. The two sets of
scores were correlated to see if they
measure the same construct.
Review: Determine the type of
reliability

2. Children’s moral judgment was


studied if it would change overtime. A
moral judgment test was administered
during the first week of classes then
another at the end of the first quarter.
Reliability:

Sample and Exercise….


REFERENCE:

Magno, C., & Ouano, J. (2010).


Designing Written Assessment for
Student Learning. Phoenix: QC
THANK YOU!!!

You might also like