Scribd
Upload
2 views

Reliability

Uploaded by

emmelh.delacerna
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
2 views

Reliability

Uploaded by

emmelh.delacerna
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 37

RELIABILI

TY
Beryl Anne B.
Daquipa, RPm
refers to the
consistency of
RELIABILI scores obtained
by the same
TY person when re-
examined with
the same test on
different
Is an index of
reliability, a RELIABILI
proportion that TY
indicates the ratio COEFFICI
between the true
score variance on a ENT
test and the total
variance
TOTAL
VARIANCE

True VarianceErrorVariance
• true differences,• variance from RELIABILI
• stable other factors
outside the TY
variable tested
COEFFICI
ENT
CLASSICAL TEST
THEORY
all the fators
associated with MEASURE
the process of MENT
measuring some ERROR
variable, other
than the variable
being measured.
Systmatic
• source of error in measuring
a variableError
that is typically
constant or proportionate to
what is presumed to be the MEASURE
true value of the variable
being measured
MENT
Random ERROR
• source of Error
error in measuring a
targeted variable caused by
unpredictable fluctuations
and inconsistencies of other
variables in the measurement
process
SOURCES
OF ERROR
VARIANCE
• TEST
CONSTRUCTIO
• item
Nsampling/conten
t sampling
2. TEST
ADMINISTRATION
• Test environment
• Test taker variables
• Examiner-related
variables
3. TEST SCORING
AND
INTERPRETATION
• scorers and
scoring
IMPORTANCE OF
RELIABILITY
• Advancement of
psychological science

• Interpretation of test
scores in practical
Methods
of
Estimatin
g
1. TEST-
RETEST
• RELIABILITY
Time-sampling
reliability
• correlating pairs of
scores from the TIM TIME
same people on two E1 2
different
administrations of
1. TEST-
RETEST
• RELIABILITY
Time interval: less
than 6 months
• correlated using
Pearson r or TIM TIME
Spearman rho E1 2
1. TEST-
RETEST
RELIABILITY
Limitations
• changes
between
conditions of TIM TIME
two testing E1 2
administration
1. TEST-
RETEST
• RELIABILITY
Carryover effect:
occurs when the first
testing session
influences the results
of the second session
TIM TIME
• Practice effect: scores E1 2
on the second test
administration are
2.
PARALLEL/ALTERN FOR
ATE FORMS M1
RELIABILITY
• Item Sampling
Reliability
• Compare two FOR
equivalent versions M2
of a test that
measure the same
attribute
2.
PARALLEL/ALTERN FOR
ATE FORMS M1
RELIABILITY
• correlated using
Pearson r or
Spearman rho
FOR
M2
STEPS IN PERFORMING
PARALLEL/ALTERNATE
FORMS RELIABILITY
• Two forms are administered to
the same group of people
(Same day or different time)
• find the reliability
2.
PARALLEL/ALTERN FOR
ATE FORMS M1
RELIABILITY
Limitation
•
have to create two
forms of the test
(most rigorous and FOR
burdensome M2
assessment)
3. INTERNAL CONSISTECY
• Inter-item Reliability
• Assesses the correlation between
multiple items in a test that are
intended to measure the same
construct
3. INTERNAL CONSISTECY

a. Split-half
Reliability
• correlating two
pairs of scores
obtained from
equivalent halves
of a single test
administered
STEPS IN PERFORMING
SPLIT-HALF RELIABILITY
• Divide the test into equivalent
halves
• Calculate Pearson r between
scores on the two halves of the
test
SPLIT-HALF RELIABILITY
a. Spearman-Brown Formula
• Derived from Spearman Brown,
estimates how many more items
are needed in order to achieve the
target reliability
INTERNAL CONSISTECY
b. Kuder-Richardson 20
Formula
• KR20
• determine the inter-item
consistency of dichomtomous items
(primarily those scored right or
wrong)
INTERNAL CONSISTECY
c. Coefficient Alpha
• Cronbach's Coeffcient alpha
• Two halves of the test have
unequal variances (non-
dichotomous)
INTERNAL CONSISTECY
c. Coefficient Alpha
4. INTER-
RATER
RELIABILITY
• degree of
agreement or
consistency
between two or
more scorers with
regard to a
particular
4. INTER-
RATER
• RELIABILITY
Kappa Statistic:
assess the level of
agreement among
raters in normal scale
• Cohen's Kappa: to
know agreement
among two raters
• Basic Research:
.70 to .90
(lower, not
reccomendable
)
• Clinical Setting:
.90-.95
THE NATURE OF THE
TEST
• Homogenous vs
Heterogeneous Items
• Dynamic vs Static
Characteristics
WHAT TO DO IF
RELIABILITY IS
LOW?
• Increase the number of
items
• Factor and Item Analysis
THANK YOU
FOR
DoLISTENING
you have any questions for
me?

You might also like