Reliability
What is Reliability?
• Reliability refers to the consistency of the
scores obtained – how consistent they are
for each individual from one administration
of an instrument to another and from one
set of items to another.
How do we determine if test is
reliable or effective?
By using this following methods:
• split-half method
• Kuder-Richardson formula
→KR-20
→KR-21
Reliability Interpretation
Excellent reliability; at
.90 and above the level of the best
standardized tests
Very good for a
.80 - 90 classroom test
Good for a classroom
test; in the range of most.
.70 - 80 There are probably a few
items which could be
improved.
Somewhat low. This test
needs to be
supplemented by other
measures (e.g.. More
.60 - 70 tests) to determine
grades. There are
probably some items
which could be improved.
Suggests need for
revision of test, unless it
is quite short (ten or
fewer items). The test
.50 - 60 definitely needs to be
supplemented by other
measures (e.g., more
tests) for grading.
Questionable reliability.
This test should not
.50 or below contribute heavily to the
course grade, and it
needs revision.
Split-half method
In this method the test is administered once and
score two equivalent halves of the test.
To split the test into halves that are equivalent, the
usual procedure is to score the even-numbered
and the odd-numbered test items separately. This
provides two scores for each student.
• To estimate reliability, Spearman-Brown
Prophecy formula is used.
Students Odd (x) Even (y)
1 4 4
2 2 2
3 4 3
4 2 2
5 3 4
• To find the reliability of the whole test we
must use the Spearman-Brown
Prophecy formula
→ rot=(2 roe)/(1+roe)
• Find the Σx, Σy, Σxy , Σx2 & Σy2 to solve
the reliability of the odd and even item
test.
x y xy x2 y2
4 4 16 16 16
2 2 4 4 4
4 3 12 16 9
2 2 4 4 4
3 4 12 9 16
Σ = 15 Σ = 15 Σ = 48 Σ = 49 Σ = 49
rxy= (n)(Σxy)-(Σx)(Σy)
√[(n)(Σx2)-(Σx)2][(n)(Σy2)-(Σy)²]
rxy= (5)(48)-(15)(15)
√[(5)(49)-(15)²][(5)(49)-(15)²]
rxy= 240-225
√(245-225)(245-225)
rxy= 15
√(20)(20)
rxy= 15
20
rxy= 0.75
Find the reliability of the original test
using the formula:
rot=(2 roe)/(1+roe)
rot= 2(0.75)
1+0.75
rot= 1.5
1.75
r = 0.86
Kuder-Richardson Formula
• Administered the test once. Score the
total test and apply the Kuder-
Richardson (KR) formula.
• The KR-20 is applicable only in
situations where students’ responses
are scored as right or wrong, true or
false, and yes or no type.
KR-20 formula estimates of reliability
provide information whether the degree
to which the items in the test measure is
of the same characteristic, it is an
assumption that all items are of equal in
difficulty.
• Another formula for testing the internal
consistency of a test is the KR- 21
formula, which is not limited to test
items that are scored dichotomously.
Students Score (x) x²
1 10 100
2 8 64
3 6 36
4 7 49
5 9 81
n=5 Σ = 40 Σ = 330
s²= n(Σx²)-(Σx)²
n(n-1)
s²= 5(330)-(40)²
5(5-1)
s²= 1650-1600
5(4)
s²= 50
20
s²= 2.5 The Standard Deviation
X̄ =Σx
n
X̄ =40
5
X̄ =8 The Mean
Solve for the reliability coefficient
using to Kuder- Richardson 21
formula:
KR21= K/K-1[1-(x̄ (K-x̄ )/K•S²)]
KR21= 10/10-1[1-(8(10-8)/10(2.5))]
KR21= 10/9[1-(8(2)/25)]
KR21= 1.1[1-(16/25)]
KR21= 1.1[1-(16/25)]
KR21= 1.1(1-0.64)
KR21= 1.1(0.36)
KR21= 0.40
Items/Questions
Students 1 2 3 4 5 Total Scores
1 1 1 1 1 0 4
2 1 0 1 1 0 3
3 1 0 1 1 1 4
4 0 1 0 1 0 2
5 1 0 0 0 0 1
Total
Correct 4 2 3 4 1 14
response
p 0.8 0.4 0.6 0.8 0.2
q 0.2 0.6 0.4 0.2 0.8
pq 0.16 0.24 0.24 0.16 0.16
p=Σp/n Σpq=0.96
q=Σq/n
pq=Σp•Σq
K=5
Σpq=0.96
s²=1.36
KR20=?
KR20= K/K-1[1-(Σpq/s²)]
KR20= 5/5-1[1-(0.96/1.36)]
KR20= 5/4(1-0.71)
KR20= 1.25(0.29)
KR20= 0.36