commit to user 68
tentative theory about why high and low scores should behave differently 2000: 110.
b. The Reliability of the Instrument.
The reliability of a test refers to consistency of the test score. Tuckman 1999: 160 states test reliability means that a test is
consistent
. Furthermore, he also explains some factors which contribute to the unreliability of a test as
follows: 1 familiarity with the particular test form such as a multiple choice question; 2 fatigue; 3 emotional strain; 4 physical conditions of the room in
which the test is given; 5 health of the test taker; 6 fluctuations of human memory; 7 amount of practice or experience by the test taker of the specific skill
being measured; and 8 specific knowledge that has been gained outside of the experience being evaluated by the test.
A test which is overly sensitive to these unpredictable and often uncontrollable sources of error is not a reliable test. Test unreliability creates
instrumentations bias,
a source of internal validity in an experiment Tuckman, 1999: 162. Ary 1985: 226 propounds that reliability is concerned with how
consistently we are measuring whatever we are measuring. Seeing the definition above, it can be said that in this study, a reliable
instrument would give the writer a consistent result. The writer uses the Kuder- Richardson procedures in order to get the reliability of the test. It is based on the
proportion of correct and incorrect responses to each of the items on a test Ary, 1985: 233.
commit to user 69
Here is the formula to count the reliability of the reading comprehension test with the form of objective test:
= − 1
1 −
∑ 5
Notes:
r
kk
: the coefficient of the reliability p
: the proportion of correct responses on a single item q
: the proportion of incorrect responses on the same
item S
t 2
: the variance of scores on the total test squared standard deviation
k
: the number of valid items on the test If r
o
is higher than r
t
, the item is reliable. From the reliability testing, it is obtained that r
o
0.937 exceeds r
t
0.361 for the level of significance α = 0.05. It
can be concluded that the test is reliable. Besides that, to know the reliability of the learning motivation
questionnaire, the formula is as follows: =
− 1 1
− ∑ 5
ĸ
∑ 5
m
Notes:
r
kk
: the coefficient of the reliability
k
: the number of valid items on the questionnaire S
i 2
: the variance of scores on the items S
t 2
: the variance of scores on the total questionnaire
commit to user 70
From the reliability testing of the questionnaire, it is obtained that r
o
= 0.970. It is consulted with r table for N = 30 at the level of significance
α = 0.05 which is 0.361. Because r
o
exceeds r
t
it can be concluded that the questionnaire is reliable.
3. Documentation