Interference Validity of the Test
itself. To sum up, a test is called reliable when it is tested on a similar group in similar condition, the result remains the same.
Ary et al assumes that reliability concerns with how consistently a researcher measures whatever shehe measures 2002: 251. Moreover, they add
that the reliability won’t concern about the meaning and interpretation of the scores, for they are more associated with the theory of validity. A measuring instrument
can be reliable without being valid. Nevertheless, it cannot be valid unless it is first reliable. As an instance, a researcher decides to measure intelligence by determining
the circumference of the head. The measurements may be reliable consistent from time to time, yet this method will not be considered valid if the circumference of
the head does not correlate with any other criteria of intelligence nor is it predicted by any theory intelligence. Regarding this research, the instrument test that will
be conducted might be reliable, yet it cannot be valid unless the instrument correlates with the theory and function of preposition for and to or contribute the
functions of the prepositions for and to which have been applied in this research. In measuring the reliability of the test, the researcher uses one of the
methods, which is the split-half technique Sprinthall et al, 1991: 35. The split-half technique of reliability is one kind of internal-consistency reliability, in which a test
requires only a single administration of one form of a test Ary et al, 2002: 256. The test is split into two halves, and correlates the individual’s scores on the two
halves; the first half the odd-numbered items was labelled X and the second half even-numbered items was labelled Y. Thus, the researcher uses The Pearson
product-moment coefficient formula to determine the reliability of the test Sugiyono, 2010: 255. The formula can be seen as follows.
Where ∑ xy = sum of the x times y scores
∑ x² = sum of the x squared scores
∑ y² = sum of the y squared scores
r = Pearson
Then, to indicate the reliability of the whole test, the Spearman-Brown Prophecy formula is needed. The formula can be seen as follows:
r = 2r half-test 1 + r half-test
Where r = The Spearman-Brown’s coefficient of reliability
r half-test = The Pearson’s Correlation coefficient
According to Ary et al, researchers must inquire into the validity and reliability of the scores derived from instruments test used in a study and must
include this information in the researcher’s thesis as a repost. Nevertheless, if the data are not included with valid and reliable instruments, the readers might have
little faith in the results gained or in the conclusions based on the results 2002: 242.