VALIDITY

AND
RELIABILITY
Asih Rosnaningsih
CONTENT
1. Validity
2. Reliability
 3

VALIDITY
Validity comes from the word “valid” which means the extent
to which the accuracy and accuracy of a measuring
instrument in carrying out its size function (Azwar 1986).

In addition, validity is a measure that shows that the variable

being measured is really the variable that the researcher wants
to study (Cooper and Schindler, in Zulganef, 2006).
 4

Validity

Criterion-
Content Face Construct Predictive
based
 TYPES OF VALIDITY
1. Content : Content validity refers to the representativeness of our measurement regarding the
phenomenon about which we want information.
2. Face : Face validity is closely related to the notion of content validity and refers to the familiarity of
our instrument and how easy it is to convince others that.
3. there is content validity to it.
4. Construct: It refers to degree to which the research adequately captures the construct of interest.
Construct validity can be enhanced when multiple estimates of a construct are used.
5. Criterion-related: Criterion-related validity refers to the extent to which tests used in a research
study are comparable to other well-established tests of the construct in question.
6. Predictive: Predictive validity deals with the use that one might eventually want to make of a
particular measure. Does it predict performance on some other measure?
Presentation title
INTERNAL VS. EXTERNAL VALIDITY
• Internal Validity
Internal validity refers to the extent to
which the results of a study are a
function of the factor that the
researcher intends.
1. Language background
2. Language learning experience
3. Proficiency level
4. Participant mortality
5. Other factors.
6
• External Validity
All research is conducted within a
particular setting and using a specific
set of characteristics. With external
validity, we are concerned with the
generalizability of our findings, or in
other words, the extent to which the
findings of the study are relevant not
only to the research population, but
also to the wider population
Presentation title 7

TESTING VALIDITY
To test the validity of this using the SPSS program.
The testing technique that is often used by researchers
to test the validity is using the Bivariate Pearson
correlation (Pearson Moment Product). This analysis
is done by correlating each item's score with the total
score. The total score is the sum of all items. Question
items that are significantly correlated with the total
score indicate that these items are able to provide
support in revealing what you want to reveal.
Valid if r-count > r table (2-sided test with sig. 0.05)
then the instrument or question items have a
significant correlation with the total score (declared
valid).
 8

RELIABILITY
Reliability comes from the word reliability. The definition of
reliability (reliability) is the consistency of measurement (Walizer,
1987).

Sugiharto and Situnjak (2006) state that reliability refers to an

understanding that the instruments used in research to obtain
information used can be trusted as a data collection tool and are
able to reveal actual information in the field.
 9

Reliability

Rater
Instrument
Reliability

Interrater Intrarater Equivalence Internal

Test-retest
Reliability Reliability of forms consistency
Presentation title 10

TYPES OF RELIABILITY
1. Rater
The main defining characteristic of rater reliability is that scores by two or more raters or between one rater at
Time X and that same rater at Time Y are consistent.
• Interrater reliability begins with a well-defined construct. It is a measure of whether two or more raters judge
the same set of data in the same way. If there is strong reliability, one can then assume with reasonable
confidence that raters are judging the same set of data as representing the same phenomenon. Intrarater
reliability is similar, but considers one researcher's evaluations of data, attempting to ensure that the researcher
would judge the data the same way at different times.

2. Instrument
Not only do we have to make sure that our raters are judging what they believe they are judging in a consistent
manner, we also need to ensure that our instrument is reliable. In this section, we consider three types of
reliability testing: test-retest, equivalence of forms of a test (e.g., pretest and posttest), and internal consistency.
Presentation title 11

TYPES OF RELIABILITY
2.1 Test-retest
• 2.1 Test-Retest. In a test-retest method of determining reliability, the same test is given to the same group of
individuals at two points in time. One must carefully determine the appropriate time interval between test
Administrations.
• 2.2 Equivalence of Forms. There are times when it is necessary to determine the equivalence of two tests, as, for
example, in a pretest and a posttest.
• 2.3 Internal Consistency. It is not always possible or feasible to administer tests twice to the same group of
individuals (whether the same test or two different versions). Nonetheless, when that is the case, there are
statistical methods to determine reliability split-half (odd-even numbers), Kuder-Richardson 20 and 21 (large
numbers of items), and Cronbach’s a (questionnaires with scales).
Presentation title 12

TESTING RELIABILITY
Criteria for Reliability
• If the alpha value > 0.7 means that the reliability is sufficient (sufficient reliability) while if the alpha > 0.80
this suggests that all items are reliable and all tests consistently have strong reliability.
• If alpha > 0.90 then the reliability is perfect. If the alpha is between 0.70 – 0.90 then the reliability is high. If
the alpha is 0.50 – 0.70 then the reliability is moderate. If alpha < 0.50 then the reliability is low. If the alpha is
low, it is possible that one or more items are not reliable.
• Click Analyze → Scale → Reliability Analysis.
Presentation title 13

ERRORS IN RESEARCH
• TYPE 1 AND TYPE 2 ERROR
A statistically significant result cannot prove that a
research hypothesis is correct (as this implies 100%
certainty). Because a p-value is based on
probabilities, there is always a chance of making an
incorrect conclusion regarding accepting or
rejecting the null hypothesis (H0). Anytime we
make a decision using statistics there are four
possible outcomes, with two representing correct
decisions and two representing errors.
The chances of committing these two types of
errors are inversely proportional: that is, decreasing
type I error rate increases type II error rate, and
vice versa.
Presentation title 14

SUMMARY
Validity is an evolving complex concept because it relates
to the inferences regarding assessment results. Focusing on
the consequences of the inferences made implies that they
should be appropriate and adequate.
Reliability reflects consistency and replicability over time.
Furthermore, reliability is seen as the degree to which a test
is free from measurement errors.
THANK YOU
asihrosna@gmail.com