Colegio de San Gabriel Arcangel

City of San Jose del Monte, Bulacan

College of Education

Program Bachelor of Elementary/Secondary Education

Course Code PROED 9

Description Assessment of Student Learning I

LEARNING ACTIVITY SHEET NO. 6

Name ________________________________________________ Score _________________

Date ________________________________________________ Section _________________

Topic Item Validation

Learning 1. Discuss the distinction between test reliability and test validity;
Objectives 2. Enumerate the different methods of establishing reliability of test;
3 Enumerate the different methods of establishing validity of test;
.
References Navarro, Rosita L; Santos, Rosita G; & Corpuz, Brenda B. (2017), Assessment of Learning I (3rd
ed):Lorimar Publishing Inc. Quezon City, Philippines
Reganit, Arnulfo Aaron R; Elicay, Ronaldo SP; & Laguerta, Cresencia C. (2010) Assessment of Learning
I: C & E Publishing , Inc. Quezon City, Philippines.
Cajigal Ronan M. and Mantuano, Maria Leflor D. (2014)). Assessment of Learning 2: Adriana Publishing
C., Inc. Manila & Quezon City, Philippines.
Sevilla, Consuelo G., Ochave, Jesus A., Punzalan, Twila G., Regala, Bella P., & Uriarte, Gabriel G.
(1992). Research Methods (Rev. ed). Rex Book Store. Manila, Philippines.

Discussion

After the item analysis and option analysis are done in the preliminary form of the test, good items with acceptable
keys and distracters are retained and included in the final form of the test. The final form of the test should be free
grammatical as well as typographical errors. The options of each item should be arranged chronologically or
alphabetically. The test questions/items should be arranged from easy to difficult based on difficulty index obtained by the
item. Administer the second try-out of the test for gathering of data purposes. Finally, after the final run, the test can now
be evaluated statistically in terms of its reliability and validity.

Interpreting Correlation Coefficient ( r )

The correlation coefficient ( r ) tells the direction of relationship between two variables, either it (X) positive variable or (Y)
negative variable. It has nothing to do with degree of relationship between variables.

Range of r Interpretation
Obtained
1.00 Perfect Positive Correlation
0.75 to 0.99 Very High Positive Correlation
0.51 to 0.74 High Positive Correlation
0.25 to 0.49 Moderately Positive Correlation
0.01 to 0.24 Very Small Positive Correlation
0 No Correlation
-0.01 to -0.24 Very Small Negative Correlation
-0.25 to -0.49 Moderately Negative Correlation
-0.51 to -0.74 High Negative Correlation
-0.75 to -0.99 Very High Negative Correlation
-1.00 Perfect Negative Correlation

Anybody who wants to interpret the result of the coefficient of correlation ( r )

Should be guided by the following reminders (Uriarte and Others, 1987).

1. The relationship between two variables (X and Y) does not necessarily mean that one is the cause or the effect
of the other variable. It does not imply cause-effect relationship.
2. When the computed r is high, it does not necessarily mean that one factor is strongly dependent on the other.
This is shown by height and intelligence of people. Making a correlation here does not make any sense at all.
On the other hand, when the computed r is small it does not necessarily mean that one factor has no
dependence on the other factor. This may be applicable to intelligence quotient (I.Q.) and grades in school. A
low grade would suggest that a student did not make use of his time in studying.
 3. If there is a reason to believe that the two variables are related and the computed r is high, these two variables
are really meant associated. On the other hand, if the variables correlated are low (though theoretically related)
other factors might be responsible for such small association.
4. Lastly, the meaning of correlation coefficient ( r ) just simply informs us that when two variables change may be
a strong or weak relationship taking place.

Evaluating the Test Reliability

Test reliability refers to the consistency, dependability, or stability of the test. If the students are given the same
twice on different occasions, the scores of the students do not differ significantly. Therefore the test is consistent,
dependable or stable. Below are different techniques of establishing test reliability:
1. Split-half Reliability. The most common technique of evaluating the reliability of the half-test is through the
odd-even split half technique. This is done by splitting the test into two, the odd-numbered items as one, and
the even-numbered items as the other. The student’s test paper has two total scores: 1) total score for all
odd-numbered items and 2) total score of all even-numbered items. If the test is 40 items, one-half of 40 is
20, therefore, there would be 20 items odd-numbered and 20 items even-numbered.
The scores of the students on the odd-numbered items (X variable) are correlated with their scores on
the even-numbered items (Y variable) (1) through the Pearson Product–Moment correlation or Pearson r and
the resulting statistic is the reliability coefficient of the half test. (2) To obtain the reliability of the entire
test, Correction Formula must be applied, which is the Spearman-Brown Prophecy Formula, also known as
the correction formula.

(1) Formula: Pearson r

NΣXY - ΣXΣY
roe = -------------------------------------------------------
√ {N (ΣX2 - (ΣX)2} {N (ΣY2 - ΣY2}

Where: N - number of cases (pair of scores) √ - Square root

(2) Formula: Spearman Brown Prophecy Formula (or Correction Formula)

2roe
rtt = ------------------------------

1+ roe

Where: rtt - is the reliability coefficient of the whole test

roe - is the reliability coefficient of half of the test

It can be noted further that the reliability of the half test is increased after applying the correction formula.

This is due to the fact that increasing the length of the test increases the reliability of the
same test.

2. Test-Retest Reliability. This is sometimes called the coefficient of stability. It is the consistency of the test
over-time. To calculate the coefficient, the test is administered twice to the same sample with a given time
interval. You may administer again the same test to the same group after an interval of two weeks. The
Pearson r is then calculated to determine the reliability of the test.
3. Alternate Forms or Parallel Form Reliability. The coefficient of equivalence is computed by
administering two parallel forms of the test to the same group of individuals. The same correlational method
is applied to compute for the reliability coefficient (Pearson r).

Computing the Split-Half-Reliability Coefficient

Using Pearson r and Spearman Brown Prophecy Formula are statistics used to determine the reliability coefficient
for half of the test and reliability coefficient for the whole test. The steps in computing the Pearson are as follows:
1. Draw a matrix (table) with 5 rows and the number of columns depends on the N or total number of cases (scores);
2. Enter the scores from the X variable in the first column (X);
3. Enter the scores from the Y variable in the second column (Y);
4. Square each score in the X variable column to get X2 and enter in 3rd column (X2);
5. Square each score in the Y variable column to get Y2 and enter in the 4th column (Y2);
6. Multiply the X variable score and Y variable score and enter the product in the 5th column (XY);
7. Add all the X variable scores to get the summation of X or ΣX ;
8. Add all the Y variable scores to get the summation of Y or ΣY;
9. Add all the X2 values to get the summation of X2 (ΣX2)
10. Add all the Y2 values to get the summation of Y2 (ΣY2);
11. Add all the XY values to get the summation of XY (ΣXY)
12. Follow the Formula of Pearson r (there are 7 steps)

For example:
 We will use the GIVEN Scores in English G-6. These are the scores of eleven 6th grade students. Using Split –
Half reliability technique, the purpose of using Pearson r is to get the reliability coefficient of half of the test only. Thus, the
computed correlation coefficient using Pearson r using Split-half reliability technique is to be treated further by means of
Spearman Brown Prophecy Formula (Correction formula) in order to the reliability index of whole test

X variable (Odd-numbered Item scores) 10 9 11 7 7 8 5 11

Y variable (Even-numbered item scores) 10 8 9 10 7 6 4 8

X (ENIS) Y (ONIS) X2 Y2 XY
10 10 100 100 100
9 8 81 64 72
11 9 121 81 99
7 10 49 100 70
7 7 49 49 49
8 6 64 36 48
5 4 25 16 20
11 8 121 64 88
ΣX=68 ΣY=62 ΣX2=610 ΣY2=510 ΣXY=546

Using Pearson r statistic to get the reliability coefficient of half of the test in English G-6.
NΣXY - ΣXΣY
Step 1 roe = ---------------------------------------------------
√ {NΣX2) - (ΣX)2} {NΣY2) - (ΣY)2}
8(546) - (68)(62)
Step 2 roe = ---------------------------------------------------
√ {8(610) - (68)2} {8(510 - (62)2}
4368 - 4216
Step 3 roe = ---------------------------------------------------
√ {4880 - 4624} {4080 - 3844}
152
Step 4 roe = -----------------------------
√ {256} {236}
152
Step 5 roe = --------------------------
√ 60416
152
Step 6 roe = --------------------
245.80
roe = 0.618 or 0.62
Step 7
Interpretation:
The computed coefficient (roe = 0.62) is the reliability index of half
only of the test in English G-6 using split –half reliability technique.

In order to obtain the reliability index of the whole test ( rtt), treat the roe =
0.62 further statistically by using the Spearman Brown Prophecy Formula
(or also called Correction Formula).

Using Spearman Brown Prophecy Formula to compute coefficient ( r ) 7of the whole test in English G-6.

Step 1 2roe
2tt = --------------------
1+ roe
Step 2 2(0.62)
rtt = ------------------
1 + 0.62
Step 3 1.24
rtt = ----------------------------
1.62
Step 4 rtt = 0.765 or 0.77
Interpretation:
The computed coefficient rtt = 0.77 is the reliability index of the whole
test in English G-6 using split-half reliability technique.

Evaluating the Test Validity

Test Validity is what the test purports to measure, or what the test intends to measure. If a test is intended to
measure what the students can do and cannot do in Algebra, supposedly, all test items/question should measure Algebra.
 1. Criterion-Related Validity. This is characterized by a prediction of relation to an outside criterion or
variable.
 Predictive Validity. The NCAE scores of students in the high school is correlated with college
performance. If the students’ score in the NCAE – Math, most likely the student would perform high in
math subjects in college, or they would enroll in commerce or engineering courses. The outside
criterion her is the student’s future performance in college which is being predicted in the NCAE-Math
score.
To estimate the predictive validity coefficient of the NCAE, the Pearson r is used.
Example: Correlation between NCAE-English scores and College grades in English

 Concurrent Validity. This also called congruent validity which is also criterion-related validity. The
test serves as to assess present status of the individual. Concurrent validity of the test is determined
by correlating the scores with the subjects” IQ in the Otis Lennon Mental Ability Test (OLMAT).
Those who scored high in the subjects likewise score high in the IQ test, OLMAT. Subject scores and
IQ scores are correlated using the Pearson r.
Examples: Correlation between Student-Made Test (SMT) scores and Teacher-Made Test TMT)
scores are correlated using Pearson r.

2. Construct Validity. Construct validity is also called concept validity for it involves discovering a positive
correlation between and among the variables/constructs that define the concept. Examples: (a) Correlation
between IQ scores of examinees and their self concept; (b) correlation between NAT scores of students and
the type of school where they are enrolled; and correlation between grades of pupils in all subjects and their
nutritional status.

Practice Exercises: Self Review

A. Describe each technique of establishing test reliability. Give your own example.
B. Enumerate the procedures in evaluating test reliability using the split half technique.
C. Using your own test data with N=10, practice the computation of Pearson r and Spearman Brown Prophecy
Formula using the split-half reliability technique.
D. What is the main difference between predictive validity and concurrent validity?

Test Activity
Direction: On a separate sheet of paper, answer the following exercises based on the discussion about test
validation. Pass a hard copy to your leader.

A. Explain the difference between reliability of test and validity of test.

B. All test are reliable but not all test are valid. Explain this issue.
C. Determine the reliability coefficient of the given test results in Math G-5 below using split-half technique. Find the
reliability coefficient of half of the test as well as the entire test.
GIVEN: Provide your own data (Take Home Quiz)
X variable - ONIS
Y variable - ENIS
Note: ONIS - means odd-numbered item scores; and
ENIS-stands for even-numbered item scores.

D. Determine the validity coefficient of the given SMT in Science G-7 below using concurrent validity method.
GIVEN:
X variable TMT scores 18 16 14 13 12 10 10 8 6 3
Y variable SMT scores 10 14 8 12 10 8 7 6 5 0
Note: TMT – refers to Teacher-Made Test in Science G-7 and SMT refers to Student-Made Test in Science G-7.

Prepared by:

JULIETA E. SICLON