Module-3

Item Analysis

Item–Test Correlation
Item–test correlation is a statistical technique used in psychological and educational
testing to examine how well each item (question) of a test is related to the total score of the
test. It helps in understanding whether an item is measuring the same ability or trait that the
entire test is intended to measure.

Meaning (In Easy Language)

Item–test correlation tells us:
How strongly a single question is connected with the overall performance on the test.
If a person scores high on the total test and also answers a particular item correctly, the item
is considered good. If a person scores high on the test but answers the item incorrectly, the
item may not be suitable.

Definition
Item–test correlation refers to the correlation between the score on an individual test item
and the total test score. It indicates the consistency of an item with the whole test.

Purpose of Item–Test Correlation

The main purposes are:
1. To evaluate the quality of individual items
2. To identify good and poor items
3. To decide which items should be retained, revised, or removed
4. To improve the reliability and validity of the test
5. To ensure that all items measure the same construct

How Item–Test Correlation Is Calculated

 A correlation is calculated between:
o The score of one item, and
 o The total score of the test
 Usually, Pearson’s correlation coefficient (r) is used.
 Often, the item score is not included in the total score. This is called corrected
item–test correlation, and it gives a more accurate result.

Interpretation of Item–Test Correlation

Correlation Value (r) Interpretation

0.40 and above Very good item

0.30 – 0.39 Good item

0.20 – 0.29 Acceptable item

Below 0.20 Poor item

Negative value Very poor item (should be removed)

 Higher positive correlation means the item fits well with the test.
 Low or negative correlation means the item may be confusing or irrelevant.

Example
In an intelligence test, if an item has an item–test correlation of 0.45, it means:
 Students who score high on the test usually answer this item correctly.
 The item is a good indicator of intelligence.
If an item shows a negative correlation, it means:
 High scorers are answering it incorrectly.
 The item may be wrongly framed or misleading.

Importance of Item–Test Correlation

Item–test correlation is important because it:
1. Helps in item analysis
2. Improves internal consistency of the test
3. Ensures uniformity among test items
4. Increases the accuracy of measurement
5. Helps in constructing standardised tests
Uses in Psychological Testing
Item–test correlation is widely used in:
 Achievement tests
 Intelligence tests
 Personality inventories
 Attitude scales
 Psychological assessments

Difference Between Item–Test Correlation and Item–Item Correlation

Item–Test Correlation Item–Item Correlation

Item correlated with total score One item correlated with another

Shows consistency with whole test Shows similarity between two items

More useful for item selection Less useful for test construction

Conclusion
Item–test correlation is a very important method in psychological testing. It helps test
developers ensure that each item contributes meaningfully to the test. Items with high
correlation are retained, while items with low or negative correlation are revised or removed.
Thus, item–test correlation improves the quality, reliability, and validity of a psychological
test.

Item–Item Correlation
Item–item correlation is a statistical technique used in psychological testing to examine the
relationship between two individual items (questions) of a test. It helps in understanding
whether different items are measuring the same construct or concept.

Meaning (Easy Language)

Item–item correlation answers this simple question:
Do two questions in a test behave in a similar way?
If students who answer Item A correctly also tend to answer Item B correctly, then the two
items are said to be positively correlated.
Definition
Item–item correlation refers to the correlation coefficient calculated between the scores of
two separate items within the same test.

Purpose of Item–Item Correlation

Item–item correlation is used to:
1. Check consistency between items
2. Identify redundant items (items that are too similar)
3. Detect unrelated or faulty items
4. Improve the internal consistency of a test
5. Support test construction and refinement

How Item–Item Correlation Is Calculated

 Scores of one item are correlated with scores of another item
 Pearson’s correlation coefficient (r) is commonly used
 Can be calculated manually or using software like SPSS

Interpretation of Item–Item Correlation

Correlation Value (r) Interpretation

0.50 and above Very high similarity (items may be redundant)

0.30 – 0.49 Good relationship

0.20 – 0.29 Moderate relationship

Below 0.20 Weak relationship

Negative value Items measure different things or are faulty

Example
In a verbal ability test:
 Item 5 and Item 8 both test vocabulary
 If their item–item correlation is 0.42, it shows both items are measuring similar skills
If the correlation is –0.18, it suggests:
 One item may be confusing
 Items may be measuring different abilities

Importance in Psychological Testing

Item–item correlation helps in:
 Ensuring homogeneity of test items
 Removing irrelevant or misleading questions
 Avoiding unnecessary repetition of items
 Strengthening test reliability

Difference Between Item–Item and Item–Test Correlation

Item–Item Correlation Item–Test Correlation

Correlation between two items Correlation between an item and total test score

Checks similarity between items Checks contribution of item to whole test

Identifies redundancy Identifies weak items

Used less frequently Used more frequently

Advantages
1. Simple to understand
2. Useful in test construction
3. Helps maintain balance in a test
4. Improves overall test quality

Limitations
1. High correlation may indicate repetition
2. Low correlation may indicate poorly written items
3. Not sufficient alone—should be used with item–test correlation
 Module-4
Standardization Process of Test – I

Relationship between Reliability and Validity

(Long answer – easy language, exam-oriented)
In psychological testing and measurement, reliability and validity are two very important
qualities of a good test. Both are closely related, but they are not the same. To understand
their relationship clearly, we must first understand each term and then see how they are
connected.

Meaning of Reliability
Reliability refers to the consistency or stability of a test.
In simple words:
A test is reliable if it gives the same or similar results every time it is used under similar
conditions.
Example:
If an intelligence test gives similar IQ scores to a person when taken today and again after
one week, the test is reliable.
Key point:
 Reliability is about accuracy and consistency
 It answers the question:
“Is the test dependable?”

Meaning of Validity
Validity refers to the accuracy or truthfulness of a test.
In simple words:
A test is valid if it actually measures what it is supposed to measure.
Example:
If an intelligence test truly measures intelligence and not memory or general knowledge, then
it is valid.
Key point:
 Validity is about meaningfulness and usefulness
  It answers the question:
“Is the test measuring the right thing?”

Relationship Between Reliability and Validity

Reliability and validity are closely related, but they serve different purposes. Their
relationship can be explained through the following points:

1. Reliability Is Necessary for Validity

A test must be reliable to be valid, but reliability alone does not guarantee validity.
 If a test is not consistent, it cannot measure anything accurately.
 Therefore, validity depends on reliability.

🔑 Important statement for exams:

“A test cannot be valid if it is not reliable.”

2. A Reliable Test May Not Be Valid

A test can give consistent results but still fail to measure what it is intended to measure.
Example:
A bathroom weighing scale always shows 5 kg extra weight.
 It gives consistent readings → Reliable
 But it does not show the true weight → Not valid
This shows that:
 Reliability does not ensure validity
 Validity is a higher standard than reliability

3. Validity Includes Reliability

Validity includes reliability, but it goes beyond consistency.
 Reliability checks how consistent the test is
 Validity checks what the test measures
Thus:
 Reliability is a part of validity
 Validity is a broader concept
4. Difference in Focus
The main difference in focus explains their relationship:

Reliability Validity

Focuses on consistency Focuses on accuracy

Concerned with stability of scores Concerned with correctness of interpretation

Statistical concept Conceptual and statistical

5. Analogy to Understand the Relationship

Think of a dartboard:
 Darts hitting the same place every time but far from the center
→ Reliable but not valid
 Darts scattered around the center
→ Not reliable and not valid
 Darts hitting the center consistently
→ Both reliable and valid
This analogy clearly shows the relationship between reliability and validity.

6. Role in Test Construction

In psychological testing:
 Reliability is tested first
 Validity is established after reliability
Because:
 Without reliability, validity cannot be established
 Both together ensure a good quality test

Conclusion
Reliability and validity are closely connected but not identical concepts. Reliability refers to
the consistency of a test, while validity refers to its accuracy and truthfulness. A test must
be reliable to be valid, but a reliable test may not always be valid. Therefore, reliability is a
necessary but not sufficient condition for validity. Both are essential for constructing a
scientifically sound psychological test.