# Item Analysis

Topics: 3, Duodecimal, 1 Pages: 8 (1687 words) Published: December 10, 2012
Item Analysis

* Both the validity and reliability of any test depend largely on the characteristics of the items. Ferguson (1989) says that high validity and reliability can be built into the instruments in advance through item analysis.

* In item analysis one determines the difficulty and discrimination indices of each test item and in cognitive test construction procedures especially in those that employ the multiple- choice type of format item difficulty and item discrimination are often used as statistical criteria for the selection and refinement of test items. In cognitive measures, item difficulty is defined as the percentage of students selecting the keyed response while item discrimination is referred to as the extent to which a test differentiates among examinees with respect to some criterion.

The U- L Index Method

* This technique was advanced by John Stocklein (1957) and is appropriate for test whose criterion is measured along continuous scale (e.g. scholastic rating, job ratings, performance records, achievement test scores) and whose individual item is scored right or wrong and positive or negative.

* The technique employs the following steps:
1. Score the test.
2. Arrange the papers from highest to lowest.
3. Separate the top upper 27 percent and bottom/ lower 27 percent of the cases. Suppose that you had try- out sample of 50 cases, would need the first 14 and the last 14. 4. Prepare a tally sheet such as the one seen below. Get the number of the examinees from the upper and lower groups who got the item correct for each of the items.

Item Number| Upper 27%N= 14| Lower 27%N= 14|
1| | |
2| | |
3| | |
| | |

5. Convert the frequencies to proportions.
6. Compute the difficulty index of each item using this formula. Dif.In= (PU+PL)/ 2
Where:Dif.In= difficulty index
PU= Proportion of the upper 27 percent who got the item correct
PL= Proportion of the lower 27 percent who got the item correct
Difficulty Index
0.91- above very easy
0.76- 0.90easy
0.26-0.75moderate or average
0.11-0.25difficult
0.10- below very difficult
7. Compute the discrimination index of each item using the formula: Dis.In= PU- PL
Where: Dis.In= discrimination index

* Items with a moderate or high discrimination index are very acceptable. Moderate discrimination index have Dis.In values which ranges from 0.20- 0.40 while high discrimination index have values ranging .41 and above and low discrimination index have values from 0.91 and below. These have limited acceptability. ITEM ANALYSIS WORKSHOP

Assume that you have constructed and administered a 50- item test to a class of 50 students. After correcting their papers, the results are as follows:

Pupil Number| Score| Pupil Number| Score| Pupil Number| Score| Pupil Number| Score| Pupil Number| Score| 1| 30| 11| 26| 21| 19| 31| 21| 41| 16|
2| 33| 12| 25| 22| 26| 32| 28| 42| 11|
3| 49| 13| 32| 23| 36| 33| 30| 43| 47|
4| 48| 14| 34| 24| 38| 34| 15| 44| 46|
5| 29| 15| 22| 25| 39| 35| 12| 45| 43|
6| 31| 16| 41| 26| 40| 36| 43| 46| 38|
7| 42| 17| 45| 27| 49| 37| 49| 47| 20|
8| 43| 18| 35| 28| 19| 38| 16| 48| 19|
9| 47| 19| 46| 39| 18| 39| 28| 49| 30|
10| 28| 20| 50| 30| 17| 40| 38| 50| 31|

Step Number 1Order the pupil’s scores, placing the highest score on top and continuing sequentially until the paper having the lowest score is place at the bottom

Pupil Number| Score| Pupil Number| Score| Pupil Number| Score| Pupil Number| Score| Pupil Number| Score| 20| 50| | | | | | | | |
3| 49| | | | | | | | |
27| 49| | | | | | | | |
37| 49| | | | | | | | |
4| 48| | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | |...

Please join StudyMode to read the full document