A test item analysis calculator computes per-question statistics—difficulty index, discrimination index, and distractor efficiency—from a table of student responses. You enter each student’s total score and their answer for each item; the calculator returns values that tell you whether a question was too easy, too hard, or failed to separate high performers from low performers.
Formula
The two core statistics are simple proportions. For each test item:
- Difficulty index (p-value) = (number of correct answers in high group + number correct in low group) / (total number of students in both groups).
- Discrimination index (D) = (proportion correct in high group) − (proportion correct in low group).
Here, nh and nl are the sizes of the top and bottom scoring groups (commonly the top 27% and bottom 27% of all test takers), and ch and cl are the numbers of students in those groups who answered the item correctly. The difficulty index ranges from 0 (no one got it right) to 1 (everyone got it right). The discrimination index ranges from −1 (low group outperformed high group) to +1 (perfect separation).
Worked Example
Suppose 30 students took a test. You rank them by total score, then take the top 10 as the high group and the bottom 10 as the low group. For one question, 8 of the 10 high-group students answered correctly, and 3 of the 10 low-group students answered correctly.
Difficulty index: (8 + 3) / (10 + 10) = 11 / 20 = 0.55.
Discrimination index: (8/10) − (3/10) = 0.8 − 0.3 = 0.5.
The calculator returns p = 0.55 and D = 0.50. A p-value near 0.5 is often considered ideal for a norm-referenced test, and a discrimination index above 0.3 is generally taken as good.
Assumptions
- Students are divided into high and low groups based on total test score, not on the item itself.
- The groups are of equal size; typical practice uses the top and bottom 27% of the distribution, but 10–15 per group is common for small classes.
- The index treats all correct answers equally—no partial credit.
- The calculations assume you have already identified the high and low groups; the calculator does not rank students for you.
How to Use the Results
Read the difficulty index as a percentage: 0.55 means 55% of the students in the two groups answered correctly. General guidelines for a multiple-choice exam:
- p between 0.3 and 0.7: acceptable difficulty for most classroom tests.
- p above 0.9: too easy; the item may not discriminate.
- p below 0.2: too hard; check for ambiguous wording or missing prerequisite knowledge.
- D above 0.3: good discrimination—high scorers got it right, low scorers got it wrong.
- D between 0.1 and 0.3: marginal; revise the item or the distractors.
- D below 0.1 or negative: the item is not working. A negative D means low-scoring students were more likely to answer correctly, which usually signals a flawed question or a miskeyed answer.
For a 100-item test, you would run this for every item and then review the items flagged as too easy, too hard, or poorly discriminating.
Distractor Analysis
Beyond the difficulty and discrimination indices, a full item analysis also looks at each wrong answer option (distractor). The calculator above gives you the two main indices, but you can manually check distractor efficiency by counting how many students in each group chose each distractor. A good distractor attracts more low-group students than high-group students. If a distractor is chosen by no one, it is not functioning—replace it. If a distractor is chosen more often by the high group than the low group, it may be a second correct answer or a clue that misleads the better students.
Common Mistakes and Misconceptions
- Using the whole class instead of high/low groups. The discrimination index requires two groups; if you use the entire class, you get the overall proportion correct, which is just the difficulty index.
- Mixing up the direction of D. A positive D is good; a negative D is bad. Some people mistakenly think a high D means the item is too hard—it does not.
- Assuming p-value means statistical significance. In item analysis, “p” is the proportion correct, not a probability from a hypothesis test.
- Using unequal group sizes without adjusting. The formula assumes equal group sizes. If your high and low groups differ, use the proportion correct in each group rather than raw counts.
- Ignoring the 27% rule. For very large classes, the top and bottom 27% gives the most reliable discrimination estimates, but for small classes (like 20–30 students), using 10 per group is common and acceptable.
How to Choose a Calculator or Do It by Hand
You can compute these indices by hand with a spreadsheet or a simple calculator. The steps are:
- Rank all students by total test score.
- Select the top 27% and bottom 27% (or a fixed number like 10 each).
- For each item, count how many in the high group got it right, and how many in the low group got it right.
- Compute p and D using the formulas above.
When choosing an online calculator, make sure it lets you input group sizes and correct counts, not just raw response strings. Some tools accept a full response matrix and compute everything automatically, but the underlying math is the same. For a quick check, the calculator on this page gives you the two key numbers instantly.
Limitations of the Classical Indices
The classical difficulty and discrimination indices are simple and widely used, but they have limits. They depend on the ability of the students in your sample—an item is only “easy” relative to the group that took it. They also assume that the total score is a valid measure of ability. If the test itself is poorly constructed, the discrimination index can be misleading. More advanced methods, like item response theory (IRT), model each item’s characteristics independently of the sample, but they require larger sample sizes and more complex software. For most classroom settings, the classical indices are sufficient to flag problematic items.
Frequently Asked Questions
What is a good discrimination index? A discrimination index above 0.3 is generally considered good. Between 0.1 and 0.3 is acceptable but may need revision. Below 0.1 or negative indicates the item is not functioning well.
What does a difficulty index of 0 mean? It means no student in the high or low group answered correctly. The item is too hard for your group, or it may be miskeyed. You should review the question and the answer key.
How many students do I need for item analysis? For stable estimates, at least 20–30 students total is recommended. With fewer than 10 in each group, the indices are noisy. For high-stakes tests, larger samples are needed.
Should I use 27% or a fixed number like 10? For large classes (over 100), 27% is standard. For small classes, use a fixed number like 10 or 15 per group to have enough data. The key is that the two groups are clearly different in total score.
Can I use item analysis for a short quiz? Yes, but with caution. A 10-question quiz with 15 students will give rough estimates. Use the results to spot obvious problems, not to make fine distinctions.
What if a distractor is chosen more by the high group? That distractor may be a second correct answer, or it may contain a clue that only the better students notice. Review the item and consider revising or replacing that option.
Related
Formula used
in n | Number of students | - | 30 | 2 in nh | Number in high group | - | 10 | 1 in nl | Number in low group | - | 10 | 1 in ch | Correct in high group | - | 8 | 0 in cl | Correct in low group | - | 3 | 0 out p | Difficulty index (p-value) | - | (ch + cl) / (nh + nl) | 2 out d | Discrimination index (D) | - | (ch / nh) - (cl / nl) | 2
Each line is one input (in, sel, list) or one result (out: label, unit, formula, decimals). How the calculators work
Method: the standard formula for this calculation with the constants and conversion factors written on the page, rounded as each result states. See method and constants, the disclaimer or report an error.
More in this section
- QPP Calculator Canada: How It Works & How to Use It
- PPP Calculator World Bank: How It Works & How to Use It2,000.00
- PPP Calculator: India vs USA Cost Comparison1,000.00 USD
- PPP Calculator USD to INR: How It Works2,335 INR
- PPP Calculator India vs UK: Real Cost Comparison103.63 GBP
- PPP Calculator India vs Dubai: Real Cost Comparison8,954.58 AED