Test Score Analysis

Paste test scores from a spreadsheet to get the mean, median, standard deviation, frequency distribution, ranks and standard scores. The "Item analysis" tab uses student answers to find each question's difficulty and discrimination, and the KR-20 reliability.

How to use

1. On the "Scores" tab, paste the name and score columns into "Score list". One score per line also works.
2. Set the "Maximum score" and "Class width" to update the summary, frequency table and histogram. "Grade bands" counts the students in each grade.
3. On the "Item analysis" tab, paste the student answers and the key to get per-question statistics and answer-choice counts.
4. Use "Copy table" or "Copy summary" to move the results into a spreadsheet or document.

How it works

• Z = (score − mean) ÷ standard deviation and T = 50 + 10 × Z. The standard deviation divides by n by default; the settings can switch it to n − 1.
• Discrimination = difficulty in the top 27% − difficulty in the bottom 27%. A KR-20 near 1 means the questions measure one ability consistently.
• Names and scores are processed on this device and kept only in this browser. "Clear saved data" removes them.

Questions and answers

How are tied scores ranked?

By default tied students share a rank and the next rank is skipped (1, 2, 2, 4). "Ranking ties" also offers the average rank (1, 2.5, 2.5, 4) or no gaps (1, 2, 2, 3).

How is the percentile rank calculated?

(students below + half of the students with the same score) ÷ all students × 100. With 15 of 20 students below and no ties, it is 77.5.

What do difficulty and discrimination mean?

Difficulty is the share of students who answered correctly; higher means easier. A discrimination below 0 means the lower group did better, so check the key and the question.

Can I use a right/wrong table instead of answers?

Yes. Set "Input type" to "Right/wrong table (O/X, 1/0)". Answer-choice counts are not shown in this mode.