**NOTE**: Most data shown in the Analytics environment are not reliable for small amounts of data (students). Beware of this when making conclusions based on Analytics.

**Overview**

**Question analysis**

There are five values that are calculated for each question, which give information about the question. These values are all calculated for multiple choice questions. The p-, Rit- and Rir-values are also calculated for open questions.

**p**: Rate of difficulty. The p-value is the percentage of students that correctly answered the question. The higher the p-value, the easier the question is.

p = [number of students that answered this question correctly] / [total number of students]

**p'**: Rate of difficulty corrected for the guess correction. The p'-value is the percentage of students that know the question. This value is minimal 0,25 and maximal 0,9. If the p'-value equals 0,5 it can be said that a maximum discrimination between good and weak students is found.

p' = p - [ (1-p) / (number of options for this multiple choice question - 1) ]

**Rit**: Relation of the question to all questions in the test. The higher the Rit-value, the better the question fits the test. A value higher than 0,40 is very good, a value between 0,30-0,40 is good, but can be better, a value between 0,20-0,29 is a doubt and a value lower than 0,19 indicates that the question is bad. The equation is given below-
- Yg = average total score of students that got this question correct
- Yf = average total score of students that got this question incorrect
- Sy = standard deviation of the total score
- Sy = standard deviation
- p = p - number of points for the question
- q = (1-p)

**Rir**: Relation of the question to all other questions (so all questions except this one). This value measures the same as the Rit-value, but the Rir-value is more exact, while it is not compared to itself. A value higher than 0,40 is very good, a value between 0,30-0,40 is good, but can be better, a value between 0,20-0,29 is a doubt and a value lower than 0,19 indicates that the question is bad. The equation is given below.

**Rk**: Corrected guess effect. The normal guess effect of for example 25% for a 4-choice question with one correct answer is corrected for the amount the other alternatives are chosen. This way the corrected guess effect can be higher than the normal gambling chance.

For every question, based on the p-, p'-, Rit-, Rir- and Rk-values, a conclusion is made. This conclusion can be:

- Good question: All values are good.
- Easy question: The p- and p'-values are probably too high.
- Difficult question: The p- and p'-values are probably too low.
- Needs attention: The Rit- and/or Rir-values can be off. Look at which number is divergent.

A line is placed under a value if the value is divergent.

For multiple choice questions a different view is displayed than for open questions.

**Multiple choice questions**

An example of a multiple choice question in the question analysis section is found in the image below. This question is marked as an easy question due to the high p-value of 0,97. Because this is a multiple choice question, per choice it can be seen what percentage of the students chose the answer.

**Open question**

**needs attention**, due to the negative Rit- and Rir-value, which are underlined. This will mean that this question does not fit in the rest of the exam.

- Contributing: This is the standard option, the question contributes to the grades of the students.
- Bonus: The question will not be taken into account in the maximum score that can be gained, but it can be a bonus for the students who got it correct.
- Full points: Every student will get the full (maximum amount of) points for this question.
- Disabled: The question and the points of the question will be taken out of the exam and will not be taken into account for the grading.

**Rank correlation**

**Grading alignment**

*WELL ALIGNED*when the averages are about equal or

*DEVIATION DETECTED*if the average values do not match. For this calculation, the number of questions reviewed is taken into account. As an instructor, you now can choose to review this question again. Ans Delft will not supply more data, information or advice on the grading of reviewers.

**Grading time**

**Objectives**

**Export and print analytics**

It may be useful to have the analytics offline or on paper. This is possible by the print option of the analytics.

In the top right corner of the Analytics page, whichever tab you are on, a button **Print** is shown, as can be seen in the first image of this article. By clicking this button, a new page will be opened including the analytics from the tabs **overview**, **question analysis**, and** rank correlation**. This page subsequently can be printed physically or to a .PDF format by printing the web page. This can be done for example by using the shortcut ctrl+P on a Windows computer and cmd+P on an Apple computer.

## Comments

0 comments

Please sign in to leave a comment.