About question metrics
When students do assignments online, Pearson collects anonymous data that is used to rate each question according to the difficulty and the time required to answer. When there is sufficient student data, this data is aggregated to provide statistics for the question.
If question metrics are available for your course, you see these statistics for each question when you view the questions for an assignment. Lists of questions include icons indicating the difficulty of the question. Move your mouse over the icon to see a tooltip that tells you the difficulty level, median time spent by students on the question, and the percentage of students who answered correctly on the first try.
Difficulty ratings are represented by the number of bars in the difficulty icon:
1 bar = easy
2 bars = moderate
3 bars = hard
4 bars = very hard
In accounting courses, the difficulty icon is replaced by an icon () that indicates the median amount of time that students took to answer the question. The amount of fill in the icon increases with the amount of time the question is likely to take your students. Each 5 minutes of fill corresponds to 5 minutes of working time. For example, indicates that the question will take approximately 20 minutes. If the clock icon is completely filled, the question will take an hour or more.
Questions with samples of more than 100 student results are used to generate question metrics. Data collection began in 2011 and is collected daily. Metric data is updated twice each year, in June/July and December/January.
|Rating||Deviations relative to average||Explanation|
Students with a skill level more than 2 standard deviations below average ability have a 50% chance of answering correctly.
Students with skill level between 1 and 2 standard deviations below average ability have a 50% chance of answering correctly.
Students with skill level no more than 1 standard deviation below average ability have a 50% chance of answering correctly.
|Very hard||Students with skill level that is average or greater than average ability have a 50% chance of answering correctly.|
These metrics are based on Item Response Theory (IRT) and in particular the Rasch model. For more information, ask your Pearson representative for copies of the paper on the theory and algorithms of IRT.