An item though consists not only of the question itself, but can also contain further information on the question, so-called meta information or metadata. They describe an item more precisely, and facilitate and simplify the selection and compilation of items for test authors. The better part of the metadata needs to be entered by an author.
All in all, more than 20 metadata attributes, according to the learning object metadata, are available for further item specification. If you are authorized to edit the item, you can modify the meta data under General, Lifecycle, Rights, Share / Authors, Educational, Item analysis and Technical. Please note though, modifications to Educational and Item analysis should only be made by an experienced author.
On top of the detail view you can find the question preview as well as information on the question settings. Clicking the "Edit" button in the top right corner opens the test editor. More information on the test editor can be found in the chapter "Creating Tests and Questionnaires", section "Test and Questionnaire Editor in Detail".
The metadata attributes can be found below the settings and are divided into 7 sections.
Meta Data
General
Contains information on categorization such as language or department as well as keywords. Coverage adds more details to the topic and narrows it down. Enter even more metadata with the additional information form. The range of subjects should cover the departments of your institution. Please contact your pool administrator if departments or subjects are missing.
Lifecycle
Specifies the version number of the item as well as its status of availability, such as "Draft" or "Unavailable".
Rights
Contains information on the author of the item, and whether the item holds a copyright. By default, 6 Creative Commons licenses are already available. Information on Creative Commons can be found in the Wikipedia and on www.creativecommons.org.
If more licenses are required, please contact your pool administrator.
Share/authors
Lists all authors as well as all the pools and groups with which the item was shared.
Educational
Information on the target audience of the item is provided by the level selection. Combined with the typical learning time, the average time needed to answer the question, the level facilitates the compilation of time-framed, knowledge level appropriate assessments.
Item analysis
Contains information on item analysis and the use of the item in tests. The Item analysis is a set of (statistical) methods, with which individual question items are evaluated and assessed pertaining to their suitability for knowledge measurement respectively assessment. Typical parameters are the difficulty index and the discrimination index.
The difficulty index expresses with a value between 0 and 1 how difficult it is to answer a question, and thus indicates how many individuals of a group of candidates correctly solve the question in relation to the maximum achievable score. The purpose of the difficulty index is to discriminate individuals with high characteristic values from those with low characteristic value. Therefore all items that could be solved by any individual or items that could not be solved by anyone, are useless (index value close to 0 or 1). Items with values close to 0 are too simple and do not distinguish between the performance of individuals and items with values close to 1 are too difficult. Please note that in a multiple choice question with 5 response options (4 distractors), there is a 20% probability that the answer was guessed. Items in the area of 0.4-0.9 or 40-90% are suitable for a good performance discrimination.
The standard deviation (of the item difficulty) expresses with a value between 0 and 1 the dispersion of individual scores on that item, thus indicating how widespread the responses were. If the test scores are distributed as a normal curve, one standard deviation comprises about 68% of the scores above and below the mean, while two standard deviations cover 95.5% of all values in the value distribution. The lower the standard deviation, the more "stable" the difficulty index on one or several test candidate universes, the larger, the more "unstable" it is. Items with a high standard deviation should therefore be selected with utmost care.
The discrimination index expresses with a value between -1 and +1 the ability of an item to discriminate candidates with good and poor knowledge of the material being tested. It provides an estimate of the degree to which an individual item is measuring the same thing as the rest of the items of the test. The discrimination index is therefore the most important parameter in the item selection process. It is calculated as the product moment correlation coefficient between student responses to a particular item and total scores on all other items on the test. In order to achieve a good performance discrimination items with distinct positive indices are required, if possible higher than or equal to .20, but certainly higher than .10. Items with a very low discrimination index do not contribute to differentiation, those with negative indices even run counter to the item selection process and should not be used in follow-up testing.
In addition, you can enter information on the number of distractors in the question and determine, whether the item is suitable for summative (evaluative), formative (diagnostic) or both types of tests. The system automatically determines if the item is already in use. You will find the exact number here. This will automatically increase when used in a test.
Technical
The technology metadata section provides information on the technical item editor. The format specifies the technical format of the item. You will also find information on the creation and last modified dates.
