In the past, tests were often constructed manually. Nowadays, digitally constructing a test may be more obvious. The (smart) software creates a balanced exam in no time at all, and at the same time human errors are eliminated. Have you already constructed your test digitally? Read more about the advantages of digital constructing a test in this blog!
In the manual construction of a test, the test developer or test constructor(s) decides, with the help of the test matrix, which questions are to be included in the test. The basic assumption here is that one test is equivalent to the next when they test basically the same thing. In addition, it is logical that important subjects should be included in each test.
These aspects make the construction of tests a labour-intensive and error-prone task. Which items (questions) will eventually be included in the test remains more or less dependent on the values and preferences of the test developer. Even if the test template is followed, it can happen that some subjects are not covered. This can make a test unintentionally predictable for candidates and you want to avoid that at all costs.
When a test is constructed digitally, it is drawn automatically from the database (item bank) on the basis of various criteria (relevant subject(s), number of questions, level of questions, etc.) which correspond to the test matrix. These criteria are drawn up in consultation beforehand. Each test is then guaranteed to have the same composition. The advantage is that it allows you to make very refined selections (at the lowest level in the test matrix). After all, the software does the work. The items in the bank must of course have been determined and checked, and must fit the test matrix.
See the diagram below for a simplified example. On the left is the structure of the database and on the right is the number of questions drawn. The number of questions per subject has a minimum (left-hand side) and a maximum (right-hand side). The structure corresponds to the structure of the test matrix. The number of questions also corresponds to the test matrix and/or is a further refinement thereof. In the example below, each test has 16 questions, so a minimum of 16 and a maximum of 16. All subjects and cases are part of each test taken. Each part subject is covered at least once. In addition, some part subjects recur in each test, in this case 1.1.3. The cases each deal with a subject A, B or C and contain, depending on the subject, 3 or 4 questions.
Constructing a test manually means searching through (separate) files to view items. It can happen that you accidentally select the wrong item. By choosing to construct a test digitally you not only save a lot of time, but at the same time you minimise the risk of mistakes and you guarantee objectivity in the selection of items.
When constructing a test digitally, you cannot only ensure that certain questions, in view of their importance, are included in every test, but you can also exclude certain questions from being included in a single exam. For example, because they overlap or because one question contains the answer to another question.
By digitally (automatically) drawing tests from a database, you can optimally control a weighted construction according to the test matrix. This guarantees that all tests are comparable and of equal quality. Constructing tests digitally also reduces the chance of errors and means a considerable saving of time. Finally, it also enhances the objectivity of the test.
Would you like to know what functionalities the Optimum Assessment Platform offers with regard to developing tests? Leave your details in the contact form below or contact us without any obligation on 013 – 528 63 63. You can also follow Optimum Assessment on LinkedIn: