Creating and Using e-Assessment
Being aware of the institutional and practical considerations is of course important, but the key issue to be resolved is that of creating the assessment tasks. As has already been mentioned, creating reliable and valid questions is a skilled task, and a lot more work is required from staff before the assessment is presented to students, especially when feedback is to be included for all of the incorrect answers in an objective test.
In common with all assessment, for e-assessment the assessment tasks should be aligned with the intended learning outcomes, so that students are able to display the extent to which they have met the outcomes for the course. Using the verbs from Bloom's taxonomy can help in the design of questions, using these to define the level of learning that are being tested.
Not all questions need to be written from scratch, and in the first instance it may prove useful to use tried and tested questions from other sources. Banks of objective questions are often associated with textbooks in further and higher education, and these can be used with permission from the academic publishers. Some of the subject centres of The Higher Education Academy (http://www.heacademy.ac.uk/3801.htm) are developing question banks, using the expertise of academic colleagues in departments across the UK. Examples of question banks are provided in the links section.
Of course departments can choose to write questions themselves. Zakrzewski and Steven (2003) suggest that staff should create their own question banks and that they should generate an extra 10% of questions each year to allow for some questions to be removed or used less frequently. This is especially important as curricula change and develop. Gipps (2003) suggest that for cost effectiveness, staff should create question banks because: 'the true costs involved mean that CAA is only really feasible for items that can be re-used' Gipps (2003: 27). Questions should be created collaboratively and tested before being made available to students to alleviate the possibility of writing overly simple questions or questions with subtle nuances that students may inadvertently pick up (Clarke et al., 2004).
The COLA project (Sclater and MacDonald, 2004) developed a Word template to help staff with item creation for multiple choice tests, including metadata to describe the item, developing the methodology that had been used successfully by the e3an project.
McAlpine (2004) has outlined a methodology for determining whether tests that are currently administered on paper could be administered online. Initial results show that the methodology has potential for discriminating between tests that can be easily administered online, and assessment items for which more research is required in terms of the use of "emerging technologies".
