The University of Birmingham Railway Systems group provide intensive training courses to London Underground staff each year, and each year the learners are assessed using extended answer and essay questions delivered on ExamOnline. The e-examinations are delivered through a standard web browser, locked down to prevent access to other web sites or resources. ExamOnline provides specific support for essay based examinations, and the blind marking process is deemed both fairer and more efficient. The University of Birmingham use the same ExamOnline system to assess students on their MSc programme in Railway Systems Engineering and Integration (RSEI).

If you are involved in the assessment of medical knowledge, you may be interested to know that the Medical School at the University of Dundee now assess all undergraduate medical students using an on-screen, computer-marked test. You may be even more interested to know that the questions are not multiple choice, they’re short-answer free-text. And that the system was recently awarded the Scottish e-Assessment Award for Summative Assessment.

The ‘key features’ approach to assessment has been extensively used in postgraduate medicine to measure clinical decision-making skills. The essential element of a key feature question is that only the critical steps in the clinical decision making process are assessed, allowing more clinical situations to be tested in a shorter time period than more traditional case based question types. This project in the Medical School at the University of Keele is using FreeText Author to implement a computer-based assessment system which will facilitate the automatic marking of short-answer free-text key feature questions.

The Open University have used FreeText Author to develop computer-aided assessments for checking and providing instantaneous feedback on questions requiring short free text answers, typically a sentence in length. The questions were developed using software provided by Intelligent Assessment Technologies Ltd (IAT). This software uses natural language processing (NLP) techniques of information extraction, but an authoring tool is provided to shield the question author from the complexities of NLP. The IAT software was used with the Open University’s OpenMark e-assessment system, thus enabling students to be offered three attempts at each question with increasing feedback after each attempt. Feedback on incomplete and incorrect responses was provided from within the IAT authoring tool, using a flagging system developed during the project.