How does the marking logic work for free text questions?
For free text questions, the author specifies how many answers are required (i.e. how many answer boxes should appear underneath the question) and must then define at least this number of answer ‘matches’. An answer match consists of one or more text strings (separated by commas). The program takes the first of the trainee’s answers and tries to find a match for it in the possible answers. If more than one possible match is found, the program takes the highest scoring match. It then considers this answer match ‘used’, and moves onto the next question comparing only with ‘unused’ answer matches. A match is found if the trainee’s answer contains at least one of the text strings defined within a given answer match.
