Digitising Examination Scripts before Sending Them to Cambridge for Marking

Mr Murali Pillai asked the Minister for Education (Schools) whether it is feasible to scan or digitise examination scripts before sending them to Cambridge Assessment in the United Kingdom for marking so as to better deal with incidents of scripts being misplaced or stolen.

Sir, each year, about 900,000 scripts are sent to Cambridge Assessment for marking. Cambridge Assessment has shared that this is the first time examination scripts were stolen and it is taking this unfortunate incident very seriously. Cambridge Assessment has assured the Ministry of Education (MOE) and the Singapore Examinations and Assessment Board (SEAB) that they will review their processes to ensure the security of examination papers and scripts and safeguard the integrity of our national examinations. This includes a full scrutiny of the courier services offered by its suppliers.

I thank Mr Murali Pillai for his suggestion of using technology to reduce the risk of physical movement of examination scripts. Indeed, SEAB is studying the option of scanning and marking answer scripts electronically. This would potentially reduce the risks, such as theft or misplacement, associated with transporting hardcopy examination scripts. However, given the large volume of answer scripts and number of subject papers offered each year, the implementation of the scanning and electronic marking of answer scripts would have to be carefully considered.

I thank the hon Senior Minister of State for his explanation. I have a supplementary question. In relation to the case of the 238 GCE "A" level Chemistry examination papers that went missing or were stolen, rather, may I ask how did the measure taken by MOE, to assess the candidates by giving them an aggregate score or to allow them to re-sit their exams, square against the need to be fair to the other candidates, and make sure that they are not disadvantaged?

The option of allowing a re-sit can be understood as being fair. It was a full examination and these were candidates that did not have the opportunity to have their submissions marked.

The issue, I suppose, is really to look at those candidates who had a presumptive grade assigned. If you look at that subset, the technique that was used is something that has been used and the model has been developed quite robustly.

One can imagine that, on a regular basis every year, there are candidates who, through illnesses or unfortunate circumstances, are unable to submit every component of an exam. In this case, it is only one out of four components of the exam which went missing.

Between MOE, SEAB as well as Cambridge Assessments, we have not just a single model but a number of mathematical and statistical models derived to predict, I suppose, or triangulate the final performance of a candidate. This is something we have developed over time. The three bodies involved did their analyses on this occasion and also backchecked the analyses with respect to the outcome for these candidates. It was felt that this was a reasonable proxy. But the choice was then given to the candidates should they wish to re-sit the paper. In the process, none of the other candidates, who had their scripts properly marked and went through the process, were in any way disadvantaged.