Dear Stephen,
Thank you for your letter of 29 July on the use of statistics by Ministers in relation to the Test of English for International Communications (TOEIC). You were concerned that an estimate of the proportion of candidates wrongly accused of cheating was used by the Home Office with undue confidence.
While the estimates in question are not official statistics, we would expect Ministers to be clear about the limitations of all data and statistics, in line with the Code of Practice for Statistics. We also expect organisations to assure themselves that they can have confidence in figures used to inform decisions, including undertaking appropriate quality assurance.
Having looked at other investigations into this issue including the National Audit Office investigation[1] and evidence provided to the All-Party Parliamentary Group on TOEIC[2] we consider that there is considerable uncertainty in the one per cent estimate of false positives. This uncertainty has not always been conveyed clearly.
I am copying this letter to John Flatley, the Acting Chief Statistician for the Home Office.
Yours sincerely,
Sir David Norgrove
[1] National Audit Office (May 2019): Investigation into the response to cheating in English language tests
[2] APPG on TOEIC (2019): Report of the APPG on TOEIC, All-Party Parliamentary Group on TOEIC
Related Links: