Serious concerns at methodology of Brexit poll which found an increase in support for Irish unity

The Press Association report on the poll that you carried on your website contained no analysis of the serious bias in the poll’s methodology.

The poll was carried out by Deltapoll who to my knowledge have no experience of political polling in Northern Ireland.

Out of the 1,199 people sampled 292 voted to Leave in the referendum.

This is approximately 26% of the sample when 44% of voters in Northern Ireland voted to Leave. 301 people or 25% of the sample support the Alliance party which is grossly out of proportion to their support in any election in recent history.

Whilst imbalances in the sample are adjusted by weighting this can lead to significant distortions in the accuracy of the poll.

The fact that the polling company could not find a representative sample of respondents is clearly cause for concern.

Given that the poll was commissioned by a continuity remain pressure group I believe limited weight should be given to this poll and any media outlet republishing its conclusions should afford it suitable critical analysis, as the News Letter did with a poll arranged by Tory peer Lord Ashcroft that had similar findings in June (‘Call for transparency over poll showing strong united Ireland support,’ July 6 2018).

Alexander Redpath, Ulster Unionist councillor, Lisburn