Update on methodology used in Yachad poll of British Jews

Recently I expressed concerns about the methodology of the Yachad poll. My concern was that the 72 ‘seeds’ in the ‘snowball’ category of the sample were mainly recruited by the advisory group of the survey, six of whom (out of ten) are Yachad signatories with three of the rest being left-leaning. The further concern was that this would bias the 496 respondents recruited by the ‘seeds’.

A guest post by Jonathan Hoffman

Recently I expressed concerns about the methodology of the Yachad poll.

My concern was that the 72 ‘seeds’ in the ‘snowball’ category of the sample were mainly recruited by the advisory group of the survey, six of whom (out of ten) are Yachad signatories with three of the rest being left-leaning. The further concern was that this would bias the 496 respondents recruited by the ‘seeds’.

To test these concerns, the lead author of the Report (Stephen Miller, Emeritus Professor of Social Research at City University), has kindly supplied an analysis of the results by the ‘snowball’ half of the sample and the rest of the sample separately. The results are set out below together with Professor Miller’s comments.  For a few responses the ‘snowball’ half is indeed more critical of Israel (eg for the statement ‘The Israeli government is constantly creating obstacles to avoid engaging in peace negotiations’) but overall – as Professor Miller says – there is no evidence of any systematic difference between the two sub-samples with respect to doveishness.  However the principle still remains: a poll needs to be seen to be independent and having a sponsor recruit around half the respondents (directly and indirectly) does not exactly encourage that perception.

snowball
More from Guest/Cross Post