This study examined pre-election polls held during six Israeli political campaigns from 1969 to 1988. There was a rapid growth in the space devoted to such polls by 15 daily newspapers.
Another important change was in sponsorship; in 1969, 95% of the polls were either unidentified or were sponsored by political parties, while by 1988 this figure had dropped to 22%, the rest being sponsored by newspapers. Over this time span, the reporting of the methodology became more complete, though it fell short of the reporting provided in the US. For example, the method of interviewing is reported in 52% of the reports in Israel, and 62% in the US. One of the more interesting findings is the strong positive correlation (0.51) between comprehensive reporting of methodology (measured by the number of methodological deficiencies reported) and the accuracy of the poll ( p < 0.01) . This is consistent with the notion that those who report more fully on the limitations of their methodology are more careful in their survey procedures.