A study tested the hypotheses (1) that reporting of methodological information needed to gauge the accuracy of public opinion poll stories in metropolitican daily newspapers improved significantly during the presidential election years from 1968 to 1980, and (2) that these newspapers were more likely to report this information in their own inhouse polls than in wire service and syndicated polls. Data were gathered from an analysis of all stories in the front sections of two Detroit, Michigan newspapers for a four-month period preceding the presidential elections from 1968 to 1980. Results supported both hypotheses. The findings suggest that the newspapers did a better job reporting poll stories over the years because of an increased awareness among journalists and journalism educators of the need to provide readers with information to evaluate public opinion poll stories. In addition, they suggest that the newspapers did a better job of reporting their own poll stories than wire service poll stories because they viewed inhouse polling as prestigious. (Author/FL)