To the Editor:
While we read with great interest your front-page story about our June-issue banking survey [May 10, "SunTrust Takes Top Spot in 'Consumer' Survey"], we would like to clarify a point raised in the last paragraph in the story. The paragraph begins:
"Some consultants warned that the survey was not scientific, and that those who participated may have done so because they had a gripe."
Someone else is quoted in the same paragraph, suggesting that the survey is not the scientific equivalent of Consumer Reports' product testing.
Our survey is indeed a scientific survey of our readers, designed and analyzed by our social scientists on staff. Using an aided-recall technique, all respondents are taken through a standardized set of questions about their experiences with their banks. By obtaining large sample sizes with the 20 banks, we were able to compare our readers' experiences and satisfaction with a high degree of statistical precision.
All the banks in our survey, even those on the bottom of our rankings, were viewed positively on average by their customers, a fact that argues against a negative bias. The banking survey was mailed to a random sample of our readers as one section of our annual questionnaire, which also contains questions on cars and more than 20 other products as well. This omnibus approach is a method we use to limit response bias.
Further, every year we conduct a parallel, separate validation study of every section on the questionnaire to see if those who do respond are more likely to have a negative point of view, and we found that they do not.
Charles Daviet
Director, survey research/information services
Consumers Union
Yonkers, N.Y.
