Evaluating public opinion polls: Designing a model for library instruction

Purpose - The purpose of this paper is to design and describe a model of instruction for librarians to effectively teach evaluation skills for public opinion polls. Design/methodology/approach - The paper show a coherent framework for instruction was identified through a survey of the literature. Learning objectives or essential core questions were identified by looking at codes and standards from five major organizations: National Council on Public Polls, Council of American Survey Research Organizations, World Association of Opinion and Marketing Research, American Association for Public Opinion Research, and the European Society of Marketing and Research. The objectives were incorporated into the aforementioned structure. Findings - The paper finds from nine documents produced by these organizations 15 elements were identified and grouped into nine essential questions: purpose of the poll, sponsor of the poll, polling organization, questions asked, order of the questions, who was polled, how were the interviews conducted, date of the poll, statistics offered to substantiate accuracy. A consistent framework was identified from an earlier work by Dodd. Research limitations/implications - The paper shows it is virtually impossible to provide a comprehensive list of all the possible ways that opinion polls can be misrepresentative. These are only a list of essential elements and questions necessary for an evaluation of polls. Practical implications - The public opinion poll continues to have a prominent role as a source of information in the societies of the world. As such, it is important that librarians as well as users are conversant in how to evaluate and place into context this type of information. Originality/value - This paper continues a dialogue within the profession about the impact of numbers as an information source. Librarians must become more conversant with numbers and, along with users, savvy consumers.