This is the gist of my recent talk to high officials and faculty members of the University of the Cordilleras, in Baguio City, on the prerequisites, and implications, of establishing one’s own survey capacity. As an advocate of local opinion polling, I address it to other institutions also.
The capacity to do surveys is a powerful support to research in all social sciences and their applications: sociology, political science, economics, business administration, public administration, psychology, anthropology, development planning, etc. Applied research generally entails data collection. A university that does surveys competently and independently is strategically positioned for research leadership.
Surveys and democracy. Opinion polling flowers in a democratic setting, where the common tao is respected and appreciated. Openly disclosed surveys help empower the citizenry by exposing their collective voices to all.
The key prerequisite to successful polling is an atmosphere of freedom—freedom of citizens to express themselves, freedom of researchers to approach respondents on freely chosen research topics and to freely analyze the data, and freedom to disseminate survey findings to the general public.
Opinion polling tends to be threatening to traditional politicians whose power does not emanate from genuine public trust. A university engaged in survey work is not only research-oriented but also committed to democracy. Democracy should be fostered at both national and local levels. Metro Manila cannot serve as a political barometer for the entire country.
For instance, the scientific way to know the feelings of the Bangsa Moro on autonomy versus independence, or of Kapampangans on the performance of their governor, or of Negrenses about the need for agrarian reform, is for independent institutes in their communities to do their own polls. The Philippine exemplar is the Bohol Poll of the Divine Word University, in Tagbilaran City, which completed its 16th annual survey this year.
In the United States, many academic centers do opinion polls, including election surveys, in their own states and cities. It is normal that private universities, rather than government-run ones, are more vigorous in such research, due to the importance of political independence. Thus, in the Philippines, no state college or university can be expected to lead in survey research; there would be too much pressure on them from politicians.
Setting the agenda. The only thing that a survey institute controls is the research agenda. Its task is to ask people a list of questions, and record their answers open-mindedly. It doesn’t tell people what to answer; it is not engaged in publicity or public relations. It doesn’t deliberately ask leading questions; actually, such questions are quite easy to spot.
The agenda should be socially-relevant. Questions on the general quality of life and the performance of local governments are always interesting. I strongly recommend taking the challenge of predicting elections, since it is the worldwide litmus test of survey quality; by the way, predicting a local election is usually easy. Surveying the popularity of collegiate courses and schools is parochially useful, but won’t establish a reputation.
Don’t be put off by complaints about questions not asked; take them as suggestions for future agenda. Challenge critics who insist on coverage of their pet issues to do, or to commission, their own polls.
The book-ends concept. The two critical parts of a survey are (a) the questionnaire design and (b) the analysis of the data; whoever does these is the survey author. The components in-between these “book-ends”—namely sampling, field interviewing, and data processing—are not rocket science; a university can learn to do them as easily as it can set up courses in nursing or computer science. An institution is the author of a survey even if it outsources all of the inner components to service providers.
A university is survey-capable once its own staff competently design questionnaires and analyze survey data. They should reject self-incriminating or otherwise threatening questions. They should limit conclusions to what the data solidly support.
Data archiving. Raw survey data are historical documents, and should be archived (typically in electronic form) for the future. Like books, they are not depletable through use. If its data get lost, a survey is impossible to redo, at any price—it’s literally priceless.
As archived data accumulate, new surveys get more meaningful. For instance, the decline in English proficiency of Filipinos became visible only because data of an earlier survey on it had been archived.
Survey economics. Use efficient sample sizes; doing as little as 300 interviews is standard for a local survey. Two rounds of 300 respondents each, at different points of time, generally produces more information than a single round of 600 respondents.
Many universities abroad earn income from surveys; oftentimes survey centers are financially self-sustaining. An omnibus survey is one with multiple components and sponsors; its non-sponsored questionnaire space can be used to run pro bono items for public disclosure.
Warning: Never, never use field survey operations to solicit funds or to market anything. It will turn respondents off to all surveys, not only your own but also those of others.
Professionalism. Always observe the rights of survey respondents, sponsors, and researchers, as found in codes of professional ethics. In particular, I recommend the code of the World Association for Public Opinion Research.
* * *
Contact SWS: www.sws.org.ph or mahar.mangahas@sws.org.ph.
