Market research can be best understood as the study of the behavior of what is being observed over time to establish patterns (trends) through quantitative means so to extrapolate likely future behavior (typically selling and buying behavior).  At its best the survey effort should be directed to both buyers and sellers in a market with all questions developed to arrive at a quantitative response. 

Market research begins with identifying the appropriate population and the % of the population most likely to deliver responses that are representative of the whole.

The survey should be carefully constructed to eliminate the bias of the researcher as well as the bias of the survey population, that is, it should be highly quantitative in nature. 

A one to ten scale survey is qualitative, and while useful, should be backed up with quantitative responses.  This tool has high uncertainty. That is, the rankings are subjective.  One person’s two is another person’s five.

Polling that is yes and no surveys, are qualitative and while useful, should be backed up with quantitative responses. This tool has high uncertainty and should take into account the likelihood of respondents giving misleading responses.

Quantitative surveys: medium to low uncertainty and must take into account how the population was selected, the survey constructed, the bias of the analyst, the bias of the subjects and the bias of the client receiving the results. 

Uncertainty should always be factored into the analysis, that is, the methodology of the population (how it was set up) should be considered in that it may not be representative of the whole; also, what you didn’t ask, the % of answers assumed to be representative of the whole, how the questions were constructed, the bias of the analyst/researcher, the degree to which noise could be filtered out (other analyses and secondary data), the degree to which the bias of the subjects was filtered due to gathering an appropriate number of data, as well as the assumed responses of those who did not answer or who were not included. 

Even the most stringent and objective analysis can be undone by the market researcher’s own desire to see a different result and/or the researcher’s desire to please the client. The goal is and should be to provide an unbiased result that accurately represents the market as well as the participants in the market. The goal is to inform and aid the client in objective decision making.