Document Type

Conference Proceeding

Conference Track

Marketing Research/ Demographics/ Consumer Behavior

Publication Date



When a marketing research project is undertaken in a single language the potential for bias through misrepresentation of those who do not speak the language is always an inherent risk. That bias can greatly increase in a market with a significant population that does not speak that language. In practice there are a small number of U.S. markets that can be significantly impacted when addressing this concern. The author identifies the conditions under which this bias can be so great as to skew results. The author extends this to show how the bias can be a significant factor in marketing research decisions. Any discussion of modification of sample frame is inherently dangerous because it requires a judgment as to exclusion of people from the sample frame. But for some products, particularly for single language products such as Universities, radio stations, and print publications, to not make such a modification can seriously alter the overall study results. The author then proposes a basic model for correcting general market population estimates to allow for only including the population that speaks the language of interest. Through analysis of U.S. Census data the author demonstrates by example how this bias could impact the sample frame for research for a major market in the U.S. Lastly, the author discusses the implications of these findings and discusses the conditions under which the proposed population adjustment would be beneficial.

Copyright Statement / License for Reuse

Digital Commons@Georgia Southern License

Included in

Marketing Commons