Arbitron Overview of Ethnic Measurement and Reporting Enhancements.
January 27, 2006
Arbitron Inc. released an overview of the two ethnic measurement and reporting enhancements implemented in the Winter 2006 Radio Survey: Personal Race/Ethnicity and Language Weighting.
In addition, demonstration data from the Summer 2005 radio surveys in Los Angeles and Miami provide insights for customers to the potential impacts of these new enhancements.
Personal Race/Ethnicity
Race/Ethnicity will be classified on an individual level based on responses to questions asked in the diary. In Arbitron surveys prior to Winter 2006, everyone who filled out a diary was attributed the race/ethnicity of their household which was determined during the initial placement call.
Language Weighting
Language usage weighting provides sample-balancing controls to ensure that the proportion of Spanish-Primary and English-Primary Hispanics matches the characteristics of the market’s population. As demonstrated in past surveys, Spanish-Primary Hispanics show different radio listening patterns than English-Primary Hispanics.
For example, in the Fall 2005 Miami radio survey, the top five radio stations for Spanish-Primary Hispanics 12+ (12 years or older) were all Spanish-language formats. The top five radio stations for English-Primary Hispanics 12+ were English-language formats. Because the Hispanic population in Miami is two-thirds Spanish-Primary, the top five radio stations for All Hispanics 12+ are also the top five Spanish-language stations.
There are also differences in language usage between markets. In the Fall 2005 San Antonio radio survey, a market with a Hispanic population that is predominantly English-Primary, the top five radio stations for All Hispanics 12+ are more reflective of the top five formats for English-Primary Hispanics.
Demonstration Data, Los Angeles and Miami Summer 2005 Radio Surveys
Arbitron examined the impact on radio audience estimates by fielding a demonstration of these two enhancements in the Summer 2005 radio surveys in Los Angeles and Miami. Ratings results were compared between three different data sets:
· The Household-weighted report: Previous technique of weighting race at the household level;
· The Personal Race/Ethnicity-weighted report: New technique of applying weights for race at the individual/personal level, (not the household level). This technique will be used in 179 markets, and;
· The Language-weighted report: Weighted for both individual race and language usage. This combined technique will be used in 21 markets.
Highlights from the Demonstration Data for Los Angeles and Miami
· Personal Race/Ethnicity will better classify the individual diarykeepers resulting in more appropriately weighted samples and provide more reliable estimates of the ethnic composition of audiences to individual stations. With the new personal Race/Ethnicity procedures, the sample composition shifted slightly. In the LA sample, the percent of Hispanic went from 45 percent to 41 percent; Black went from 10 percent to 9 percent, and; Other, went from 45 percent to 50 percent. Similar results came out of the Miami sample.
· The impact of Personal Race/Ethnicity will be more pronounced in markets with more mixed-race households.
· Individual markets will show variance in results from applying Personal Race/Ethnicity in the 179 metros with Black and/or Hispanic controls, and Language Weighting in the 21 language controlled markets.
These markets are listed at:
Personal Race/Ethnicity:
http://www.arbitron.com/radio_stations/comingup_WI06race_markets.htm
Language Usage Weighting:
http://www.arbitron.com/radio_stations/comingup_WI06luw.htm
A more detailed review of the findings can be found at:
http://www.arbitron.com/downloads/ArbitronEthnicMeasurement.pdf.
History of Language Usage Weighting
“In November 2002, Arbitron told the industry that we would add weighting by language preference as one more incremental enhancement in the services we offer our subscribers. And we are delivering this enhancement on the schedule that we announced in January 2003,” said Owen Charlebois, president, Operations, Technology, Research and Development, Arbitron Inc. “More importantly, now that we are evaluating race and ethnicity at a personal level and weighting by the language preference of Hispanic respondents, our subscribers can have more confidence in the stability of our radio audience estimates. These improvements also allow stations and advertisers that target specific segments in the Hispanic community to reach those consumers more effectively.”
