Nielsen has used self-reports recorded in a paper diary for television audience measurement since the 1950s. Nielsen sends bilingual Spanish and English materials to TV Ratings Diary service households that had any indication of Spanish usage in the household. This has long been assumed to be the best practice, as anyone in the household can use the materials whether they prefer to read English or Spanish. However, the bilingual materials may be perceived as more complex by survey respondents, but there was no empirical data to support that position. As viewing choices have increased and television technology has evolved, respondents increasingly have difficulty accurately and completely recording all viewing information in a paper-based diary.
In response, Nielsen recently developed a “mailable meter”, a smaller self-installed television meter that captures tuning data. With this system the respondents complete a very simple viewing log in lieu of the more complicated viewing diary. Nielsen has been conducting an ongoing series of methodological tests to determine the viability of this concept. As part of the research and development of this new system, in July 2008 Nielsen conducted a field test with more than 400 households in a heavily Hispanic market. In order to maximize the benefit of a more technological solution, one must still examine the burden placed on respondents in following instructions to complete the survey. Since this study focused on how Hispanic households would react to this new methodology, respondents were given the option of what type of materials (English only or bilingual) they would prefer to receive. Therefore, respondents in that study who spoke Spanish during recruitment calling, or answered a question that indicated that they or someone in their home was a Spanish speaker, were asked an additional question about their preference for the language of the log and survey instructional materials. Respondents were advised that their entire household would be expected to participate in the study and had agreed to this prior to being asked about their preference for materials.
Increasingly in the survey industry, the concept of acculturation is coming to the forefront as an important consideration in survey design and methodology. Webster defines acculturation as “cultural modification of an individual, group, or people by adapting to or borrowing traits from another culture”. Padilla and Perez (2003) argued that social cognition, cultural competence, social identity, social dominance, and social stigma all play a role in the rate and degree of acculturation that an individual or group would experience. The process of acculturation is complex and not well appreciated by many researchers. However, when viewed from a survey participation perspective, the level of acculturation of the rapidly growing Hispanic respondent population is a critically important factor in how a person perceives information and thus how they react to survey materials. Though perhaps somewhat of an oversimplification, two of the most commonly used measures of acculturation are the language used at home and how well the respondent speaks English. Depending on the level of acculturation of a given respondent, there may be a positive or negative effect of providing English only, bilingual or Spanish only survey materials. Thus the language treatment given to a Hispanic target group must be an important consideration in survey planning.
A random digit dial (RDD) sample of 9,586 phone numbers was selected in the El Paso, Texas market. The Nielsen Call Center recruited 459 households that agreed to install meters and complete logs of their household’s viewing for one week. Of these 459 households, 431 were mailed survey packets. Of those 431 households, 280 respondents indicated that they or someone in their home speaks Spanish. Two hundred and thirty-seven of these 280 respondents were Hispanic and were asked a follow-up question on the level of Spanish use in their household (only Spanish, mostly Spanish, Spanish and English equally, mostly English or only English at home). These 280 respondents were asked their preference for language of viewing log and instructional materials, and mailing of materials was based on that response. In addition, a post survey follow-up questionnaire was mailed to all households that received meter packets to gather information about their survey experiences.
Summary of Findings
Surprisingly the results showed that nearly three out of four respondents (73%, 203 of 280) to this preference question requested English only materials. There was a predictably strong connection between the amount of Spanish use in the home and a request for bilingual materials. As shown in Table 1, more than 90% of respondents who spoke only Spanish asked for Spanish materials and 56% of those that spoke mostly Spanish asked for bilingual materials. Conversely, only 4% (6 of 134) of households that spoke an equal amount of Spanish and English, mostly English or only English requested bilingual materials. There were 43 non-Hispanic respondents that were asked their preference of materials, as we learned on the phone that there was at least some Spanish spoken in their home. Only two of these respondents requested bilingual materials.
Age did not appear to have much influence on the respondents’ choice of materials, as 64% of those with a householder under the age of 35 selected English only materials, whereas 68% of those with a householder age 35 to 49 and 71% of those with a householder over the age of 50 did so. This was despite the fact that 51% of those with a householder under the age of 35 stated that they spoke only or mostly Spanish at home and only 38% of those with a householder over the age of 35 did so.
The broad implication of this research may indicate a need to examine the methodologies one employs when sending survey materials to bilingual populations.
- To maximize the usability of the survey instruments when sending survey materials to bilingual populations don’t assume all bilinguals should receive bilingual materials or that all Spanish-speaking people should receive Spanish materials.
- Providing a choice of language of materials for the bilingual survey respondent is an important consideration that may improve response rates and data quality.
- In considering what special treatments one might need to give unacculturated Hispanics, consider the importance of language used to this group.
Some interesting insights were gained through the post survey follow-up questionnaire regarding their experiences with the materials. Among English only materials households only 4% reported that instructions were difficult and less than 7% would have preferred to have received bilingual materials. Among bilingual materials households: 66% reported looking at Spanish instructions more, while 17% looked at both languages equally and 17 % looked at English more. Also, 31% of bilingual materials households reported never looking at English instructions, and only 2% reported never looking at bilingual instructions. Only 2% of bilingual materials households reported that the instructions were difficult.