Editor's note: Lee Slurzberg is president of Lee Slurzberg Research, Inc., Fort Lee, New Jersey.

As a result of two decades of interviewing Hispanics in the United States, our firm has learned some things about the advantages and disadvantages of telephone vs. door-to-door interviewing in this market. This article talks only about the method of interviewing and excludes other parameters such as assuring that the Spanish language questionnaire is appropriate for the specific sample of Hispanics.

The growing interest in niche marketing over the last five years means more advertisers are concerned about the Hispanic market. Concurrently, interest in interviewing Hispanics has increased. For those of Hispanic descent, this means that if you have a Spanish surname and have a listed home telephone number, you have a greater chance of being contacted by a telephone interviewer than a non-Hispanic. Remember, Hispanic surnames are not generally deleted from a typical regional or national telephone sample, so English speaking Hispanics are usually included in typical telephone studies. "English speaking" could mean monolingual in English or bi-lingual.

A 1990 study by the National Center for Health Statistics on "Aids Knowledge and Attitudes of Hispanic Americans" interviewed 1 out of 5,300 Hispanics but only 1 out of 10,500 non-Hispanic whites. Thus, Hispanics had twice the odds of being in the study as compared with non-Hispanic whites.

A typical telephone-listed household could be contacted (dialed) 10 or more times a year by a research agency, but someone with an Hispanic surname could be contacted 15 to 20 times.

Our experience indicates that Hispanics can be interviewed by trained Spanish interviewers in the home or on the telephone, so the choice is usually a function of budget and concern about the attitudes and habits of the market segment without telephones.

Telephone sampling generally uses the Census list of 12,500 Spanish surnames as the frame. This list is in the public domain and is used by sampling firms and directly at our firm and others. There are less than five million residential listings with Spanish surnames. The sampling frame can be expanded by using RDD sampling in high density Hispanic areas (HDHA) to maximize inclusion of newly listed Hispanics and those Hispanics who do not have Spanish surnames. Approximately 20% of Hispanic women marry non-Hispanic men and therefore are excluded from Spanish surname samples.

Telephone advantages

1. Telephone sampling in the Hispanic market, as in the general market, permits greater sampling dispersion. Clustering, as used with in-home probability sampling, is not required. This difference is most apparent in Los Angeles, the largest Hispanic market. The Los Angeles-Long Beach MSA includes over 100 different directories. All of these might be used in drawing a sample of Hispanic surnames, but covering 100 primary sampling units (PSU's) in a door-to-door sample in that market would be extremely difficult and prohibitively expensive.

2. Quality control, in terms of the consistency of interviewing, seems to be better since all our interviewing is done from a central location. Having the interviewing for the four or five most-used markets (typically, Los Angeles, New York, Miami, San Antonio, and Chicago) done from one location tends to maintain consistency of handling the questionnaire.

3. Responses are less likely to be influenced by other members of the household. It is sometimes impossible to isolate a selected respondent for an in-home personal interview. All the "classic" field manuals instruct the interviewer to attempt to conduct the interview without others in the family listening. This good rule is hard to follow in door-to-door interviewing in many Hispanic and African-American PSU's.

4. Callbacks on not-at-homes, busy signals, or no answers are economical--even more efficient than the initial dialing. The list of numbers to be called back tends to exclude the business numbers, fax machines and modems, so it is more efficient. This effort contributes to statistical reliability. The higher the completion rate (the lower the non-response error), the more reliable the statistics.

5. A telephone study can be fielded more quickly than an in-home study involving probability samples. It is not as affected by weather and does not require as high a ratio of supervisors to interviewers.

6. The lower cost of telephone studies compared with the cost of in-home interviewing is an overwhelming advantage.

Telephone disadvantages

1. There are high numbers of Hispanics in all major markets without a telephone. "Owning a telephone" is different than telephone "availability," a term used by the Department of Commerce in its American Housing Study, which asks, "T may have to clarify something with you after checking my work. Is there a telephone on which you can be reached?" Their definition includes a telephone number at work, at a neighbor's or friend's house where they can be reached. This study did not employ a Spanish language questionnaire, but allowed "Spanish speaking field representatives (to) conduct an interview in Spanish when required."

Further, approximately 40% of the Spanish speaking Hispanics in an ongoing personal, in-home, omnibus in New York report that they do not own a telephone.

Our own data show that those without telephones are more likely to be younger. not fluent in English, and from Central or South America. Those without telephones also tend to have different media habits. They spend more time with Spanish TV. Differences also appeared in some personal care product categories.

2. Flash card can't be used.

3. Quality of responses may be affected by Hispanic perception of the telephone.

4. Interview length is relatively limited.

In-home advantages

1. The sampling frame includes Hispanics without Spanish surnames and/or without a telephone.

2. In-home studies permit use of English and one or two Spanish translations to describe different products categories or segments. Flash cards can be used for this purpose and to show brand lists, materials, product descriptions. For example, samples of fabric can be used to identify corduroy and denim. Even with flash cards, it may be necessary for the interviewer to read the words to illiterate respondents.

3. Face to face interviews generate better rapport between interviewer and respondents. Hispanics feel more comfortable and at ease with identifiable Hispanic interviewer. Respondents tend to be more hospitable to interviewers. The same in-home interview conducted among Anglos and Hispanics will tend to take longer among Hispanics than among Anglos because of the need to "socialize." Therefore, there are fewer terminates in the middle of the interview.

4. A longer interview is possible - usually involving a co-op fee.

5. Where necessary, cigarette brand, type, flavor and length can be verified by observation. Telephone ownership, makes of appliances (subject to recall error on the telephone) can all be verified.

In-home disadvantages

1. Door-to-door interviewing is much harder to control. For example, the inclusion of respondents without telephones prevents telephone verification of interviews with those respondents. Many respondents - especially in New York apartment houses, projects - may not have usable mail boxes, a fact that limits post card verification of interviews.

Telephone number matching has limited value because several families in the same building unit may use the same telephone.

2. Because of necessary clustering, Hispanics in low density Hispanic areas (who may be more acculturated - more likely to be living "in the high income suburbs") have less chance of being included in a door-to-door survey.

3. The cost is significantly higher than telephone studies and may be the major disadvantage for most consumer marketing research studies. The cost of door-to-door callbacks increases geometrically while the cost of telephone callbacks is lower.

Specific PSU's may require bodyguards or chaperones to accompany the interviewers.

Summary

On balance most consumer goods marketers can effectively use telephone studies in the Hispanic market. They work for ongoing tracking studies where changes over time are of greater concern to marketing management than absolute levels. They work for basic attitude and usage (A&U) studies and for a wide variety of research information needs.

Consider in-home interviewing when your product/service is known to, or supposed to, operate differently among that segment of the Hispanic population without telephones, but with a permanent address. For example, if your product is international money orders, would purchase habits be different between those with and without telephones. Is that difference likely to affect your marketing strategy?