Editor's note: Michael Halberstam is president of Interviewing Service of America, Van Nuys, Calif.

Demographic sources, including the 1990 census, show there are now at least 7.3 million Asian- Americans living in the United States, a 110 percent increase over 1980. It's not a homogeneous group, either: The figure breaks into 30 cultural subgroups. Members of each group have their own opinions on contemporary issues, are making their own decisions on where their dollars will be spent, and are speaking their own language. These same census figures also show that these new Americans send more children through college, hold down more white-collar jobs, and possess more disposable income than any other ethnic group.

But in 1985, when we began exploring this emerging market, we found that the first rule of surveying Asian-Americans was understanding each culture and how it differed from working with Hispanics or any one of the 50 cultural groups we survey on a regular basis.

It was then that we devised the Asian Surname Database, a segmentation tool that we now use in the 55,000 to 60,000 surveys we complete each year, to differentiate each of the ethnic communities based on surname. (We also provide the database to other companies.) Before the availability of this database, ethnic sampling was done with Asians as a single bloc. No one differentiated between groups. Since Japanese are different from Koreans, and Koreans differ from Vietnamese, and so forth, researching these markets was usually done in English. This precluded about 80 percent of native-language Asian Americans, a staggering percentage to leave out of any supposedly projectable sample.

Eager participants

It was practices like these that led many companies down the wrong path in getting a foothold in the Asian-American market. But today we are able to effectively survey in 11 Asian languages and dialects including Cambodian, Cantonese, Hindi, Japanese, Korean, Laotian, Malay, Mandarin, Tagalog, Thai and Vietnamese.

One of the most astounding things we first learned in this business was these communities' interest in participating in surveys. As many of you who work in the marketing research business know, it is increasingly difficult to find willing respondents, but the Asian communities are different. With a refusal rate at approximately 40 percent of the industry norm, their desire to participate is driven by wanting to become a larger part of the American mainstream without losing their cultural identity.

This is why we use bilingual Asian employees, many of whom were born outside the United States, to survey respondents in their native language. These employees have a keen awareness for cultural, conceptual and idiomatic considerations and nuances when collecting data, translating questionnaires and moderating focus groups.

Companies who spend the time and effort respecting Asians and their culture, and communicate with them in their native language, find they are rewarded with enduring brand loyalty.

The use of native language is also important because many of the groups we have tracked are still primarily native-language speakers. This is most prevalent in the Vietnamese, Cambodian, Korean, Chinese, Laotian and Thai communities where at least 70 percent of the group members rely on their native language.

The value of native-language interviewing can be seen with open-ended questions where respondents tend to provide longer-than-normal answers that must be recorded in the native language to get the full flavor of the response.

Children in these communities between 2 and 15 years old who were born here or have been in the United States for at least five to seven years speak English as well as their native language. However, they often cannot write or read the native language.

Cultural sensitivity a must

Household composition also presents an interesting challenge. While mainstream American households typically consist of a single family, Asian-American households are usually a compilation of a number of extended family members and close friends. It is not at all uncommon for such a household to include grandparents, nephews, nieces, or even close friends who share equally in the management and financial obligations of the home. This is why market research within these communities always includes a household composition question. Two areas we do treat very sensitively with these communities are education and income.

Another important consideration in surveying Asians revolves around using numerical rating scales versus verbal rating scales. Using words like "excellent," "very good," "good," etc., doesn't work in these communities because there is no conceptual way to translate or make respondents understand the difference that mainstream Americans make between "very good" and "excellent" or "not very good" and "poor."

But simply following the rules outlined here is not enough. Any native language research or subsequent advertising or communications to these communities must also be "culturally correct." Let me cite an example. It concerns a video ad campaign targeted at older, high-income traditional Japanese men whose companies might buy a condo in the United States for business purposes.

The actor chosen was a Japanese-American whose family had been in the United States for generations. He spoke no Japanese and had little contact with his native culture. During the filming, our intercultural consultant was brought in to supervise the script content and setting to avoid offending the target audience. Marketing research, both qualitative and quantitative, had been conducted two months before the shoot. One of the first things the consultant noted was that the Japanese actor was wearing slippers and a robe. After the consultant learned the setting was intended to represent a traditional Japanese male, he suggested three key changes.

First, the producers learned that slippers are not worn at home. Second, in one scene the actor was exposing the bottom of his feet, which is considered an insult in the Japanese culture. Third, the way the actor wore the robe was reminiscent of the way a robe is used in Japanese burial rites.

This story is just one example of how lack of respect for these communities, and not conducting proper research or using an intercultural consultant before embarking on an Asian-American marketing effort can have a disastrous effect on the ultimate goal.

I always think of three words when I'm dealing with an Asian-American person or project: face, honor and respect. If you never forget these three words, you'll considerably reduce the risk of making a critical error that could prove embarrassing and costly.