Editor's note: Felipe Korzenny, Ph.D., is president of Hispanic Marketing Communication Research, and its Asian division Asian Marketing Communication Research, Belmont, Calif. Betty Ann Korzenny, Ph.D., is executive vice president of the firm.

What do you do when you need to have your questionnaire translated into Spanish (or another language)? Would you ask an analyst who took Spanish (or the required language) in college to translate it? Would you ask a native speaker in your office to do the job? Or would you call a friend from the country where that language is spoken to help you with this minor task? While all three options might seem feasible in theory, you could very likely find yourself with many difficulties.

There are many perils lurking in the shadows of cultural diversity for the unaware researcher. Just think about the possibility of someone in a research house using a computer program to translate an instrument for his or her latest study. The results could be disastrous, not because a computer program should not be used at all, but because thinking, judging and cultural knowledge are as important as a big vocabulary in the target language. Research procedures and instruments are full of intentions that need to be deciphered properly; this cannot be done out of context, no matter who - or what - does the translating.

The fallacy of cultural interpretation

The unaware researcher is likely to miss the nuances of cultural interpretation and conclude that adapting research to another language is a relatively trivial exercise. Translation is more complicated than it may seem, however. It is easy to assume that anyone with university-level knowledge of the target language (let's take Spanish in this case) should be able to produce an acceptable Spanish version of an original questionnaire in English. This is what we call the elusive fallacy of cultural interpretation.

This fallacy consists of thinking that any intelligent and educated speaker of two languages can interpret one language into the other. The reader may notice that we have stopped speaking about translation and are now speaking about interpretation.

Interpretation is the task of attaching meaning. Translation, on the other hand, deals with transforming one set of symbols into another. This is an important difference because interpretation is less concerned with the code (or words) and more concerned with meaning.

Cultural interpreter vs. translator

Adapting messages, questions and instructions from one language into another is an ambiguous task. Finding the right word may be what one needs in some instances. On other occasions, finding cultural equivalence may be the key.

When conducting research, the interpreter must be versed in both the original and target languages. She or he must also be cognizant of cultural nuances in both cultures. And to top it all off, the interpreter must be knowledgeable about research methods, conventions and customs.

An interpreter culturally adapts text so that it can be understood in the target language in the same way it was understood in the original language. The interpreter must redo the text from scratch. There is no feasible compromise.

When you combine idiomatic expressions in consumer language that are available in one culture, but not in the other, you get a very complicated set of circumstances. An innocent "translator" faces a formidable task when trying to translate such words and phrases. When he or she is done with the translation, the outcome may make no sense to the speaker of the target language.

Making sense of interviewer instructions

When a naive translator takes on a questionnaire, she or he will likely have a series of problems understanding the intended meaning. One particularly sensitive area of confusion in translation is instructions to the interviewer. The lingo used in the industry for research instructions can be perplexing when words alone are the focus of attention.

These instructions, if literally translated, can make very little sense. How will a naive translator phrase something like "If yes on A2b go to B3c. If no on A2b and yes on B 1, terminate. If yes on A2b and no on B 1, skip to C30po"? In our experience, these and simpler types of instructions tend to be awkwardly and incorrectly translated.

Just imagine the interviewer's confusion as he or she tries to make sense of already complicated instructions that have been expressed badly in another language. If there is not an outside check and verification, the data obtained could be simply trash.

The importance of conceptual understanding

Another essential area of research expertise for interpreting a research instrument or protocol has to do with sensitivity to subtle meanings offered as responses to a question. Take the case of asking about television viewing patterns. The optional response of "prime time" television has no equivalent in the daily parlance of Spanish speakers. The best one could do in that case would be to list, by hours, the actual time segments during which people might watch TV

Country music sounds easy to translate, right? Well, for Mexicans, for example, "country music" may be ranchera or simply "Mexican or Latin music." These latter types of music have absolutely nothing to do with the "country" music most commonly referred to in English, the kind Kenny Rogers sings.

So what would be an appropriate translation? Your best bet is to simply leave the term untranslated. If the respondent listens to U.S. "country music," she or he will be familiar with the U. S. English term. If you want to learn whether the respondent also listens to ranchera, simply use this well-understood music category in Spanish.

If, as we have found in some cases, country music is translated into Spanish as ranchera the client may waste large amounts of money. She or he may buy time for country music stations. That money, however, should have been placed in Mexican ranchera stations.

An instrument may not contain relevant categories. If the interpreter doesn't know about the research intention and/or the target culture, she or he may allow the client to only ask about categories that are irrelevant to the culture.

The respondents may not be given a chance to identify what they actually watch or listen to. Imagine asking Spanish-dominant Mexicans exclusively about country music, rock, classical, easy listening and Top 40.

That would not allow them to indicate they listen to boleros, ranchera, romantica, salsa, banda and many other Mexican music genres on the radio. With this omission, it would appear that this public listens to the radio very little, which would be erroneous.

An informed instrument interpreter would probably keep in mind that response options need to be exhaustive and mutually exclusive. A naive translator is likely to assume that what she or he was given to translate is final. That assumption tends to be generally dangerous. Rarely is everything included when moving from one language and culture to another.

Back interpretation - do it right

"Back interpretation" can be very useful in instrument adaptation. The method is not difficult but requires attention to detail. First, you need two interpreters, one whose native language is English, and one whose native language is Spanish.

The native Spanish speaker first translates into Spanish, then the native English speaker translates back into English. An informed researcher compares the two English versions; if they are comparable in intended meaning you are likely to be on safe ground. If the versions diverge, the two interpreters and the researcher need to discuss any discrepancies and adjust them.

Cultural verification through empirical research

A more scientific but costlier approach to interpretation and cultural adaptation is to verify the interpretation by submitting it to a panel of bicultural and bilingual judges. This can be done in a focus group setting in which bilingual members of the target audience are asked to debate and compromise.

These individuals are presented with the English and proposed Spanish versions and then asked to verify the proposed Spanish version. The discussions produced in this manner are lively and also insightful.

If the interpreter sits in the back room, she or he is usually surprised to hear the comments and discussions. The interpretation thus produced is likely to contain the intended meaning for the intended audience.

Cultural adaptation is more than linguistic interpretation

Culture is composed of a variety of elements, some implicit (subjective) and some explicit (objective). Language interpretation tends to focus on the explicit symbols of language expression. In conducting research across cultures, particularly in the Hispanic market, other cultural issues emerge that affect the design and operation of research.

Cultural adaptation beyond questionnaires

Here we will touch upon a series of other research design issues that require cultural interpretation and understanding. Going beyond language stresses the importance of considering cultural adaptation in broader terms.

Qualitative research assumptions

In qualitative research sessions, whether focus groups, in-depth interviews or other formats, researchers make cultural assumptions. These assumptions are taken for granted. For example, it's assumed that consumers can easily participate in imagery exercises.

What type of person would a candy bar be?

The idea that candy bars can have personalities seems to be taken for granted among U.S. qualitative researchers. One commonly hears "Let's imagine these candy bars are people . . . What type of person would each of these candy bars be?"

Even more interesting, "What type of animal would each candy bar be?" For respondents coming from other cultures, these exercises can prove difficult, if not nonsensical.

Cultural adaptation of these types of research procedures calls for understanding the expectations of members of the culture.

General market respondents in the U.S. are quite familiar with TV game shows and with other types of playful ways in which one can explore meaning. Among members of certain cultures, particularly Hispanic, a group setting tends to be taken seriously. Group settings are taken so seriously by some Hispanic respondents that they stand up when they speak.

For these respondents, a group setting is one where one has to be careful not to lose face. Common comments these writers have heard include: "But candy bars are not people," or "How can a candy bar be an animal?" Respondents may be baffled at how to respond and may tend to view the group process with less seriousness.

What score would you give each of these concepts?

Other common research procedures that may require cultural adaptation when moving from one culture to another are rating or ranking exercises. While qualitative research should not generally be used for the collection of quantitative data, it is common practice to do so occasionally. For example, respondents are given the opportunity to rate concepts so they make a private commitment before discussing the concept.

There are members of Hispanic cultures who are not used to ranking or rating conceptual statements. The notion that abstract ideas can be quantitatively compared may be not be a universally shared notion. In addition, quantitative skills are not universally shared.

There are occasions in which respondents are embarrassed by a client who insists on obtaining rank- order comparisons among concepts.

What does it take to get results?

To alleviate these problems, the researcher, acting as a cultural interpreter, adapts these exercises to the requirements of the cultures involved. In the case of the imagery exercise, one would ask for relevant conceptualizations, e.g., "What type of person would eat that candy bar?"

A more powerful imagery question would be "What images come to your mind when you try this candy bar?" This more direct approach tends to be productive. It asks for experiential reactions, something the person can relate to.

In the case of the rating or ranking exercise, respondents should probably be asked to read with the moderator each concept at a time. In groups where literacy levels tend to be lower than in the U.S. general market, those who can't read still have the option of following along. Then respondents are asked to think of the concept that convinced them the most about the product. The outcome is discussed.

Then respondents are asked to think about the concept that would follow the one just discussed, and so on. This approach makes the task less burdensome because the complexity of the judgment is confined to one item at a time.

Conclusion

Cultural relevance depends on the ability of the researcher to adapt instruments and procedures to the cultural groups she or he works with. Adaptation requires skills beyond those normally held by a representative of the culture.

A cultural interpreter for research combines research skills, bicultural sensitivity and experience in cultural interpretation. Successful intercultural research goes beyond culture in searching for universality. At the same time, the intercultural researcher must be willing and able to look through the multipaned window of culture to make sense out of uncertainty.