This is an article I didn't want to write. Telephone vs. door-to-door research? Enough said, to my way of thinking. Yet alas, here I go, for there has been a new load of telephone prattle that must be addressed.


First, I invite you to ask yourself. Where does this pro?phone propaganda come from?

One source. One person.

Of all those research practitioners specializing in Hispanic work, I know of just one who consistently promotes telephone research. Unfortunately this lonely guy pays a PR firm to be his megaphone, so we hear him from many directions. And that one voice, many places, has been saying, in effect: Look, only about 15% of Hispanics don't have phones. What's the big deal?

To start, the percentages of Hispanics with telephones is 70-80%. I know the 1980 Census said just over 80%, but the Census tended to miss those least likely to own telephones. The Census people themselves are now admitting to around three million undocumented Hispanics. What percent of illegal aliens rush out to sign up for the phone? Maybe 20%-50%? Let's say 50%. Add 1.5 million Hispanics without phones (that's 50% of three million) into the revised Hispanic population total and re-calculate phone ownership and what do you have? About 75% with phones.

Moreover, the typical definition of "Hispanic" for a marketing program is a Hispanic consumer who is less than fully assimilated. (This is why we often have a screening requirement that respondents spend a given number of hours per week with Spanish media). Unassimilated Hispanics are less likely to own phones, thus, the operational proportion with phones is even less than 75%.

Even so, a big-picture person might resist, arguing that the majority do have phones. And, such a person might add triumphantly, "Dauten, you yourself have stated that door-to-door research can only reach Hispanics in 'Hispanic areas' and thus overlooks 15-20% of Hispanics. If the telephone misses 25% and door-to-door misses close to 20%, the error is about equal."

This tit for tat would work if the tit and the tat were a decent swap. They aren't. The individuals overlooked by door-to-door research are those who live outside "Hispanic areas" and thus are those most likely to be assimilated. Those without telephones are least likely to be assimilated. (As I have explained at length elsewhere, telephone ownership is a cultural variable; the U.S. is a telephone culture whereas Latin America is not. There are only about seven phones per 100 people in Mexico).

A chart might help. In Illustration 1, you can see that the majority of Hispanics have telephones and live in "Hispanic areas." On the left are those individuals without phones. On the right are those living outside Hispanic neighborhoods. When you buy decent Hispanic research, you get the group in the middle, one way or the other. Door-to-door adds the group to the left, while telephone adds the group to the right.

illustration 1

It might at first seem like a nearly even trade and if it weren't for the assimilation variable, it would be. Does that assimilation variable matter? It does.

Let me review a test recently made public.

Place: Chicago.
Contenders: Door-to-door vs. telephone.
Sample: More than 1,000 Hispanic respondents.
Object: Measure the impact of methodology on estimates of Spanish media usage.

The idea of this test was to measure the share of audience going to Spanish-language vs. English-language radio stations. The Spanish stations were concerned that telephone research might underestimate their audience. And indeed it did. The Spanish station's aggregate share (averaged across day-parts) was as follows:

If you are one of those people who believes that Hispanic marketing is a waste of time, a distraction from the primary market, then undertake telephone research. By ignoring the portion of the market most likely to respond to the campaign you will greatly increase your chances of "proving" that Hispanic marketing doesn't work.

Look again at the Chicago radio data reported above. The telephone sample would have convinced you that nearly half of Hispanics' radios were tuned to English-language stations. You can then argue that you did not need a special radio campaign to reach Hispanics, that the regular mainstream campaign would do the trick. On the other hand, if you had opted for door-to-door research, you would have found that English-language radio had only a 17% share and thus concluded that a Spanish-language campaign was essential.

The same would often be true for making conclusions about a Spanish language ad campaign: To choose the telephone is to choose to ignore much of the unassimilated population and hence to choose to ignore much of a good campaign's effectiveness.

A few points about the comments above. First, the assumption was that all respondents with telephones would be eligible to be interviewed via telephone research. This is rarely the case. Most telephone research is done by Spanish surname and thus excludes Hispanics with unlisted phone numbers and those who do not have Spanish surnames. The upshot is that the actual percentage of Hispanics eligible to be interviewed is not the 75 % who have phones but the 40?50 % who have a telephone and a phone book listing, and a Spanish surname.

Second, there are advantages to telephone interviewing beyond just sample quality. Door-to-door research allows for:

  • The possibility of longer interviews;
  • Lower refusal/non-cooperation rates, and
  • The opportunity to use show materials and brand photographs.

Finally, as Hispanic market expert Juan Melendez recently pointed out to me, there is another factor which is often overlooked in choosing research methodology: Comfort. Because the majority of households in Latin America do not have telephones, there is virtually no telephone research there. Some research is done in central locations but most is door-to-door. Thus many Hispanics in this country find telephone research strange and door-to-door familiar.

Once again, I must conclude that door-to-door research gives the more accurate reading of the Hispanic market. That being said, let me add with a personal message. Our reservations about phone work are not self-serving. We at Research Resources not only do telephone research, our work lives would be a bit softer if we were to push it.

It's a lot less work to call up Survey Sampling and order a telephone sample and conduct the interviewing via a phone bank here in Los Angeles than it is to construct a door-to-door sample and coordinate the interviewing with several cities. Yet we continue to do most of our research in person because it offers a sample with double the coverage of the typical phone sample.

And we believe the sample can make the difference between a Spanish-language marketing campaign being judged a success or a failure. It's just not fair to our clients, or to the advertising agencies that assist them, to provide diluted data.

So back to our question: Does that minority within a minority, Hispanics without telephones, really affect the results of research? It's the tail that wags the research dog.

Research Resources, Westlake Village, Cal., is a minority-owned company that specializes in Hispanic and black market research. It was founded by Dale Dauten and Teresa Menendez and offers complete market research services in both minority and general market research.