When words mean one thing to researchers, and another to subjects

Sometimes the simplest words can be misunderstood when people are interviewed by an investigator, and this is why cognitive interviewing research is growing.

The main purpose of cognitive interviewing is to obtain information about the processes involved in answering questions. Two basic procedures are involved, and these are think-aloud and verbal probing.1

While cognitive interviewing has become accepted practice among survey researchers, it's just beginning to gain notice for the use in the research informed consent process, says Gordon Willis, PhD, a cognitive psychologist with the National Cancer Institute in Rockville, MD.

"Clearly when it comes to testing survey questionnaires and the use of cognitive interviewing is widespread and no longer seen as radical or experimental," Willis says. "It was within the last 15 years that cognitive interviewing has gone from a new technique to an accepted, widespread application."

Willis is one of the pioneers in using the technique for the purpose of improving informed consent.

The reason cognitive interviewing is needed is because research participants often misunderstand the purpose of a study, and without cognitive probing, investigators will never know there is a problem, Willis says.

Willis offers these examples of how easily people can misunderstand questions asked by investigators:

• The phrase "sex partners" can mean different things.

For instance, when people are asked how many sex partners they've had, they might report having had two sex partners in the last 12 months when, indeed, they've had sex with more people than that, Willis says.

"But you asked about partners, and they think that means an ongoing relationship," Willis says. "So you have to ask them about the people they've had sex with to find out more information."

• Medical professionals use simple words that the public might not understand.

Even terms like inpatient and outpatient, which medical professionals use routinely, are not well understood by the general population, Willis notes.

Other simple questions that can lead to misconceptions are questions about work, Willis says.

• Does 'work' mean only legal, paid activity?

When people are asked how many hours they worked in the last month or last week, it's very problematic without further definition because it's not clear whether the interviewer or survey is referring only to paid work or includes volunteer work, Willis says. And some paid activities may not be considered work as the investigator thinks of it because they are illegal or nontraditional.

Likewise, questions about leisure time and physical activity cause confusion because some people might not consider exercise a leisure activity, and some cultures do not have a concept of "leisure time," Willis adds.

These simple examples show how easily survey questions and informed consent documents can be misunderstood.

Researchers also could use cognitive interviewing to avoid other types of misunderstandings, Willis says.

For example, the U.S. Census Bureau wanted to save time and money in collecting some information by asking individuals if they could collect the requested data from other federal agencies, Willis says.

Investigators, through the use of cognitive interviewing, found that as simple as the request was, it was misunderstood by the participants, Willis says.

More than half of the people thought the Census Bureau was asking if it could share its information with other agencies, which was the reverse of the intention, Willis explains.1

"On a receiving end, the message does not come across the same way, and cognitive interviewing is very effective in demonstrating that regular wording used to convey this information needs to be fixed up," Willis says.


  1. Willis G. Cognitive interviewing as a tool for improving the informed consent process. JERHRE. 2006:9-24.