Nondirective Probing

Probing inadequate survey answers for the additional information that may be necessary to fully meet a question's goal(s) is an important element of standardized survey interviewing. In training interviewers to probe effectively, an important distinction should be drawn between nondirective and directive forms of this technique. Unlike directive probing, nondirective probing is designed to encourage and motivate respondents to provide clarifying information without influencing their answers. That is, this approach is specifically designed to be neutral in order to avoid increasing the probability that any specifie type of answer is encouraged, or discouraged, from respondents. When nondirective probing is employed, an answer is never suggested by the interviewer. Some examples of nondirective probing of closed-ended questions include slowly repeating the original question or repeating the full set of response options (e.g. “Is that a ‘Yes’ or a ‘No’?”). When asking open-ended questions, some nondirective probe examples include repeating respondent answers, using neutral statements such as, “Could you tell me a little bit more about that?” “I'm not sure I understand what you mean here,” “So why do you feel that way?” and “Is there anything else you wanted to say about this?” or simply pausing while respondents collect their thoughts.

Nondirective probing is also important when requesting numerical information. Useful strategies when probing answers to these types of questions include asking respondents to provide more exact information (e.g. “I need a more precise figure if possible”), asking them to select a single number from a range of values initially reported (e.g. “Would you say 2 or would you say 3?”), and asking them to perform any necessary calculations when they provide information using a format other than what was requested (e.g. question: “How old are you?”; answer: “I was born in 1955”; probe: “So how old would that make you?”).

In contrast, directive probes are not neutral. They may inadvertently bias respondent answers by limiting the potential range of responses available or by suggesting that some answers are more preferable than others. In probing a closed-ended question, an example of a directive probe would be presenting a truncated range of response options (e.g. answer: “My health is on the low side”; probe: “So, would you say your health is ‘only fair’ or ‘poor’?”). Interviewers often also will construct directive probes to open-ended questions by attempting to reword a respondent's initial answer (e.g. “In other words, you are opposed to income taxes because they are a disincentive to work?”). Similarly, an example of a directive probe to a numeric question might be, “So that means you were 12 when you first smoked a cigarette?”. These latter two examples highlight the fact that directive probes can often be answered with a “Yes” or “No” answer.

Although an important element of standardized interviewing, nondirective probes themselves ironically can be only partially standardized and hence are both employed and worded to some extent at the discretion of the interviewer. This variability should also be considered a potential source of measurement error, one that is best confronted through careful training of interviewers regarding the critical nature of their role in conducting standardized survey interviews, as well as the specifie goals of each question included in the survey instrument.

...