Guidelines for Writing Effective Market Research Survey QuestionsBy: Kristin Bright
Clients often ask us about the particular wording of market research survey questions and whether questionnaire wording affects the outcome of a survey. Wording of certain types of questions can definitely impact the interpretation by a respondent. Even small nuances can alter the response to a question and therefore yield misinterpreted results. However, it is easier to discuss the ways not to compose survey questions that cause bias or ambiguity than it is to discuss how to write a proper question. The following quick, five point checklist is a helpful tool when composing questions in any type of survey.
Use vocabulary in your research survey that is simple, direct and familiar
You need to choose words that are easily understandable by all respondents. Often technical jargon used in questionnaires is unfamiliar to the respondent and can mislead the respondent’s answer. Choosing wording that is simple and easy for a respondent to understand, regardless of their educational level, will ensure a valid response. This should also be considered when dealing with cross-cultural studies. Be careful not to use words that may having a different meaning in different cultures or ones that do not lend themselves easily to interpretation or translation.
Do not use survey questions that are leading or loaded
A leading question is one that attempts to guide the respondent’s answer or reveals the researcher’s opinion. This can be done easily by framing a question, “Wouldn’t you say half the neighborhood likes to swim?” Certain questions can also be more subtle in nature by not including all possible responses such as “don’t know” or “none of the above”. These are called loaded questions. For example, “When reading for pleasure, do you prefer to do …?” What if the respondent doesn’t like to read or is unable to read. Also, it is important to keep in mind that organizations or groups may have emotional associations. Using them in questioning can almost certainly bias the response. For example, “Doctors from Hospital X suggest that, …: Do you support or disagree with…? It is often wise to refrain identifying the sponsor to avoid this type of bias.
Avoid using double barreled survey questions
A double barrel question is one in which a respondent can agree with one part of a question, but disagree with another. This leaves the respondent unable to answer the question honestly. They are being forced to accept or decline a particular assumption in the question that simple may not be true. For example, Do you plan to move and buy another house within the next year? You could plan to move, but not plan to buy a house or not move, but plan to buy a second house. Another example might be “Did you receive friendly and accurate service?” This type of questioning leaves the respondent in a quandary again yielding misleading results.
Avoid using words with ambiguous meanings
It is up to the market researcher to provide the respondent with an adequate time frame of reference. This is a common mistake when using words like, often, sometimes or occasionally. When the respondent is not given the proper time reference, they will make up their own. When respondents make up their own time reference, results become impossible to interpret. For example, if asked “How often do you play golf? Do you play often, occasionally, seldom or never?” A respondent who has relatives that golf everyday, may consider once a week occasionally, whereas another individual who golfs weekly may consider that often. In this situation, you might ask, “how many times a week do you golf?” Another example would be, “Where are you from?” Do you mean where do you live now, where were you born, where did you spend your youth, or where have you lived the longest? Use words that are time or place specific to avoid confusion.
Avoid forcing survey respondents to answer questions that may not apply
Often times respondents who claim to have utilized a particular service are asked to rate the performance of an employee in several areas. “On your last contact was the sales representative friendly? Did they shake your hand?” Maybe the last point of contact was via the telephone, thus shaking hands would not apply. Try to ask qualifying questions to avoid these sorts of situations and thereby allow your respondent to skip questions that aren’t relevant to them. If asked up front what the nature of their last contact with the sales representative was, asking whether or not you shook hands mistakenly might be avoided.