When developing and analyzing surveys, there is a considerable amount of time and effort in developing the questions format, sequence, length, and the effort and time required by respondents to answer them.
Questions have a technical aspect to them, (see our Paper on Question types) but also developing questions has to be done in a specifically defined framework.
When asking questions, first, understand who will be taking the survey. The framing of questions around your audience is paramount. If you have different respondent groups, you should consider different surveys for each group. For example, when surveying teachers and parents about software or social media, the lingo and defining of terms would be different for each audience. The correct use of terms and starting point of your audiences understanding can be quite distinct. You do not want to alienate either audience by talking down or simplifying to one and alienating another by not defining the terms that are not universal to them.
The framing should be obvious. In this way, the respondent knows the context in which the question is asked. Framing the context of a question is often omitted in order to keep surveys short. Framing the context is done by establishing the facts before asking a question. An example would be: "What should be done to get students to complete their homework?"
To some parents, this may imply that their student may not be completing homework, when the parent in fact knows her student spends three plus hours a night on homework. A better way to ask this question is to establish framework around facts, which might read:
Within our school district teachers assign 2 hours of homework per night. However, teachers state that only 50% of students are coming to class with completed assignments, which disrupts the lesson plan for the day not allowing new material to be presented as scheduled.
After this establishment of facts within the survey, questions may be asked within this framework such as:
Do you believe your student completes his/her homework every night?
Does your student spend more than 2 hours per night on homework?
Do you think 2 hours is the appropriate amount of time for homework?
What do you think can be done to get more students to complete their homework?
It appears at first, that this type of framework around questions makes a survey lengthy, it does, but this is where targeting a survey to one specific topic plays a role. The survey cannot be about the cafeteria food, school sports and homework. It has to be specific to the subject to be studied, where one specific problem to be researched or understood is the only theme of the survey.
While keeping survey short is paramount in eliminating high abandonment rates, asking questions within an established framework on one topic at a time, keeps the responses meaningful and worthy of drawing conclusions and recommendations for understanding and action.