Instrument Structure

The following outline provides a basic format for surveys. There will be times when alternate forms are appropriate, but the general rules should apply to most circumstances.

• Begin the survey with a brief introduction or cover letter.
  
  
• Organize the survey so that the first questions are easy for the respondent to answer (i.e., ask for an opinion) and are interesting so that they will be motivated to continued filling it out.
  
• Bad example: Which Industrial Technology courses were part of your area of concentration when you attended HSU?
• This requires respondents to figure out factual information from the past which may be difficult, and is not an interesting or engaging question.
  
• Better example: Which Industrial Technology classes were most valuable in preparing you for your current job?
• This requires the respondent to remember the specific classes, but asks for an opinion that is probably readily accessible and not too difficult to answer.
  
• Best example: What part of your Industrial Technology education did you feel was most valuable?
• This simply asks for an opinion, it does not involve remembering specific courses from the past, and it shows that the authors of the survey value the respondents’ views.
  
  
• Group related items in order to keep a logical organization for the respondent. (See Question Checklist)
  
  
• Write a brief (one sentence) introduction to each section of the survey to alert the respondents to the new topic being addressed in that section.
  
  
• Keep items short and to the point.
  
  
• Group items that use a common type of scale (e.g., Likert) This can make a survey much shorter, easier to answer, and more attractive.
• Click here for an example
  
  
• Simplify response scales so that items can be grouped.
  
  
• Keep response scales simple and consistent.
  
  
• Wait until the middle part of the survey to ask difficult or objectionable questions. (By this time the respondent will be invested in completing the survey.)
  
  
• Leave the demographic questions until the end of the survey.