Programming and testing a survey can prove to be a very time-consuming process for an analyst. Oftentimes tinkering with one question can impact a large branch of logic that can cripple your programming sequence if done incorrectly. When programming a survey, it is extremely difficult to multi-task because of the focus that is needed to handle one piece at a time. If a programmer gets too concerned with fixing multiple items (logic, randomization, grammar) by juggling back and forth, it can become easy to overlook items.
Over the years, I've found that checklists in market research, such as a programming checklist provide several key benefits for users. The checklist allows you to remain focused on one item at a time and allows you to make sure encompassed all angles of survey testing which gives you peace of mind that everything has been covered before your soft-launch. If the checklist is itemized enough you'll have confidence the users will have the experience that is intended and they do not get confused or frustrated. The checklist helps provide a more global look at the testing process and it is highly recommended.
Here are some general survey programming areas you need to test:
Check for single versus multiple response questions
Check to determine if your question offers N/A, unsure, or another option to bypass if needed
Check to see which questions are forced answers and which are not
Check answer sequences (random, inverse, forward, backward, etc.)
Check to see if your questions offer "other" to ensure it all-encompassing - use the MECE rule
Check your scaling and label categories on questions
Check for open text boxes and character limits
Check your routing and branching logic
Check to see if your dis-qualifiers route to the end of the survey
Check to see if the order and flow of your questions make sense
Drive Research is a market research company located in Syracuse, NY. Visit our contact page to fill out a form to contact our team with any questions or comments you might have. Call us at 315-303-2040.