Surveying the landscape

I love giving and taking surveys. I just sent this one out via GoogleForms on Friday. The first of many I'll be using to get feedback from our 1000+ staff here in the district.

The more I use surveys, the smaller and more focused they get. One of the first ones I created back in the mid-90s was called the Mankato Survey of Professional Technology Use, Ability and Accessibility (the titles are getting shorter too.). While I can't find the original, written in FileMaker, here is a link to the 2003 incarnation. I'd honestly forgotten I'd written it until somebody ran across a reference to it and asked for a copy this week. Took awhile to locate.

My CODE 77 Rubrics have proven to be useful survey tools. I've written surveys to help users evaluate their library programs (Here's the teacher version in SurveyMonkey) and library facilities (Here's a version for student evaluations.) I've tried to get a teacher attitudes toward technology implementation with this survey. And of course, a needs survey is a part of every long-range tech plan we've written in the district. (We used a state survey for the latest tech plan.) This past spring and this fall we are surveying students to determine the percent who have home Internet access using this tool.

I've learned a few things about writing surveys. 

1. Don't give as survey unless you have a use for the information it will provide. Yeah, duh.
2. Test surveys before you send them. If you send your to just a few guinea pigs before sending it to the masses, your survey will not only have fewer typos, but fewer questions with confusing questions and useless responses.
3. Don't allow neutral answers. When using a Likert scale (Strongly agree 4 3 2 1 Strongly disagree), force people to choose a response either on the positive or negative side.
4. The more questions, the smaller number of responses. Make your surveys as short as possible. I think for every question you ask, you lose about 10% of your potential respondents. How do you make the survey shorter? Don't ask a question unless you have a use for the information it will provide (see #1).
5. The more convenient the survey, the greater the number of responses. Although the graphic above shows a web-based form, I actually sent the survey out to my staff in the body of a e-mail. Click two buttons and then submit. Finished.
6. Share your results. When people are asked their opinion, they like to know they've been heard and I believe most have a genuine curiosity about the results of surveys in which they've participated.
7. Give people a change to comment. Give people a choice of 10 things, someone will come up with a reasonable 11th. I often find that comments provide as much or more information than the questions.
8. Use the results as a guide, not gospel. Since I tend not to worry about the same things professional survey makers worry about (like sample sizes, validity tests, etc), I use the results as indicators not definitive answers.

My plan this year is to do a weekly 30-second survey. And maybe repeat them in future years to get some longitudinal data. It'll be interesting.

Results from the survey above:

 And my favorite comment:

Some of us do not Facebook or Twitter.  Those are evil time wasters.  Have you seen the episode of South Park where Stan gets sucked into the internet by his profile?  I mean, really, No Siree, I do not want to go down that route.  Please continue with email postings. ;-)