08 June 2007

Online Survey Instrument

Hi Folks,
You may be interested in the online survey instrument produced by http://www.SurveyMonkey.com . They offer a free trial of unspecified duration, for an unspecified number of surveys. On most of their webpages they present various features - for which you need to 'upgrade now' (this translates basically as you must pay money). However, the trial membership is an excellent way to get people to try it out, and find - as I did - how easy it is to use, and quite effective educationally.

I plan to 'upgrade' to full service tomorrow. Today, I would like to cover just a few points about online surveys generally, and SurveyMonkey in particular. Generally, such surveys could be divided into whether or not they are being used formatively or summatively. A formative survey could be used to gather data from current actors in order to inform a future potential revision in a course (as performed by WestGa), or as summative to judge the learning achieved (usually respondents not anonymous) or the teaching (usually anonymous) or any other completed activity. I am against anonymous surveys as a rule since anonymity can provide cover for incomplete or disruptive responses, that are difficult to repair or recover, and moreover I believe it is better to value and honour the time and effort of the respondent by acknowledging their participation.
My present survey is formative, administered in the middle of a course, and inviting feedback on future potential revisions.

I have designed and applied a SurveyMonkey to 20 students at college, and they should complete this by 6pm. No doubt, then I will be in a better position to report any glitches. The actual survey html code is very complex and looks as if it was developed by google.analytics.com based on their /urchin program. It is not adaptable by the user (i.e. me), which is a drawback. Since it is standard design protocol for surveys to use the native-language of the respondents and my students are Japanese, so then I would want to write the survey in Japanese. Even in English, I would like to adjust the amount of 'white space' in the presentation - SurveyMonkey has everything quite densely packed on the screen-page. The margin widths for my label to go beside their textbox is limited to only nine character-spaces (although in their sample they have 12, maybe because the upgrade version has more flexibility). I wrote the survey questions on paper first and reflected on these for a day or so, before entering their website and creating my survey. Everything went smoothly in retrospect, but in the creation process I did have several worries - for example we do not need to number our questions - they will do this for us. We do not have to 'save' the final creation anywhere or by any special button, they do this for us automatically. It would have been more comfortable for me if they had said we can just shut down our computer and the survey so far will be safely preserved until we next visit and log-in, (who said learners shouldn't take risks?) The 'Edit Question' button was a useful button to allow later changes. Everything was left-aligned in this trial version. I hope the full upgrade version allows for us to align text as we choose.

This morning, I decided it would be interesting in the analyses to look at the responses with respect to the time of day they were written. I can perform a wave analysis to see how results would look if I had gathered data from an unlimited large number of respondents (more on this tomorrow). My final version had ten questions ; - the first asking for the date and time (and forcing a completed response), the second their name, email address and so on (not forced) to deny anonymity, third, fourth, and sixth asked them to select one of a comprehensive range of given answers, fifth write their top three favourite learning activities using three singleline textboxes, and then seventh, eighth, and ninth used comment/essay boxes. The final tenth question asked them to select one of a given range presented on a single horizontal line. Overall, I used four types of question. SurveyMonkey offers these templates for about fifteen different types of question.
You can view my survey here , but please do NOT complete it,
because you would be corrupting my data \\\ thank you ;-)

I am keen to try to use their software to design my own template, but first I want to look at the quality and depth of their analyses, and to check if I can retrieve the raw data easily or not. More later ...
All Best Wishes
Paul

No comments: