In my web-based survey the sampling bias is likely to influence the initial questions about basic use of computers. I assume the participants will not be expected to have ‘no skills’ in using a computer connected to the Internet when participating in a ‘on-line survey’, but these questions are equivalent to the previous survey that are used for comparison. Teachers that don’t use computers will have an increased ‘drop-out rate’ and it will be hard to tackle the possibility of multiple submissions, incomplete responses, and data security.
I will in my thesis address potential problem when conducting Web-based surveys:
2. Sampling error
3. Measurement error
4. Nonresponse error
(Dillman, Tortora, & Bowker, 1999)
I’m planning to use Google Docs’ spreadsheets survey/forms feature and played around with the “Blank survey template”.
You enter the question in the Title and if you want supply help text that assists the users completing your survey. Select the Type of question (Text, Multiple Choice, etc.) and mark if the question is required (forces users to enter a response). At a later stage you may re-order your survey by dragging-and-dropping the question to the desired location. Finally to distribute your survey, you may use “Email this Form” (this generates an e-mail with the link to your survey form), or click “Embed” from the “More Actions” menu. Google will then display the iFrame code that you may copy/paste into your e-mail or to use on your web page. Responses are collected and stored in a Google Documents Spreadsheet (compatible with MS-Excel). When required you may disable your form by clicking “Accepting Responses” from the “Form” menu. Then you can also create a copy of an existing survey form by selecting “Create a Copy…” from the “File” menu and new data can be collected.