Transcript Advantages of web surveys

Collecting Information via the
Web
Stephen Porter
Director of Institutional Research
[email protected]
Michael Roy
Director of Academic Computing Services
[email protected]
Wesleyan University
Advantages of web surveys






Low costs - only software, no printing, envelopes or
postage needed.
No data entry costs.
Minimal data entry errors.
Easy to correct problems during survey
administration.
Much quicker completion time for project.
Response rates are comparable.
Population




It’s important to keep in mind your survey
population.
Do you have good email addresses for sample?
Does everyone use email software that allows
hyperlinking?
Do some people not use email at all?
 Students (especially grad) may not use institution
assigned emails. Do you collect home email
addresses?
 Some low-tech faculty don’t use email or check
infrequently.
Security and web surveys


Security can be a big issue – you have two choices:
Ask for name and other identifying information such
as userid and password.
 Multiple responses can be eliminated.
 Responses can be linked to institutional data.
 No evidence this affects response rates, but be
aware many people are nervous using SSN’s on
the web.
 Important to state that only aggregate data will be
released.
Security and web surveys

Allow anonymous responses.
 In theory, anyone can answer the survey.
 Multiple responses will vary with topic and survey
length.
 Cannot use incentives.
 Research indicates that promises of anonymity
and/or confidentiality do not affect response rates
except for sensitive topics such as sexual behavior.
Security and web surveys



Sending out id numbers and passwords can be
cumbersome.
It’s possible to embed an id number and password in
a unique hyperlink that is emailed to each survey
participant.
This approach has worked well for us, but make sure
participants know that the survey is not anonymous.
Survey structure





Keep it simple: avoid complicated designs with lots of
colors.
Make the survey one (screen) page in length.
For drop down boxes, make sure first response
category is not the visible line of the box.
Keep the survey as short as possible.
See Dillman (2000) pp. 352-401 for a good set of
detailed recommendations.
Survey administration



The Dillman (2000) “tailored design” approach
 Pre-notice, survey, reminder, follow-up survey, and
so on.
 For web surveys it is simply multiple emails.
 Considered the best method possible for surveying
- used for example by the Census Bureau.
If you do only one survey mailing, you WILL get a
low response rate.
With multiple emails you need to take out
respondents from your email list.
8/
0
4/ 2
9/
4/ 02
10
4/ /02
11
4/ /02
12
4/ /02
13
4/ /02
14
4/ /02
15
4/ /02
16
4/ /02
17
4/ /02
18
4/ /02
19
4/ /02
20
4/ /02
21
4/ /02
22
4/ /02
23
4/ /02
24
4/ /02
25
4/ /02
26
4/ /02
27
4/ /02
28
4/ /02
29
4/ /02
30
/0
5/ 2
1/
5/ 02
2/
0
5/ 2
3/
5/ 02
4/
5/ 02
5/
0
5/ 2
6/
02
4/
Cumulative response rate as of midnight
Survey administration
Response Rate for 2002 Senior Survey (Web Version)
100.0%
Second reminder
email
Third reminder
email
80.0%
First reminder
email
60.0%
Initial
email
40.0%
20.0%
0.0%
Survey administration





The email should be concise, with hyperlink to survey
visible when participant opens the email.
Avoid sending emails on Monday.
We usually wait a week between reminder emails.
How many? 2 or 3 reminders are fine, especially if
you allow participant to opt out.
Offering a web option via “snail mail” does not seem
to work.
For more information



Dillman, Don A. (2000) Mail and Internet Surveys:
The Tailored Design Method. New York: John Wiley &
Sons.
National Computer Systems, Inc. (NCS) has a great
website with a sample size calculator, survey research
glossary and a set of short reports on various survey
research topics: http://www.ncs.com/researchnotes/
Talk with your institutional research or assessment
office – no need to reinvent the wheel.