W3C home > Mailing lists > Public > public-html@w3.org > August 2007

Re: Likert scale for survey? (on design principles)

From: Laura Carlson <laura.lee.carlson@gmail.com>
Date: Mon, 6 Aug 2007 12:44:48 -0500
Message-ID: <1c8dbcaa0708061044v74edfda6i155bcd6a939b696c@mail.gmail.com>
To: "Dan Connolly" <connolly@w3.org>, public-html@w3.org

> Thanks for the input on the options; I'm not sure how best
> to do this survey...

Surveys can be difficult to design.

> I chose options like "Yes, but only if changes are made..." because
> I know what action to take when someone chooses them.

Oh, I thought that the survey was more an informational tool to gauge
attitudes, because it [1] says, "This survey is a non-binding,
informational-gathering exercise." Likert-scale questions can be good
for measuring the degree of respondents' feelings or attitudes about

> I don't know what the editors or the WG should do differently
> based on a reply of "Disagree" vs "Strongly Disagree".

Then perhaps consider a scale of:

- Agree
- Somewhat Agree
- Neutral
- Somewhat Disagree
- Disagree


Agree =  Full support as written.
Somewhat Agree = Has suggestions for consideration.
Neutral =  Neither agree or disagree.
Somewhat Disagree = Some changes are needed.
Disagree = No support.

That would even out the scale and still provide actionable responses
for the editors and WG.

The generic text box would be the mechanism for respondents to use to
communicate rationale for all choices, suggestions, needed changes,
and/or conditional provisions. It could also alert us to any multiple
dimensions in a topic, and convey a richness of feeling often missing
from closed suggested answers.

Best Regards,

[1] http://www.w3.org/2002/09/wbs/40318/dprv/

Laura L. Carlson
Received on Monday, 6 August 2007 17:44:53 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:25 UTC