[Open-data-census] Customized Questions and Scores

Sat Feb 7 08:12:55 UTC 2015

Hi Waldo,

Quick precision: actually the questions are editable (just fork this
spreadsheet
<https://docs.google.com/spreadsheet/ccc?key=0AqR8dXc6Ji4JdE1QUS1qNjhvRDJaQy1TbTdJZDMtNFE#gid=3>)
in theory without changing the underlying code. However, the instructions
<http://census.okfn.org/doc/#configuring-your-census> rather ominously say:

*We strongly recommend against customizing the questions. The app may well
break if you change the questions.*

So it could be a case of raising an issue
<https://github.com/okfn/opendatacensus/issues?q=questions+is%3Aopen> about
this to find out what exactly is the problem (hard coded IDs, etc.) and
making the project more generic.

A word of warning, you said it first: as with OpenSpending, one of the
fundamental goals of the Census is a basis for comparability between global
and local evaluations. Deviating from the question format will make your
census less relevant internationally. Please consider advocating for, and
working with the core team, on revising the approach across the board.

Secondly, we have been slowly starting a similar effort here to compare
Cantons and Cities, and there is still some resistance / misunderstanding
of the Census methodology. I think some improvements could be made in the
way the project is documented. I.e. I think we and the community could
really do more to pool our experience.

Lastly, there have already been over 45 forks of the code, and it looks
like Code for America is doing active development
<https://github.com/codeforamerica/opendatacensus/commits/services>. If
you're not the same team, you should get in touch :)

Cheers,
Oleg

*Opendata.ch* - Swiss Chapter of Open Knowledge <http://okfn.org/>
____________________________________________________

On Fri, Feb 6, 2015 at 10:59 PM, Waldo Jaquith <waldo at usopendata.org> wrote:

> Folks,
>
> I’m working with Emily Shaw, at the Sunlight Foundation, to put together a
> census of U.S. states’ open data efforts. We’re going to change the
> evaluation criteria a bit, to work better for U.S. states, notably by
> increasing scores for datasets that are served over SSL (or are otherwise
> automatically verifiable). Changing the questions and their scores isn’t
> something that’s supported by Open Data Census, because OKF wants to be
> able to provide global comparisons, which requires constant metrics. We’re
> going to fork the project to make the necessary changes, though, so that's
> an easy problem to solve, although it’s a shame to work somewhat outside of
> OKF’s ecosystem.
>
> I’m curious, though—does anybody else find that they want/need to change
> questions or scores? Or is it just us? :)
>
> Best,
> Waldo
>
> ---
> Director
> U.S. Open Data
> http://usopendata.org/
> 202-719-5315
> _______________________________________________
> open-data-census mailing list
> open-data-census at lists.okfn.org
> https://lists.okfn.org/mailman/listinfo/open-data-census
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/open-data-census/attachments/20150207/87f9e081/attachment.html>