[ddj] Graphing Obesity

Tue Jul 5 17:46:50 UTC 2011

Hi Roberto,

Thanks for this interesting dataset!

Have you tried using a log scale on this graph?
http://lists.okfn.org/pipermail/data-driven-journalism/attachments/20110705/f0d6816a/attachment.png

When dealing with proportions, log scales can help get closer to what the
data means.

That said, I've looked at the data and there seems to be little correlation
between the increase in GPD and the increase in obesity. There's actually a
negative correlation when plotting GDP increase with BMI increase. Below is
the result I got from STATA for the stats nerds among us:

---------------------------------------------------------------------------
     diffgdp |      Coef.   Std. Err.      t    P>|t|     [95% Conf.
Interval]
-------------+----------------------------------------------------------------
     diffbmi |  -67.15838   427.6736    -0.16   0.877    -965.6673
 831.3505
       _cons |   16566.22   6001.733     2.76   0.013     3957.049
 29175.39

Some evidence seems to point to a *negative *relation between obesity and
wealth (http://www.ncbi.nlm.nih.gov/pubmed/17258803 and
http://eurpub.oxfordjournals.org/content/17/1/53.full.pdf page 57).

Seems counter-intuitive, would require more research, I guess.

Best wishes,

nicolas.
(ps: We met at the World service when I was an intern there back in '07,
great to see you here!)
--
Datajournalist since 2007
nkb.fr <http://nkb.fr?m>
InfluenceNetworks.org <http://InfluenceNetworks.org?nkb>
+336 50 57 53 80

On Tue, Jul 5, 2011 at 6:58 PM, <
data-driven-journalism-request at lists.okfn.org> wrote:

> Send data-driven-journalism mailing list submissions to
>        data-driven-journalism at lists.okfn.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>        http://lists.okfn.org/mailman/listinfo/data-driven-journalism
> or, via email, send a message with subject or body 'help' to
>        data-driven-journalism-request at lists.okfn.org
>
> You can reach the person managing the list at
>        data-driven-journalism-owner at lists.okfn.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of data-driven-journalism digest..."
>
>
> Today's Topics:
>
>   1. Graphing Obesity (Body Mass Index >30) vs Gross National
>      Income per capita - Feedback needed (Roberto Belo-Rovella)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Tue, 5 Jul 2011 17:52:50 +0100
> From: Roberto Belo-Rovella <roberto at erbr.net>
> Subject: [ddj] Graphing Obesity (Body Mass Index >30) vs Gross
>        National Income per capita - Feedback needed
> To: data-driven-journalism at lists.okfn.org
> Message-ID:
>        <CALa1hNmJJw9975DAT0mEZ95-EMWMRnrOKkHY__tNs6nPHgRQJw at mail.gmail.com
> >
> Content-Type: text/plain; charset="utf-8"
>
> Hello, my name is Roberto Belo and I work for the BBC World Service. I
> need your statistics expertise on the project we are working on.
>
> We are looking into a set of data provided by the WHO, which we
> complemented with GNI per capital (Atlas method) from the World Bank:
> http://bit.ly/kF6fyr . We are just looking at the period from 1980 to
> nowish, with data from 1980 (except for Russia), 1990, 2000 and 2008.
>
> We are trying to show the relation between a higher income and the
> increase in Obesity (% of population with BMI bigger than 30),
> especially focusing on the G7 countries on one side, and on the BRICS
> countries (Brazil, Russian, India, China, South Africa) on the other.
>
> 1) Originally we produced this graphic, showing the variation of
> Obesity vs GNI per Capita, and as the variations within the years seem
> to be linear, we just included 1980 and 2008 (except for Russia, we
> took it from 2000) - NB Just focus on the graphic, the rest is
> furniture.
>
> <<obesity_bmi_v_gni.png>>
> This shows that the increase in obesity cases have been steeper in
> BRICS countries than in G7 ones.
>
> 2) This one shows Obesity vs the % variation of GNI per Capita - and
> here the picture seems to change radically, with a mixed picture.
>
> <<obesity_bmi_v_percent_increase_gni.png>>
> 3) Finally, this one, which shows the % variation in Obesity vs the %
> variation of GNI per Capita - which I think falls into very muddy
> waters, but hey, the designer was very enthusiastic
>
> <<obesity_percent_increase_bmi_v_percent_increase_gni.png>>
> Given all these, would someone be able to shed some light into which
> of these three options would be the most honest and statistically
> correct way of showcasing the data - and why?
>
> Otherwise, if you could point me in the direction of someone who would
> be able to make total sense of this, I would be very grateful.
>
> We have been putting a lot of effort in getting data journalism into
> the agenda, and we can't afford a mistake in this presentation.
>
> Thanks a lot.
>
> Roberto
>
> Roberto Belo-Rovella | BBC WS Future Media Editorial
> Room 500NE, Bush House, London WC2B 4PH
> Phone: +44 20 755 70611 (Ext: 02 70611)
> E-mail: roberto.belo at bbc.co.uk
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: obesity_bmi_v_gni.png
> Type: image/png
> Size: 281334 bytes
> Desc: not available
> URL: <
> http://lists.okfn.org/pipermail/data-driven-journalism/attachments/20110705/f0d6816a/attachment.png
> >
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: obesity_bmi_v_percent_increase_gni.png
> Type: image/png
> Size: 287957 bytes
> Desc: not available
> URL: <
> http://lists.okfn.org/pipermail/data-driven-journalism/attachments/20110705/f0d6816a/attachment-0001.png
> >
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: obesity_percent_increase_bmi_v_percent_increase_gni.png
> Type: image/png
> Size: 280141 bytes
> Desc: not available
> URL: <
> http://lists.okfn.org/pipermail/data-driven-journalism/attachments/20110705/f0d6816a/attachment-0002.png
> >
>
> ------------------------------
>
> _______________________________________________
> data-driven-journalism mailing list
> data-driven-journalism at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/data-driven-journalism
>
>
> End of data-driven-journalism Digest, Vol 4, Issue 2
> ****************************************************
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/data-driven-journalism/attachments/20110705/5a33407c/attachment.html>