[Futuregen] Fwd: Re: Examplary graphs for the Data Navigator discussion

[rodrigues at euro.centre.org]

Dear all,

Susan raised some very good points already regarding the Data Navigator 
and the data we shared - please see below.
We thought it would be good to share these with you in advance of the 
zoom on Thu as you may have had similar doubts.

Looking forward to the discussion.

Best wishes,
Ricardo

-------- Original Message --------
Subject: Re: [Futuregen] Examplary graphs for the Data Navigator 
discussion
Date: 2019-12-17 17:19
 From: Susan Phillips <noasue at gmail.com>
To: Ricardo Rodrigues <rodrigues at euro.centre.org>

finite copy all. I will hold off writing more and enjoy the discussions
tomorrow.

Susan

On Tue, 17 Dec 2019 at 11:10, <rodrigues at euro.centre.org> wrote:

> Hi Susan,
> Thanks a lot for your thoughts. They are very helpful to go forward
> and
> to some extent reflect some of our own questions.
> I copy Eszter and Stefania here and if you agree I would also copy the
> 
> rest as this is very relevant for the zoom on Thu. Would this be OK
> with
> you?
> My comments are in the text below.
> 
> On 2019-12-17 15:27, Susan Phillips wrote:
>> Hi Ricardo,
>> 
>> The graphs are a huge amount of descriptive info. Do you think
> others
>> will use them or are we putting them on the data navigator because
> we
>> said we'd do so?
> Most of what I will say pertains to care, on which our knowledge is
> greater, but I think some of this is also applicable to
> health/functionality.
> You would be surprised at just how little is systematically reported
> on
> a comparative basis on care-giving and receiving.
> I think that having trends of care-giving or receiving by gender and
> socio-economic SEC would be used quite a lot. Even more if we can
> represent these by cohorts. Many databases in Europe did not even
> covered care till the 2000s.
> I think the same applies to ADLs and IADLs, again especially if
> represented by cohort.
> 
> It is a bit trickier for some of the determinants (like education by
> gender).
> For some, Eurostat (the statistical office of the European Union)
> already publishes data on trends and there is absolutely no point in
> replicating some of that data.
> Some of the data may not even make sense to be presented in a cohort
> manner (e.g. income quartiles). However, given the scope of FUTUREGEN,
> 
> it might be useful to see if the gender parity in education groups or
> income (but how to account for the fact that we have more women than
> men
> at higher age groups?). Or age-differential at marriage and living
> arrangements between men and women or among those with different
> socio-economic condition (or intersection of the two?).
> 
>> 
>> I now understand why data from Canada, even if it is not from the
>> CLSA, should also be on the Data Navigator. However, am not sure we
>> have the person-power here to do the same with the GSS or CCHS data.
>> Would this actually be part of WP 1 or 2 and something others might
>> do? I know I sent you one of the datasets (I forget which) and, of
>> course, could resend it).
>> 
> Famous last words: once we have figured out the representation, to
> re-run this on other datasets would not be very time consuming.
> But, for the most part, WPs 1 and 2 (and perhaps WP 4) would bear the
> brunt of the descriptive.
> One thing we worry about though is which Canadian data we would be
> able
> to use: CLSA might not be usable, GSS (if we recall correctly) has no
> information on ADLs and IADLs and care is defined differently as well,
> 
> CCHS I don't remember.
> If we send you the definition of care for SHARE, do you think Afshin
> could look it up and see if they are somewhat similar to those used in
> 
> CLSA, GSS and/or CCHS data?
> 
>> We are really working on the systematic review and thinking through
>> strongest but least complicated methods for getting at
>> intersectionality. It has been very interesting but reading through
>> and thinking through the non-regression type methods is complicated
>> and slow. Sneak preview of our thoughts to date - no matter what
>> method is used sex-disaggregation of the data is key to
> understanding
>> intersectionality. The method, itself, may be less important that
>> this.
> 
> OK, my turn, and this might be a very silly question (apologies in
> advance if it is so) :): "sex-disaggregation of the data is key to
> understanding intersectionality", wouldn't we expect that to be so?
> Can you even analyze intersectionality (or gender differences for that
> 
> matter) without such disaggregation?
> That the methods would not necessarily be that relevant - or yield
> similar results (but that is for the proof of concept, I guess) - or
> that they each capture different aspects/angles of intersectionality
> is
> more a bit more interesting.
> But we would be keen on hearing more!
> 
>> 
>> Susan
>> 
>> On Mon, 16 Dec 2019 at 11:37, Ricardo Rodrigues
>> <rodrigues at euro.centre.org> wrote:
>> 
>>> Dear colleagues,
>>> 
>>> Ahead of our zoom on Thursday we would like to share with you some
>>> graphs that would stand as possible examples for the data
>>> representation in the Data Navigator.
>>> 
>>> The original idea of the Data Navigator is to represent trends and
>>> evolution of "outcomes" (e.g. ADLs and IADLs, care giving and
>>> receiving) as well as their determinants (e.g. how education
>>> attainment, living arrangements, etc are evolving) across time,
>>> place (countries), cohorts and gender. Attached are files with
>>> different graphic representations for care giving and receiving
>>> based on SHARE data.
>>> 
>>> More specifically:
>>> 
>>> We have thought of depicting this information initially through
>>> traditional trend lines across time. These are represented in the
>>> file "FC_trend_country" attached for formal care receiving.
>>> 
>>> After seeing Stefan and Johan's cohort graphs, however, we thought
>>> that these might provide a better portrait of the information we
>>> want to convey. We therefore put together similar cohort graphs for
>>> care receiving (formal and informal) and giving (informal only) for
>>> several cohorts, by gender and by education. These you will find in
>>> file "FC/IC/ICgave_country_gender" and
>>> "FC/IC/ICgave_country_education" attached. The data is presented by
>>> gender and by education levels for all countries that participated
>>> in at least 3 waves of SHARE. All lines have been smoothed using 5
>>> year groups (i.e. we estimated the prevalence rates for 5 year,
>>> rather than year-on-year). The figures in the legends (e.g. 35-44
>>> female) refer to the years of birth rather than the age of the
>>> cohorts.
>>> 
>>> Some "sensitivity graphs":
>>> 
>>> We also tested the possibility of presenting data for care
>>> receiving not only for the total population in that
>>> cohort/age/gender group - which is what is shown in all the graphs
>>> above - but also for those with care needs only (defined as having
>>> at least on ADL, OR one IADL, OR difficulties in mobility). We did
>>> this for formal care receiving in the file "FC_country_gender_need"
>>> attached.
>>> 
>>> As you will notice, the graphs on education do not always result
>>> very readable. The data depicted in many of them refers to very
>>> small numbers as those with primary and tertiary education are a
>>> relatively small group for these cohorts. We therefore depicted the
>>> graph for formal care received by education with only the upper and
>>> lower education groups in the file "FC_country_education_2"
>>> attached.
>>> 
>>> To overcome the issue of small numbers for those with primary and
>>> tertiary education, we also calculated the prevalence rates for
>>> formal care use for different income and wealth quartiles (1st -
>>> poorer - and last - richer).  These you will find in the file
>>> "FC_country_wealth_1" and "FC_country_income_1" attached.
>>> 
>>> Finally, we also thought it could be interesting to have the data
>>> above depicted not only per country (i.e. each graph refers to a
>>> country), but to combine different countries in one graph to
> compare
>>> the evolution across countries. An example of this for formal care
>>> receiving is represented in the file "FC_bycountry_AT_DESW"
> attached
>>> (for Austria, Germany and Sweden only as examples).
>>> 
>>> We know this is a lot of information to digest so we thought of
>>> detailing what we are sending you as much as possible. If you have
>>> any questions, please do not hesitate to contact us ahead of the
>>> zoom. We will also post this in FREEDCAMP and tomorrow we will have
>>> the agenda for the zoom there as well.
>>> 
>>> Best wishes,
>>> 
>>> Ricardo
> 
> --
> Ricardo Rodrigues, PhD, MSc
> Deputy Director
> Head of Health and Care Unit
> European Centre for Social Welfare Policy and Research
> Berggasse 17,
> 1090 Vienna, Austria
> Tel. 0043-(1)319-4505-35
> www.euro.centre.org [1]
> Follow us on Twitter! @EuroCentre_SWPR


Links:
------
[1] http://www.euro.centre.org

-- 
Ricardo Rodrigues, PhD, MSc
Deputy Director
Head of Health and Care Unit
European Centre for Social Welfare Policy and Research
Berggasse 17,
1090 Vienna, Austria
Tel. 0043-(1)319-4505-35
www.euro.centre.org
Follow us on Twitter! @EuroCentre_SWPR