[R-sig-Geo] Shiny app for Spatial Econometric classes

Roger Bivand Roger@B|v@nd @end|ng |rom nhh@no
Mon Jul 12 11:16:16 CEST 2021


Perhaps it is just my connection, but it is not possible to load any data, 
so I cannot evaluate the other components in practice.

I do however think that you need to acknowledge the results of Piras and 
Prucha, https://doi.org/10.1016/j.regsciurbeco.2014.03.002, and try to 
point out the dangers of pre-testing, rather than estimating the most 
inclusive model first, and simplifying from there.

Roger

On Fri, 9 Jul 2021, Roger Bivand wrote:

> Returning to the "Tobler" shiny app:
>
> The main serious weakness is the name and the approach. It is not and has 
> never been the case that "Everything is related to everything else, but near 
> things are more related than distant things." This is and has always been an 
> oversimplification, disguising the underlying entitation/support problem. Are 
> the units of observation appropriate for the scale of the unobserved spatial 
> process?
>
> Tobler (1970) https://doi.org/10.2307/143141 was published in the same SI as 
> Olsson (1970) https://doi.org/10.2307/143140 (Olsson precedes Tobler), but 
> Olsson does grasp the important point that spatial autocorrelation is not 
> immanent in spatial phenomena, but often is engendered by inappropriate 
> entitation, by omitted variables and/or inappropriate functional form. The 
> key quote from Olsson is on p. 228:
>
> "The existence of such autocorrelations makes it tempting to agree with 
> Tobler (1970, 236 [my interpolation, the original refers to a conference 
> paper]) that 'everything is related to everything else, but near things are 
> more related than distant things.' On the other hand, the fact that the 
> autocorrelations seem to hide systematic specification errors suggests that 
> the elevation of this statement to the status of 'the first law of geography' 
> is at best premature. At worst, the statement may represent the spatial 
> variant of the post hoc fallacy, which would mean  that coinincidence has 
> been mistaken for a causal relation."
>
> The status of the "first law" is very similar to the belief that John Snow 
> induced the cause of cholera as water-borne from a map. It may be a good way 
> of selling GIS, but it isn't accurate; Snow had a strong working 
> microbiological hypothesis prior to visiting Soho, and the map was prepared 
> after the Broad street pump was disabled as documentation that the hypothesis 
> held (see https://doi.org/10.1016/s0140-6736(00)02442-9 Brody et al. 2000).
>
> So the framing of the shiny app is unfortunate, and spatial econometrics is 
> not a toolbox, rather a way of loooking at mis-specification when space is 
> involved, I think.
>
> Hope this isn't too critical,
>
> Roger
>
> On Thu, 8 Jul 2021, Raphael Saldanha wrote:
>
>>  Estimado Spencer,
>>
>>  Thanks for your kind words. Actually, that construction is not standard in
>>  Portuguese also (although that does not sound pedant or with some bad
>>  intention for our ears). I revised the text three times and used an app
>>  for grammar correction before sending. Anyway, it was a mistake by all
>>  means.
>>
>>  To learn Spatial Econometrics and other fields require from us additional
>>  efforts. There is just a few books or materials in Portuguese. But I try
>>  to approach this difficult as an opportunity to learn.
>>
>>  Indeed, one of the reasons for building the app was to facilitate the use
>>  of the methods for people that are learning R. *Professor Eduardo and I*
>>  teach annually a course on that field and we face this challenge. With the
>>  pandemics and Zoom classes, we hope that the app and our scripts will help
>>  the students to learn R.
>> 
>>
>>  Um forte abraço,
>>
>>  Raphael Saldanha
>>
>>>  Em 8 de jul. de 2021, à(s) 08:45, Spencer Graves
>>>  <spencer.graves using prodsyse.com> escreveu:
>>>
>>>  Estimado Raphael:
>>> 
>>>
>>>     I don't know any Portuguese either, and I'm very grateful that you've
>>>     made the effort to ask this group in a language not your own.
>>> 
>>>
>>>     In primary school, I learned that constructions like, "Me and
>>>     Professor ..." too often sounded like, "Mean Professor ...".
>>> 
>>>
>>>     Un fuerte abrazo ("A big hug", as a Uruguayan man whom I had not
>>>     previously met ended a letter inviting me to talk at a conference).
>>> 
>>>
>>>     Spencer Graves
>>> 
>>>
>>>  On 7/8/21 5:32 AM, Raphael Saldanha wrote:
>>>>  No offense taken. Thanks for the remark and opportunity to learn.
>>>>  Cheers,
>>>>  Raphael
>>>>>  Em 7 de jul. de 2021, à(s) 23:20, Rolf Turner <r.turner using auckland.ac.nz>
>>>>>  escreveu:
>>>>>
>>>>>
>>>>>>  On Wed, 7 Jul 2021 18:09:50 -0300
>>>>>>  Raphael Saldanha <rfsaldanha using gmail.com> wrote:
>>>>>>
>>>>>>  Hello all!
>>>>>>
>>>>>>  Me and Professor Eduardo Almeida (UFJF, Brazil) are creating an R
>>>>>>  Shiny app for Spatial Econometric classes.
>>>>>
>>>>>  <SNIP>
>>>>>
>>>>>  I hope that you won't find this comment offensive, but I could not
>>>>>  resist sending it to you, since I am rather fanatical (pedantic?) about
>>>>>  correct English usage.
>>>>>
>>>>>  The expression "Me and Professor Eduardo Almeida" is both (a) bad
>>>>>  grammar and (b) bad form.  (a) Instead of the pronoun "Me" (accusative
>>>>>  or
>>>>>  objective case) you should use "I" (nominative or subjective case),
>>>>>  since
>>>>>  this expression is the subject of the sentence.  (b) It is "bad form"
>>>>>  (it sounds vaguely self-promoting)  to put a first person pronoun
>>>>>  before the
>>>>>  names of others to which it is linked.  You should say "Professor
>>>>>  Eduardo Almeida and I are creating ....".
>>>>>
>>>>>  I realise that English is surely a second language for you, and your
>>>>>  English is infinitely better than my non-existent Portuguese.  Your
>>>>>  English is actually quite good, probably better than that of many
>>>>>  native English speakers.  Nevertheless the expression that you use is
>>>>>  not correct and you should strive to avoid such usage.
>>>>>
>>>>>  Again let me express the hope that you won't take offence at this
>>>>>  comment.
>>>>>
>>>>>  cheers,
>>>>>
>>>>>  Rolf Turner
>>>>>
>>>>>  --
>>>>>  Honorary Research Fellow
>>>>>  Department of Statistics
>>>>>  University of Auckland
>>>>>  Phone: +64-9-373-7599 ext. 88276
>>>>>
>>>>>  _______________________________________________
>>>>>  R-sig-Geo mailing list
>>>>>  R-sig-Geo using r-project.org
>>>>>  https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>>>>  _______________________________________________
>>>>  R-sig-Geo mailing list
>>>>  R-sig-Geo using r-project.org
>>>>  https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>> 
>> 
>
>

-- 
Roger Bivand
Emeritus Professor
Department of Economics, Norwegian School of Economics,
Postboks 3490 Ytre Sandviken, 5045 Bergen, Norway.
e-mail: Roger.Bivand using nhh.no
https://orcid.org/0000-0003-2392-6140
https://scholar.google.no/citations?user=AWeghB0AAAAJ&hl=en


More information about the R-sig-Geo mailing list