DRW Notes: Max Planck Institutes

From InterSciWiki
(Redirected from Max Planck Institutes)
Jump to: navigation, search

Social Anthropology

Kinship group- Bettina Mann and Andre Thiemann

Also: User:Mihai Popa

Re: [Fwd: MPI for Social Anthropology Halle (Network Analysis)]] From: "Andre Thiemann" <thiemann@eth.mpg.de> Date: Fri, May 20, 2011 9:52 am To: drwhite@uci.edu Cc: "Tabea Scharrer" <scharrer@eth.mpg.de> (more) Options: View Full Header | View Printable Version | Download this as a file | View Message Details

Dear Doug,

thank you very much for the kind offer to meet in Leipzig. My co-organizer Tabea is doing fieldwork at the moment, so from the organizers I will come alone.

I would like to learn more about your ideas on network analysis, esp. how to best organize the learning for a group, in a workshop format or otherwise.

Maybe there would be some colleagues interested in the suggested training session, would you mind if I bring them along? Which day would you suggest for the meeting? Before Saturday 25th or after Sunday 26th of June?

As for the hardware and software requirements for long distance teaching. We have a computer pool with 8 PCs in the institute, Team Viewer was used some time ago there for another workshop. Some researchers are using MAC Laptops, others Laptops with Windows.

Best, Andre


Bettina Mann wrote:

...wer antwortet hier (in cc) an die anderen?

MPI for Social Anthropology Halle (Network Analysis)]
From: drwhite@uci.edu
Date: Tue, 17 May 2011 07:20:06 -0700
To: "Bettina Mann" <mann@eth.mpg.de
 
Dear Bettina, Tabea, Andre,

Short reply for the moment ... are you Mac Users or PC? And yes, TIPP is also
a good place for training, not for p-graph analysis which is simpler to
interpret (as in all the articles listed in the link below) but gives a rich
analysis of types of marriage and their overlaps.

Now that I have the Leipzig analysis organized it strikes me that since we
dont meet at Leipzig on the weekend of June 25th-26th, you three might like to
come down for the day to Leipzig with your laptops and we could have an
introductory training session. Bring others as you wish.
Best, Doug

Dear Doug,
thanks from your prompt reply and interesting information and hints.
Your suggestion to have a skype session would be also an excellent way
for exchange. We will discuss your suggestions and come back to you as
soon as possible.
I heard about the TIPP project which also seems to be very promising.
Best regards
Bettina

Demographic Research

Rostok Max Planck Institute for Demographic Research

The Causality Seminars

Hi Doug,

Let us have a seminar at 2pm on Monday (June 18th). Most likely, there will also be an archaeologist present, John Bintliff from the Netherlands, and that should lead to interesting discussions.

Best, Juergen

  • Monday 18th informal 9am Practicum: Using 3SLS modeling tools for the Mac and PCs
    • 2pm seminar: Overview, sample results, and sketch of a "modeling Causal graphs" perspective (John Bintliff, ...)
  • Tuesday 19th am or pm?: The Topology of Causality and Bias thru the conceptual lens of Sander Greenland
    • Not "states of nature" but endogenous model of changes in heterogeneous states of nature
    • Diagrams represented only to spot biases allowed by assumptions
    • DAG diagrams similar to kinship flows of ancestries, no cycles, converging arrows or "colliders" have no lateral transmission between respective "parents".
    • Idea is to take a subgraph inside the causal network.
    • Arrow absences imply null hypotheses.
    • To control for biases consider the graph as a flow diagram (directed piping), unblocking converging arrows on a node "opens" that valve
  • Wednesday 20th:
Send link to Greenland’s pdf
Stephan Poppe
Stephan.poppe@mis.mpg.de,
lizier@mis.mpg.de,
thoumar@mis.mpg.de,
mlynar@mis.mpg.de,
otbrich@mis.mpg.de,
ffl@mis.mpg.de,
ehrig@mis.mpg.de,
Hande.Tuncel@mis.mpg.de,
shiping@mis.mpg.de,
Lorenzj.taggi@mis.mpg.de,
Sarmient@mis.mpg.de,
Christian.rodriguez@mis.mpg.de

Causality Seminars SCHEDULE

That's perfect. Then if you agree we could set up the 10 days there half the day with some presentations, discussions and instructions in how to use the software, and the other half (before/after lunch) still open but more oriented towards actual research and use of the software.

(software and usage make me ask if your faculty/postdocs are mostly mac users, but also if there are a few PCs around that I will need to use occasionally for Fortran executables and new compilations. Also if you have any R programmers there)

On monday after lunch then I'll do an overview, sample results, and sketch to be discussed Tuesday on analyzing results in a "Causal Bayesian networks" modeling perspective.

Tuesday could feature a powerpoint pdf by Sander Greenlake, the leading teacher, exponent of of the mathematical definitions and theory, applications author, coauthor with Pearl, and critic of Causal Graphs and its related empirical and methodological issues.

Then would you want me to go on to sketch topics day by day or to have less of a schedule i.e., more open ended.

Would it be ok if I invited three more junior researchers from the Soc Anth MPI at Halle for Mon and Tues?

People are mostly worked with Linux here, some PCs and a Mac or so should also be available

> On monday after lunch then I'll do an overview, sample results, and sketch to be discussed Tuesday on analyzing results in a "Causal graphs modeling" perspective.

Overview on Monday after lunch is of course ideal. We could then have the causal graph models on late Tuesay afternoon. We should also have some time on Monday to discuss with John Bintliff and other archaeologists (not yet finalized, however)

> Tuesday could feature a powerpoint pdf by Sander Greenland, the leading teacher, exponent of of the mathematical definitions and theory, applications author, coauthor with Pearl, and critic of Causal Graphs and its related empirical and methodological issues.

Good, as mentioned above, on late Tuesady afternoon, we should discuss causal graphs, from different viewpoints.

> Then would you want me to go on to sketch topics day by day or to have less of a schedule i.e., more open ended.

Let's have it more open. In any case, some obligations will prevent me to be present all the time, and also, since I am personally not an expert on software, I shall of course skip those parts.

> Would it be ok if I invited three more junior researchers from the Soc Anth MPI at Halle for Mon and Tues?

That sounds like a good idea, because we then can gain a wider perspective, and I am looking forward to meeting the folks from Halle anyway.


Doug,

since I am not programming myself, I am also not directly working with R, but since this is, as I understand, the standard statistics language, there should be no problem.

BTW, I have also arranged for a visit by the archaeologist John Bintliff from Leiden during June 20/21, because there seems to some common ground for very interesting discussions.

Best Juergen

June 6 2011

Doug, here are some proposals for the program.

  • Can you give an introduction to the project for interested people on June 20

at 11, say?

  • At 2pm, there will probably be a lecture by the archeologist John Bintliff

which we hope will also find your interest.On June 21, say at 5pm, we could have the other planned lecture from your group (Ih ave a doctor's appointment earlier that afternoon).

  • On June 23, I'll have ohter seminars, but on June 23, between 3.15 an 5.30,
  • I could participate in a network meeting.
  • also, part of the afternoons of June 27/28 will probably be free (those dates are no longer blocked).
  • Best

Juergen

Mathematics in the Sciences

Preprint: Talk on "R Software for 2sls inferential statistics (2sls-is)" pdf and doc