Writing a paper on network reduction for "Ethnography kitchen"

Hello @Jan, @Richard @melancon @bpinaud

Our paper was rejected by SM&R:

I read your manuscript with great interest and found the ideas intriguing. However, the manuscript is written at a level of abstraction that I think unlikely to aid social scientists. That is, it contains too many broad assertions without proof, explanation, or applied examples. This makes it difficult to ascertain what has been learned, or how other scholars might apply your method to their own cases.

Any ideas? Should we try to re-read it with this feedback in mind? Or simply resubmit somewhere else?

Should we try to get some more detailed feedback from the editor, who wrote that email? He is Felix Elwert, University of Wisconsin-Madison.

Bummer! It sounds as if we’d be able to sell our paper more easily if we included empirical examples which show how the network reduction works in practice. The Polish data from POPREBEL is probably the richest. Could we use that as a case study? As @Jan is knee-deep in revisions to deliverables, we could ask the Polish ethnographer and the Polish coder to help with this. Let me discuss this with them at tomorrow’s weekly meeting. R

I had the same fear.

Great idea.

Dear @Maniamana and @Wojt

As you may know, @Jan, @Alberto and I, along with two French colleagues, have written an article about the Edgeryders methodology and, in particular, about how to reduce networks of ethnographic codes co-occurrence (attachment below). The editors of Sociological Methods and Research liked the article but thought it was too abstract. So, we now think we need to include empirical examples which show how the network reduction works in practice. In this connection, we think that the Polish data from POPREBEL would make a good case study. Would you be able to help us with this?

@Alberto, can you explain what we would need Mania and Wojt to do?

Sociological_Methods___Research_paper.pdf (12.5 MB)

1 Like

PS: @alberto

I found another potential journal: https://journals.sagepub.com/home/smx

Dear All, I am somewhat overbook as you might have noticed, but I will gladly help in expanding this paper. But either @Wojt or @Maniamana will have to lead the project. And @alberto I would love to chat with you soon. Let me know when you are available. Ciao!

This is quite tricky, because we risk watering down the methodological soul of the paper; and because we are at the beginning of writing the final report for WP2, based on the coding. Hopefully that report will be interesting enough to justify its own paper – however, in this case, not a paper on methods, but a paper on populism.

@melancon, @bpinaud, do you see another outlet for the paper as is, in the NetSci space?

How about Thursday afternoon CET or Friday after the ethno call?

I see https://www.cambridge.org/core/journals/network-science or https://appliednetsci.springeropen.com/

1 Like

Hi @alberto

I’m not sure I understand your point about ‘watering down the methodological soul of the paper’. Wouldn’t demonstrating the practical value of the different reduction methods strengthen the claims the article is making? Or have I misunderstood what you meant?

But happy to go with one of the NetSci journals if they would accept the article as is.

1 Like

I explained myself very poorly. What I was trying to say is this: the present form of the paper focuses on techniques for reducing large networks of co-occurrences netween ethnographic codes. It uses data from three projects, and tries to present these techniques as generalizable, independently of the specific study that each dataset comes from. The paper is content-blind.

Parts of it are appropriately so. The effectiveness of each technique at reducing the number of nodes and edges is independent of the codes those nodes represent. Using three datasets, we show exactly that techniques work in the same way across different studies, and produce different effects.

If we bring content back into the picture, we need to add a lot of length to support sub-section Mapping network reduction techniques onto four major approaches in sociology and anthropology. It’s almost a second paper, growing inside of the existing one.

Unless… @Richard, @Maniamana, @Wojt: do you think it is possible to find, say, two examples per each technique, if I take your data and apply each technique to the “mother” graph?

I have to bail on that, @Jan, sorry. A sudden family event dictates that I travel back to Italy, leaving now and back on Sunday.

Ah, ho capito! On reflection, then, it will be more straightforward to try one of the Network Science journals Bruno suggested.

I changed my mind on this, @Richard. I now believe that, with the correct emphasis, we can indeed improve the article this way. @Maniamana , @Wojt, can you make it to Monday’s seminar? Things will become much clearer after we present the paper.


Sure, looking forward to getting the link for that!

Eranda sent the link to everyone. Anyway, it’s here:

Helllo @icqe22_authors, reviving this thread as I am preparing to submit the latest version of this paper to Applied Network Science. But I have a strategic problem, and would like your opinion on it.

Recall how the editor of Sociological Methods & Research had suggested that our paper would make a lot more sense with an example. Well, we now do have that example: it’s the analysis we submitted to ICQE22. However, in the ICQE22 paper, the description of the reduction methods themselves is much less developed than in the SM&R paper. So, we discussed a strategy was to revive the SM&R paper, and graft to it a section Application, which would be lifted from the ICQ22 paper.

The strategic problem stems from the issue of text recycling. The SM&R paper was never published, but some parts of it were used for the iCQE22 one, which is also, at this time, not published, but it could be (after revision). The new ANS paper would then consist of two parts: one introducing the method (partially recycled) and one with the example (totally recycled). The situation is complicated by the fact that, of course, ICQE22 is not a journal.

To the more experienced academics in the group: what do you think we should do?

1 Like

Conference papers are basically expected to be published in journals. As long as the ICQE paper does not in and of itself become a basis for another publication, I think that should be fine.


The journals on which I am a member of the Editorial Board would not consider articles, large chunks of which exist elsewhere. (Of course, there’s no definition of what constitutes a ‘large chunk’.) The Applied Network Science paper would thus have to offer some data or analysis that is new. From an academic perspective, a journal article is worth much more than appearing in ICQE22. Are you required to publish with ICQE22?

I am not sure, Richard. Here is an overview of the Frankenpaper:

ICQE22 paper ANS paper
Introduction same same
Related work same same
The codes co-occurrence network and its interpretation shorter longer
Data and pre-processing no yes
Techniques for network reduction:
What makes a good technique for network reduction? slightly shorter slightly longer
Association depth short much longer, with figures
Association breadth short much longer, with figures
Highest core value short much longer, with figures
Simmelian backbone short much longer, with figures
Comparing reduction techniques no yes
An application same same
Discussion and conclusions same same

It is basically the same paper, with an extra (quantitative) comparative analysis and some more detail in describing the four techniques.

Hi @alberto

As this is your baby, the final decision about where to publish should absolutely be yours but, speaking as an academic, publishing in conference proceedings (which is how I have understood the ICQE22 output to be) is of no value. If I were you, I would write to the editor of ANS and ask whether the proposed article submission would be considered sufficiently different from the potential ICQE22 publication (as per your table above) to constitute an original article. If not, I would prioritise ANS. But, of course, the decision is yours.