Check rejected papers for manual effect size option

[kfrasier]

The combined list of papers we're considering is here: https://docs.google.com/spreadsheets/d/1LMyiHs5ptr3YZEc9_k72Xni4Ps7uXh-OUZi1d3yydzc/edit#gid=1765537852

In the "Relevance" column, the reason for rejecting papers is listed. We need to review papers that say "No effect size" to see if there's enough information to calculate an effect size manually from the information provided (e.g. from data in tables or supplementary info). It might be worth scanning the "No biomass" papers too, but those are less likely to have what we need.

If a paper is changed from a "no" to a "yes" then it should be added to GoodPapers_April2018.xls.

[kfrasier]

I started saving pdfs in /Meta-analysis search CSVs/Meta-2018/papers/ but there's a long way to go.

[ailich]

@kfrasier @berisman what data needs to be in the papers in order to calculate effect size?

[kfrasier]

@ailich I think we need some measurement of CPUE or a related biomass metric, under at least two conditions of whatever driver is being investigated (e.g. reef vs. no reef). Often there will be a table of measurement values, and if the authors provide means, and some metric of variability (CV, std. dev., etc), then we could conceivably compute our own effect size from those pieces. If there are values relating to sample size or effort, those values would be very helpful/possibly essential for the calculations, depending on what exactly is being measured. The values definitely need to be measuring red snapper, we're not considering other closely related species or numbers that include red snapper along with other species lumped together. As I recall, there were a lot of papers where this lumping together an the issue. @berisman, please correct me if I'm missing anything here.

[kfrasier]

@ailich Heads up: The link in the description for this has been updated, to point to the latest and greatest spreadsheet.

[ailich]

Ok, do the environmental variables need to be in groupings? What if the study is not a factorial design and has environmental variables are measured on continuous scales?

[kfrasier]

If they have measurements of red snapper biomass in relation to continuous measurements of some variables, then we could definitely use that. I think we could calculate an r^2 from it. We would probably need the values of both the environmental variable and the response (snapper) to do it.

[berisman]

Sorry guys - had to check out of the snapper business for this week to catch up on things prior to the workshop. I’ll try to respond to the meta- questions in short order. I should have time on Friday to get back to you.

In the meantime, I had a very productive conversation with Steve Szedlmeyer about red snapper. As you know, he’s probably published more papers than nearly anyone on red snapper biology.

I briefly described our snapper projects to him and mentioned that we need to get ‘em done and submitted. He laughed and said he is finishing up a book on Red Snapper (edited by him and Steve Bortone with chapters written by various experts; CRC Press) and is in need of a few chapters. Briefly, all submitted chapters undergo peer-review and a few submissions were rejected, which opened up a few slots.

He was very excited about what we are doing and invited us to submit a chapter or both if we wanted. He’d want a draft submitted by June 30th, however. Yikes! That said, it would help us meet our deadline. Also, I bet we can push that out a bit. After all, it’s a book, and these things are always late.

Anyway, I thought I’d share this with the group in case we are interested in publishing one or both papers in a book rather than a journal. We’ll have to decide quickly.

As a side note, he did also have very strong feelings about the validity of SEAMAP bottom trawl data as a proxy for red snapper distributions. I won’t belabor the point other than to butcher his quote: “It’s like counting mountain goats down in the desert”. 🐫 I’ll entertain you with the longer version next week.

See y’all soon!

Brad

Brad Erisman Assistant Professor of Fisheries Ecology Marine Science Institute, University of Texas at Austin 750 Channel View Drive Port Aransas, TX 78373 Phone: (361) 749-6833 Email: berisman@utexas.edu Web: https://fisheries.utexas.edu

On May 26, 2018, at 10:49 PM, Kait Frasier notifications@github.com wrote:

@ailich https://github.com/ailich I think we need some measurement of CPUE or a related biomass metric, under at least two conditions of whatever driver is being investigated (e.g. reef vs. no reef). Often there will be a table of measurement values, and if the authors provide means, and some metric of variability (CV, std. dev., etc), then we could conceivably compute our own effect size from those pieces. If there are values relating to sample size or effort, those values would be very helpful/possibly essential for the calculations, depending on what exactly is being measured. The values definitely need to be measuring red snapper, we're not considering other closely related species or numbers that include red snapper along with other species lumped together. As I recall, there were a lot of papers where this lumping together an the issue. @berisman https://github.com/berisman, please correct me if I'm missing anything here.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/NCEAS/oss-fishteam/issues/28#issuecomment-392303415, or mute the thread https://github.com/notifications/unsubscribe-auth/Aa25neX8ikEuEpmtjc4CUryrJUROJM8jks5t2iJFgaJpZM4TVjr5.