Sebastian Karcher
LightNews LightNews
banner
adam42smith.bsky.social
Sebastian Karcher
@adam42smith.bsky.social
1.2K followers 740 following 1.7K posts
Director, Qualitative Data Repository (personal account).
Data, Zotero, Social Science Methods
https://sebastiankarcher.com
Posts Replies Media Videos
adam42smith.bsky.social
It's a weird piece - I'm actually mentioned by name in it (based on an email w/o permission), and I did say that sociologists aren't active users of QDR, but added a caveat. This would have allowed for an interesting comparison (I don't think UK sociologists are further to the right?) 1/3
Although I largely agree with her analysis, it is worth noting that qualitative scholars in other social science fields, especially in political science, have been far more serious than sociologists about improving transparency and data sharing. For example, I have discovered in my conversations with peers that very few sociologists know about the Qualitative Data Repository – a dedicated archive for storing and sharing data generated or collected through qualitative research in the social sciences. My impression was confirmed in an email by Sebatian Karcher, the current director of the QDR. Given the prominence of qualitative research in sociology, it is worth asking why it took political scientists to spearhead this initiative to improve transparency and reproducibility in qualitative research (Elman & Kapiszewski, 2018). Yes, that’s absolutely in line with my impression in the US (and QDR in particular) – I think it looks differently when you go to the UK. Sebastian
November 3, 2025 at 9:23 PM
adam42smith.bsky.social
When they say "plummeted" they really do mean _plummeted_. I'm not even sure there's a precedent for a >30ppt drop in attendance . E.g. Met Opera attendance fell from 75% pre-COVID to 61% in the first post-COVID season 2021-22 (and has since gone back up to 72%).
Gift link: wapo.st/4hxOdFo
Since early September, 43 percent of tickets remained unsold for the typical production. That means that, at most, 57 percent of tickets were sold for the typical production — and some tickets may have been “comps,” which are given away, often to staff members or the press. That compares with 93 percent sold or comped in fall 2024 and 80 percent in fall 2023.
October 31, 2025 at 3:25 PM
adam42smith.bsky.social
We have a farming family at temple & because the Torah portion is Noah, we had farm animal havdalah -- all of which is to say, I recommend goats. This is "Vanilla"
A white goat surrounded by lots of legs looking right into the camera, big ears pointing straight out to both sides
October 26, 2025 at 2:25 AM
adam42smith.bsky.social
We have just swapped out "legally" for "traditionally" in media coverage. I guess new indicator for the state of the rule of law just landed?
At least four Republican senators, including the Senate majority leader, Senator John Thune of South Dakota, publicly signaled that they would have opposed his nomination to the office, which is a traditionally independent corruption-fighting agency that safeguards federal whistle-blowers and enforces some ethics laws. Traditionally, the United States has dealt with suspected maritime smuggling as a law enforcement problem, using the Coast Guard — sometimes assisted by the Navy — to interdict boats. That follows a familiar pattern: Police officers arrest people they suspect of dealing drugs; it would be a crime for an officer to instead summarily kill such a suspect.
October 24, 2025 at 7:37 PM
adam42smith.bsky.social
This is an observational study & their main model has worrying 39 covariates, etc. etc. so treat with appropriate caution, but the effect sizes are staggering. Look at these survival graphs (Y axis is 'overall survival %).
Survival rates of patients with & without COVID 19 mRNA vaccine for 5 types of cance patients. Survival rates after 40 month are about 20 percentage points higher for patients who received the vaccine
October 24, 2025 at 12:44 AM
adam42smith.bsky.social
My younger has helpfully provided a DAG
An abstract kid's drawing with lots of lines in lots of colors
October 23, 2025 at 11:22 PM
adam42smith.bsky.social
The Open Pool Fellowship for #IQMR 2026 call is open:

docs.google.com/document/u/1...

Each year, w support from NSF, we fund ~25 early career researchers to attend the 2-week Institute for Qualitative & Multi-Method Research for free. More on IQMR: www.maxwell.syr.edu/research/cen...
Jim Mahoney being very happy about giving a lecture about qualitative methods to a full auditorium
October 22, 2025 at 8:40 PM
adam42smith.bsky.social
Here's how the settings should look. And here is an example item from our library that has a PDF attached & accessible to members (open on a desktop computer, not mobile) - you can see the PDF in "Attachments", but it's greyed out and you can't access it: www.zotero.org/groups/48771...
LIbrary Settings for a Zotero group with Group Type set to Public Closed
October 15, 2025 at 1:52 PM
adam42smith.bsky.social
There's a freely accessible version here: www.fouirnaies.com/s/fouirnaies... (the paper is fairly technical, though, depending on your background might also not help much).

The core logic is in these two paras, more details later in the paper
of the primary does not change, divisive primaries prolong the period during which the eventual nominee is still unknown. In this way, the runoff’s increased length is a good approximation for a divisive primary. Nevertheless, in subsequent sections we attempt to unpack the various mechanisms by which runoffs affect general-election outcomes, and our evidence suggests that it is more the divisiveness, itself, rather than the added length of the campaign that drives the effects Are runoffs an appropriate proxy for “divisive primaries” more generally? This is a key question for thinking about how to interpret the estimates we present below. First and foremost, runoffs clearly do make the primary more divisive. Using the definition of divisive primary from Hacker (1965), namely, any primary in which the winning candidate’s vote share is less than 65%, every runoff is by definition a divisive primary.4 But the main issue is that runoff primaries do not only make the primary more divisive, in the sense of increasing and pro- longing close competition between candidates, but they also make it last longer. This is slightly different from divisive primaries, which do not technically alter the official end date of the primary election the way runoffs do. However, many di- visive primaries, especially for president, do actually extend the length of the de facto campaign. Although the official end date
October 15, 2025 at 3:45 AM
adam42smith.bsky.social
Just putting the obligatory Utrecht before & after photo here... against a politics of hopelessness 😉
Two photos of attracted. One in black and white. One in color. The black and white photo shows lots of big rows and lots of concrete. The color picture shows the same part of the town but now lots of green people riding boats, trees etc
October 12, 2025 at 9:05 PM
adam42smith.bsky.social
It's certainly not a simple story, but there's good evidence for trade offs of all kinds. This is one of my favorite articles e.g. www.sciencedirect.com/science/arti...
1/2
Among 5 808 passing bicyclists, there were large differences in helmet and light use: only 2.2% of Paris bicyclists wore helmets compared to 31.5% in Boston. In contrast, 46.8% of nighttime Paris bicyclists had working head or tail lights compared to only 14.8% in Boston.
October 12, 2025 at 8:45 PM
adam42smith.bsky.social
Yes, likely system locale. Zotero lets you override that, I don't think Zbib does.
Best workaround I can think of is to use a style that fixes the language.
There's a "University of York - Modern Language Association" style that almost works, except for UK style quotation marks
Zotero bib bibliography with University of York - Modern Language Association style. Watson, J. D., and F. H. C. Crick. “Molecular Structure of Nucleic Acids: A Structure for Deoxyribose Nucleic Acid.” Nature, vol. 171, no. 4356, Apr. 1953, pp. 737–38, https://doi.org/10.1038/171737a0.
October 2, 2025 at 2:33 AM
adam42smith.bsky.social
via @conradhackett.bsky.social this 2022 survey is telling:
www.pewresearch.org/religion/202...

Here's the crosstab that shows pretty strong evangelical dislike for Mormons
Cross-tab showing how people with various faiths feel about other faiths (views favorable - unfavorable percentage). Shows that Mormons have overall positive feelings about everyone, whereas everyone but Catholics has negative feelings about Mormons, with evangelical Christians at -12% (atheists at -55%)
September 29, 2025 at 7:50 PM
adam42smith.bsky.social
The free, offline (i.e. run entirely on your machine), #transcription tools we now have are incredible. Here's my intro to a focus group exactly as auto-transcribed. I have a slight accent that tools in the past didn't love.

This is done w noScribe github.com/kaixxx/noScr... #qualResearch
S08: [00:00:32] My name is Sebastian Karcher. I'm the director of the Qualitative Data Repository, and I'm also faculty in political science at Syracuse. My role today will be in the background. So I'll turn off my camera, turn off my mic, and mainly take notes. And if anything goes awry, I'll help. And if something, if you need something, you can chat me directly, but otherwise you won't hear much from me today.
September 27, 2025 at 1:55 AM
adam42smith.bsky.social
Retroactively would be near impossible, I agree, but the exec order this is about doesn't do that: (via en.wikipedia.org/wiki/Executi... )
The executive order states that these provisions are only effective for people born 30 days or more after the date of the order; it would have only applied to children born beginning February 19, 2025 had it not been blocked.[2]
September 27, 2025 at 12:46 AM
adam42smith.bsky.social
This is what the AAP paper (which isn't a study of Long Covid prevalence in kids but a SotA Review) actually says.
I think making confident statements about prevalence based on this type of evidence, let alone comparing them with actual Asthma prevalence numbers is _hugely_ irresponsible.
The incidence of PASC in children is less well-characterized than in adults and varies widely, ranging from 4% to 62% across existing large studies, with more studies reporting estimates closer to 10% to 20% within the first 6 months after acute infection.9,13,–21 Overall, the wide range of estimates of PASC incidence relates to differences in study design, setting, population, follow-up period, variable ascertainment methods, and variable diagnostic criteria. To date, the majority of studies of PASC in children can be characterized as small, case-based, cross-sectional, retrospective, clinic-based, or convenience samples. PASC can also be difficult to diagnose because associated signs and symptoms are broad, affecting numerous organ systems, and can overlap with underlying comorbidities.
September 25, 2025 at 2:58 PM
adam42smith.bsky.social
I say this without any inclination to minimize long COVID or to defend Oster, but this (unreferenced) paragraph (on which the quoted post is based) is worse statistical malpractice than anything Oster has written.
Long COVID is common, affecting up to 10% to 20% of children with a history of COVID-19. With almost 6 million US children potentially affected, this is higher than the number of children with asthma, the most common chronic health problem in children.
September 25, 2025 at 12:20 AM
adam42smith.bsky.social
I'm so confident they should have because there are explicit guidelines by the Committee on Publication Ethics (COPE) for handling post-pub critiques: publicationethics.org/guidance/flo... and they are very clear about this step: You provide the authors the chance to respond prior to pub 6/n
Share the critique with the original authors and invite them to provide a response within a deadline. Inform the authors that if they do not provide a response by the deadline or decide not to respond, the critique will be published regardless. In such cases, a note can be published with the critique explaining the absence of a response. consider if it is appropriate to also peer review the response or seek advice from the wider editorial team/editorial board. ideally, both critique and response should be published simultaneously
September 24, 2025 at 2:30 AM
adam42smith.bsky.social
I'm not a genocide scholar, but I have heard this repeatedly. Here's a striking passage from Lee Ann Fuji's work on this (doi.org/10.1080/0963...)
Omer Bartov, in a recent Throughline episode (www.npr.org/transcripts/...), talks about this as a common feature of genocides. 1/2
When violence began in his cellule, Christophe went to a neighbor's house to hide.Footnote 3 Fearing too many people were there, he fled to another house. While he hid, a member of one of the local killing groups (which local residents refer to as Interahamwe), stopped to alert Christophe that the rest of the Interahamwe was on its way to kill him.Footnote 4 Christophe fled and survived the genocide. Why would an Interahamwe help him like that?, I asked during one of many interviews with him. Christophe replied matter-of-factly, “Because he was my friend.” Christophe's experience was not unique. Many survivors of the genocide tell similar tales of aid and succor from Hutu, even Hutu who were active participants in the genocide. DONNELLA: Yes, it does. But yeah, Omer said that at this early stage, everyday people are still doing a lot of caveating in their minds about, like, who's a good version and who's a bad version of whatever group is at issue. BARTOV: It's not your friend. Your friend seems to be a perfectly nice person. But generally, that's what they're like.
September 19, 2025 at 1:43 PM
adam42smith.bsky.social
I went looking for why they'd go for MMR+V over MMRV & it seems dumb w very limited consequences (most kids get MMR+V anyway), so I'm not sure this is worth the online rage.
The overall incompetence seems like a bigger issue (as does the looming vote against Hep B)
www.nytimes.com/2025/09/18/h...
The M.M.R.V. vaccine has been available since 2005. It is an alternative to administering two separate shots: the M.M.R. shot, against measles, mumps and rubella; and the shot that immunizes children against varicella, or chickenpox. At the meeting on Thursday, scientists from the Centers for Disease Control and Prevention presented data showing that the combination vaccine slightly increases the risk of seizures caused by a fever. Such seizures also can occur with any childhood illness, including ear infections, but do not cause lasting harm. The C.D.C. had long recommended administering the shots for M.M.R. and chickenpox separately in children under 4, because of the risk of seizures.
September 19, 2025 at 1:54 AM
adam42smith.bsky.social
Competitive High School marching band is even more bonkers. They don't let you record shows, but here's the Syracuse School District setting up for it's show "When a Line Meets a Dot" (a love story; spoiler -- they do get each other).
A group of uniformed high school students lining up on a football field. THere are three giant banners with a line, a squiggle, and a dot.
September 8, 2025 at 1:25 PM
adam42smith.bsky.social
In sports households, the year is split into football & basketball season. In music households, there's marching band season & musical season.
Marching band is so uniquely American, absolutely bonkers, and super fun. The SU #prideOfTheOrange band has 220 kids this year...
September 8, 2025 at 1:25 PM
adam42smith.bsky.social
The priorities are exactly what you'd expect. Here's question 3 (followed by 3 more that make absolutely sure you have the right views on trans kids) and by Q8 we've made it to the constitution...
(The whole thing is of course also terribly buggy)
Question 3 of 34: How is a child's biological sex typically identified. The 'correct' answer is visual anatomical observation and chromosomes. Question 8 of 34: What are the first three words of the Constitution (Correct answer is We the People).
August 30, 2025 at 2:37 AM
adam42smith.bsky.social
We also compared our findings of qual pieces with articles looking at all DAS in PLOS. Biggest differences is the heavier use of other articles of suppl. info & the much higher share of qualitative pieces who say they can't share data (makes sense: more human participants, more sensitive info) 4/n
Table showing the difference between our study, Jiao et al's analysis of ~130k articles and Federer et al's of ~50k. Both Jiao and Federer find around 70% of articles claiming data is in paper of SI. Much higher share of 'on request' and "cannot share' statements for qualitative data
August 8, 2025 at 1:58 PM
adam42smith.bsky.social
We looked at the "on request" category separately. Very oddly (& due to PLOS instructions) a lot of requests are directed to IRBs and Ethics Boards. We're convinced that few of them are equipped to handle external data requests or would even be able to get the data. This is a problem. 3/n
Bargraph of Types of Data Available on Request: 45% from IRB; 33% from author institution; 26% from authors; 6% from original source
August 8, 2025 at 1:58 PM