iiinova — where your digital self keeps evolving

Ricardo Ciacci Strina @riccardociacci.bsky.social · 10/10/2026

@grufian.bsky.social Ojalá ERC forme parte del Frente Amplio, tal y como ERC formó parte del Frente Popular hace 90 años

000

Reposted by Ricardo Ciacci Strina

Sumar @movimientosumar.es · 05/10/2026

Estamos listas y somos más. El próximo 29 de Noviembre el Frente Amplio es la llave para garantizar el derecho a la vivienda, a la salud, al tiempo, a la vida digna.

3517995

Reposted by Ricardo Ciacci Strina

Jóvenes Más Madrid @jovenesmasmadrid.bsky.social · 19/07/2026

Hoy tenemos una cita histórica 🔥🔥🔥 Ven a vivirla con nosotros desde las 20:00 ¡Vamos a por la segunda! 🇪🇸❤️‍🔥 📍Paseo de las Acacias, 52

¿Y POR QUE NO?
Desde las 20.00h | Sede Más Madrid
Paseo de las Acacias, 52
ENTRADA LIBRE
ESPAÑA vs ARGENTINA

042

Ricardo Ciacci Strina @riccardociacci.bsky.social · 08/07/2026

results – is a disservice to the profession, not an act of scientific correction. 8/8

000

Ricardo Ciacci Strina @riccardociacci.bsky.social · 08/07/2026

publish the critique and the reply (after peer review) side‑by‑side, and let readers decide. Retracting an article because some specifications are not robust to alternative choices – and because the critics choose not to believe the 7/8

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 08/07/2026

If retractions were based on whether a result survives every alternative coding choice, much of economics would be retracted. The proper way to handle this debate was always the path taken by the American Economic Review in similar cases (e.g. Card & Krueger / Neumark & Wascher): 6/8

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 08/07/2026

The journal's own independent committee reviewed all evidence, considered retraction (even a corrigendum) and discarded them. A disagreement about how to code seasonality is a matter of interpretation – what we call robustness. 5/8

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 08/07/2026

If they truly believed in our profession and were impartial, they would be the first to state that they do not believe the results, but that a retraction is not justified. 4/8

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 08/07/2026

but simply that they do not believe it. It is sad to see that scholars who recognise the large statistical effect is not due to an error, and that it is possible to find it in the data, nevertheless accept a retraction – only because they do not believe the results. 3/8

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 08/07/2026

not a statistical mistake in the analysis itself. Even after they "correct" it, they acknowledge in their own appendix that "it is possible to get large and significant estimates from a correctly implemented RDiT" (Appendix, p. 8). Their argument is not that the result is statistically invalid, 2/8

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 08/07/2026

@ollefolke.bsky.social claims the Retraction Watch article is incomplete because their peer‑reviewed critique "corrects" a coding error in my follow‑up paper. But what they call a "coding error" is not an error in the code – it is a description typo in the PDF about how seasons were coded, 1/8

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 20/06/2026

Quando bullizzava gli altri le andava bene, quando bullizza lei no. Coerenza meloniana

000

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

In 28 hours I sent the replication file you asked via email and later posted that email online. Then, with those files you opened the investigation at the journal. The journal asked me to provide the files via the investigation and so I did. I did not say I'd provide the files, I provided them

000

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

“Anyone who carries out research knows writing replication do files takes time.”. That is asking for more time. I mentioned the time constraint issue also in our correspondence that may be accessed here: www.dropbox.com/home/Online%... file "Correspondence thread 2.pdf" 2/2

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

As it can be seen in the picture you shared before here bsky.app/profile/olle... On 20 March 2024 I wrote: “it takes ages to prepare replication do files for a paper of 120 pages”, “You gave me six hours, this hectic behaviour speaks on its own” and 1/2

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

As it might be clear to anyone, all these sentences convey a different message from the one that actually happened and you finally admitted N/N

000

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

and even in Twitter/X 3) "No, it was based on the following: he first said no but eventually agreed to share only a small portion of the do-file." source: x.com/OlleFolke/st... 5/N

x.com

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

1) "The author refused to provide replication material" source abstract of this file: www.econstor.eu/bitstream/10... 2) "The author, Riccardo Ciacci, also refused this request." source: Appendix of this article journals.sagepub.com/doi/10.1177/... 4/N

econstor.eu

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

Please acknowledge in your articles that I only asked for more time, I never said “no”. Please clarify that you interpreted it as a refusal. Currently in your articles you write statements of this sort: 3/N

econstor.eu

200

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

Moreover, I did not say you would find errors. I said "most likely you will find something you do not agree with merely to discredit the results of the paper without furthering research on this issue" implying that sharing files would not end the debate – and it did not. 2/N

econstor.eu

Making sure you're not a bot!

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

Here you write I "clearly say no to sharing the replication files." bsky.app/profile/olle... Now you state you "interpreted" my responses as saying no. I think it is clear to everyone who may read these messages that such two claims are substantially different. 1/N

200

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

Sorry, can you tell me where I say "no" to sharing them in that thread? There I only state I need more than 6 hours (the time you gave me) to prepare the replication files of a paper that has 120 pages with the appendix and that sharing them would not end our debate (as it did not)

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

the original paper was later retracted – based on one of your reports (again desk rejected by the journal) that you submitted to COPE. You also describe this event in the paper your published this month. N/N

000

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

What I find nonsensical is that after the journal considered your reports not publishable (i.e., desk rejected them) and decided not to correct my original paper but to have a second (complementary/additional) paper, 16/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

If the committee had concluded that retraction was warranted, as I show in the correspondence I uploaded they considered that option but did not eventually choose it, I would have agreed with it. 15/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

After such an amount of robustness checks was evaluated, and retraction was not the outcome chosen by the committee, not even correcting the original paper, I thought such a decision was final. 14/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

since, in that complementary paper, I was asked by the journal to run about 76 robustness checks per specification (for a total of around 310 robustness checks). Some of them, in a regression discontinuity in time framework, merely boil down to a naive pre and post comparison. 13/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

as I also showed in my Institute for Replication paper. Such results are robust to some checks and not robust to other, as any other finding in our field. It was clear that result was not going to be robust to all the robustness checks I was asked to perform 12/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

In the only specification that shared the identification strategy with the original paper (no population weights, using control variables), the results are statistically positive and statistically equal to my original estimates. This implies that the Stata issue did not cause those results, 11/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

implying that the Stata issue was not considered an error serious enough to warrant a correction. Instead, they opted for a complementary paper estimating a different target parameter (weighted by population). 10/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

Back to your question: In my post I was describing what happened. The committee considered the option to retract my paper but eventually did not opt for it after seeing the results. They also explicitly discarded the option to correct the original paper, 9/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

might be a way to estimate a precise zero. If you do not want to work with me is fine, but please do not forget about these topics. They are important and deserve to be studied regardless of whether my paper is retracted or not. 8/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

My aim is to push knowledge further on these topics. As I have said before, I am willing to collaborate with you using better data, to which I lack access because I am not affiliated with a Swedish institution. If you believe there is no effect, analysing better data (together or alone) 7/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

However, I also believe that your way of proceeding – writing (what I consider) "aggressive" posts on social media, asking for a retraction in your reports, contacting my university, and eventually filing a complaint through COPE – did not help to achieve that constructive outcome. 6/N

600

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

I think the journal should have addressed this issue (similarly) as the AER handled the Card & Krueger / Neumark & Wascher debate: through a publishing your report via peer review and my reply via peer review as well. 5/N

270

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

I am glad you found the issue, it could have been key to the estimation and this is why I wrote a reply for the Institute for Replication to your second report. 4/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

within 28 hours of receiving your email. You also claim that my papers do not describe well enough what I do in Stata. That is at odds with reality: the issue you found was described so clearly in my paper that it could have been grasped without seeing the Stata code at all. 3/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

Yet, your report contains statements about my behaviour that I can easily prove to be untrue by sharing our messages via this link www.dropbox.com/home/Online%.... For example, you wrote that I refused to share my code. However, I sent you the part of the code you considered most relevant 2/N

dropbox.com

200

Ricardo Ciacci Strina @riccardociacci.bsky.social · 28/05/2026

I am sorry I am gonna be long but I will address the whole situation. In my post I used exactly those words: bsky.app/profile/ricc... I am glad you were able to publish your report. I believe that the only way to better understand any topic – including this one – is to study it. 1/N

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 27/05/2026

To my understanding, a desk rejection means the journal decided not to publish your reports. That is what I meant by “not considered publishable”. I was only stating what happened

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 27/05/2026

I am sorry Olle, I was merely describing the situation. The journal desk rejected your reports; as far as I understand, that means they did not consider them publishable. Please correct me if I am wrong.

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 26/05/2026

I mean: the journal rejected to publish your two reports. Its committee found no need to correct my article. The retraction came later – with no published evidence except my 2nd article, which the committee called complementary, not corrective. That is the contradiction.

100

Ricardo Ciacci Strina @riccardociacci.bsky.social · 27/04/2026

For transparency, I have uploaded the correspondence throughout this process. Reading it, it becomes clear that an article which the responsible editor, after an internal investigation, found no need even to correct 1/2

100

Reposted by Ricardo Ciacci Strina

Javier Gil @javigil.bsky.social · 27/04/2026

1.Hoy publicamos nuevo informe. El problema de la vivienda es de concentración de la propiedad. 🔴Cada vez más personas sin propiedades 🔴Cada vez menos personas con una propiedad. 🟢 Cada vez más personas son multipropietarias. La vivienda parte el país en dos. 🧵

11383319

Ricardo Ciacci Strina @riccardociacci.bsky.social · 20/04/2026

I agree replicators deserve fair treatment, but that is a separate editorial decision. As for me, I now find myself in a Kafkaesque situation: an article that the responsible editor deemed not in need of correction was later retracted – under external pressure, not science. 2/2

120

Ricardo Ciacci Strina @riccardociacci.bsky.social · 20/04/2026

I understand the frustration, but the second article was not a reward – it was a direct requirement from the journal’s independent committee after a ten‑month investigation. They explicitly asked for a follow‑up article addressing a different research question, not a correction. 1/2

100