"Database replication strategies shape latency, throughput, and consistency—benchmarks show which approach shines for your workload. #DatabasePerformance #Replication" https://milvus.io/ai-quick-reference/how-do-replication-strategies-affect-database-benchmarks

This last talk #WoReLa1 before the final keynote is a very original one: Mike Dorin, who is a software engineer, is presenting an attempt to reproduce the findings of a study on an indigenous language spoken in Peru that was originally conducted by his wife when she was a student several decades ago! Mike used an impressive array of ML tools to transcribe and analyse the written and oral recordings that his wife and co-students made as part of their fieldwork. #replication

#WoReLa1 continues with "How replicable are bilingual interactive processing effects? A pre-registered close replication and extension of Dijkstra et al. (1999)" presented by Jurriaan Witteman. #linguistics #replication

Opening the second day of the First Workshop on #Replication (#WoReLa1), we are treated to a keynote about "Research on Bilingualism as Discovery Science" by Anne Beatty-Martínez. Anne has lots of food for thought for us all. #linguistics #metascience

The first day of the Workshop on #Replication in the Language Sciences (#WoReLa1) is closing with a keynote by Timo Roettger provocatively titled... "Against replication"! Timo intends to argue that #linguistics is simply not ready for replications.

Lucía Vieitez Portas is continuing with the topic of grammatical gender with a talk entitled "Grammatical gender and beyond: Uncovering differences across women and men on the interplay between grammatical and emotional processing". So far, Lucía has made clear that the literature has produced very mixed results, using both ERP and behavioural methods motivating future work on this reaction question. #WoReLa1 #linguistics #replication

Tamara Bouso-Rivas's talk on "Mapping links within the network of English Objoid Constructions: A #replication study using T-Maze" was also very interesting and triggered discussions on many different aspects including what "unrelated constructions" in Construction Grammar are (and how degrees of relatedness might be measured) and how long differences in reaction times need to be for us to consider that a priming effect has been detected. @tamarabouso #linguistics #WoReLa1

In conclusion, we need to do more replications (but maybe call them something else) and move away from binary expectations of success vs. failure for findings to replicate. #replication #linguistics #WoReLa1

Angela de Bruin agrees with Nosek and Lakens that "[t]here is no such thing as an exact replication."

She will discuss two major challenges to #replication in the language sciences, with a particular focus on #bilingualism and executive control:

1. Literature does not provide us enough information.
2. Not enough consideration or theoretical understanding of how "changes" matter.

We begin with a keynote by Angela de Bruin on „The importance of both direct and conceptual #replication studies when studying language users from different language backgrounds“. #linguistics #WoReLa1

Brompton and I have made it to the conference venue: very much looking forward to the 1st Workshop on Replication in the Language Sciences! https://sites.google.com/view/worela1/programme #linguistics #replication #WoReLa1

"Database replication strategies shape latency, throughput, and consistency—benchmarks show which approach shines for your workload. #DatabasePerformance #Replication" https://milvus.io/ai-quick-reference/how-do-replication-strategies-affect-database-benchmarks

#statstab #378 Selective Inference: The Silent Killer of Replicability

Thoughts: Benjamin overviews the replicability crisis, alternatives to p-values (and their issues), and suggests selective reporting is a large issue itself.

#replication #crisis #pvalue #nhst #selectivereporting #replicability #reproducibility

https://hdsr.mitpress.mit.edu/pub/l39rpgyc/release/3

🔁 Be ready for the unexpected and set up disaster recovery for Azure VMs!
See how to enable replication and disaster recovery to another region:
🔗 https://msft.it/6016SmiEn
#Azure #DisasterRecovery #Replication

#statstab #374 Replication Mirror

Thoughts: Not sure what to make of this, but it has a useful reference list on the discussion of the OSC replication results.

#replication #criticism #openscience #critique

http://www.psi-chology.com/replication-mirror/

Phil McAleer & Helena M. Paterson Improving pedagogy through Registered Reports #oSoTLSummer #SoTL #AcWri #highered #Vol1No1_21 #pedagogy #scholarship #publication #registeredreports #replication https://osotl.org/osotl/article/view/13/2

#PostgreSQL folks: I'm seeing a thing where stalled cursor transversal doesn't seem to be subject to #replication timeout cancellation. Which means that a backend with a stalled cursor read can block replication indefinitely. Is that a known problem, or should I start working on a test case?

#statstab #371 Safeguard Power as a Protection Against Imprecise Power Estimates

Thoughts: tl;dr - when replicating a study, use the lower end of the CI of the original study as your effect in a power analysis.

#poweranalysis #samplesize #replication

https://journals.sagepub.com/doi/10.1177/1745691614528519

#statstab #369 Don't Trust Internal Meta-Analysis

Thoughts: I don't agree with all the assumptions, but interesting to know how bad these can be.

#metaanalysis #QRPs #minimeta #datacolada #falsepositive #replication

https://datacolada.org/73

#statstab #366 Type M error might explain Weisburd’s Paradox

Thoughts: Learn about type M error while you learn about the issues in criminology!

#replication #typeM #typeS #QRPs #paradox #power #effectsize #errorrate

https://sites.stat.columbia.edu/gelman/research/published/weisburd_28.05.2017.pdf