Afternoon Juners tiny experiment playbook that saved our team a week of back‑and‑forth I fed the creative variants, metric definitions, and a short analysis rubric into a Privacy session, asked June to: 1) preregister hypothesis sample size calc, 2) generate a fail‑fast checklist, 3) synthesize a tiny privacy‑safe QA dataset and schema exportable as JSON for our analytics pipeline, and 4) draft a post‑mortem template with exact fields to collect after the run. Locked the plan with @askjuneai, pushed an automated Notion CI snippet, and launched without leaking user inputs Who else uses June to standardize experiments and ship faster what fields do you always pin in a testing playbook? JM Juners #experimentdesign #privacy #productivity $JUNE

Es ist schon auffällig, wie wenig Menschen einsehen, dass schon das Experimentdesign ungeeignet, gebiased, manipuliert sein kann, und dann prinzipiell überhaupt nichts rauskommen kann, was mit der vermeintlichen Frage zu tun hat. Pandemievibes.

Sometimes, the path to groundbreaking statistical methods starts… with a polite cup of tea. ☕ #DataScience #Statistics #ExperimentDesign #MachineLearning #ResearchMethods #ABTesting

🤔 Curious about real-time ML? 💬 Drop a 🔄 below if you'd like a follow-up tutorial 🔁 Retweet this to help others explore smarter testing 📌 Follow @shreycapital for more intuitive technical data science content #DataScience #ML #BanditAlgorithms #ABTesting #ExperimentDesign #ReinforcementLearning #BayesianStats

📣 What’s your go-to power analysis tool? 🔁 Retweet to help your DS friends avoid underpowered tests 📌 Follow @shreycapital for stats made intuitive (with rigor) #DataScience #Statistics #ABTesting #HypothesisTesting #PowerAnalysis #ExperimentDesign

Und dann erklärst Du uns bitte noch, wieso auch noch ein absolut betrügerisches Grunddesign verwendet wurde, da Calu-3-Lungentumorzelllinie TMPRSS2/ACEII wesentlich gesteigerter expressiert, als bspw. A549, welche sich hier bedeutend eher im Experimentdesign angeboten hätte.(4/x)

This new benchmark reveals the limitations of AI in conducting expert-level research. Original Problem 🤔: Researchers lack comprehensive benchmarks to evaluate LLMs' ability in assisting with expert-level research tasks like equation validation, experiment design, and paper reviewing. ----- Solution in this Paper 🔧: → Introduces AAAR-1.0, a benchmark with 4 distinct research tasks: - EquationInference: Tests equation correctness validation - ExperimentDesign: Evaluates experiment planning capabilities - PaperWeakness: Assesses paper criticism abilities - ReviewCritique: Measures review quality assessment skills → Data quality ensured through: - Expert annotation by senior AI researchers - Multi-round peer review process - Rigorous filtering using GPT-4 - Custom evaluation metrics for each task ----- Key Insights from this Paper 💡: → LLMs struggle with equation validation, performing just above random chance → LLM-designed experiments are creative but often lack feasibility → LLM-generated paper criticisms lack depth and specificity → Models show limited ability in identifying deficient reviews ----- Results 📊: → EquationInference: Top models achieve 60% accuracy (random baseline: 25%) → ExperimentDesign: High creativity but low feasibility in experiment proposals → PaperWeakness: Generated criticisms broadly applicable but lack paper-specific depth → ReviewCritique: Low F1 scores even from top-performing models

A study shows that minimising experimental aberration is critical to improving the accuracy of measuring latent psychological variables 🔬🔍. #experimentdesign #latentvariables #MathPsych #Psychology #Research #Science @sciencedirect sciencedirect.com/science/ar…

🧑‍🔬 #AIinResearch Benefit 💡 Automated #ExperimentDesign--AI designs optimal experiments, reducing trial and error, and accelerating scientific breakthroughs. #ScientificInnovation

How do you isolate impact? 🔍 For @Quantium's chip project, I selected control stores using correlation analysis to match trial stores. This let me accurately measure new layout effects. Deets: joshuanwachi.github.io/ #DataAnalytics #ExperimentDesign #CausalAnalysis"

Bin auf jupyter umgestiegen (in Anaconda), dat reicht für meine Anwendung erstmal. After all möchte ich ja nur ne Automatisierung für Experimentdesign und Datenanalyse/-prognose schreiben, Excel is ja schön und gut, aber ich krieg zu viel bei manchen großen Datensätzen

Hello #BioIT23 Come visit us at booth 922! We are ready to discuss helping you accelerate your science! #multomics #experimentdesign

Good morning #BioitWorld23 - we cannot wait to talk #multiomics and #experimentdesign with you at Booth 922. Come see us!

We'd love to hear about your ideas, roadblocks, and frustrations, and share our platform for making your team more organized, productive, and efficient. Let's talk #multiomics #workflow #experimentdesign

Eben. Man sollte sich erst einmal um sauberes und einheitliches Experimentdesign bemühen, das Störungen bestmöglich eliminiert, und dann kann man weltweit messen. Solange da jeder empirische und experimenteller Laie seine eigene Idee eines Nachweises verwirklicht wird das nix.

Leverage our pre-built ERP experiment sandboxes, free and open-source software and APIs, and full integration with common experiment design tools. #Mentalab #MentalabExplore #OpenSource #CognitiveNeuroscience #Cognition #Psychopy #ExperimentDesign #Biodata #DataPipeline #MoBI

"Tato kniha je pro každého, kdo chce porozumět principům statistiky a interpretovat statistiku, aniž by se pouštěl do podrobných matematických výpočtů." ➡️directory.doabooks.org/handl… #EARA #datainterpretation #experimentdesign #statistics #ANOVA #correlations #Biomedicine #Biology

2/ Talk to your #Genomics and #Bioinformatics groups BEFORE doing your experiment. They have likely done these experiments before and can help catch a fatal #experimentdesign flaw before spending the dollars.

Sign up for our webinar on Experiment Design, next Wednesday, june 15th at 12:30 PM CET. #Growthhacking #Growth #Experimentdesign vwo.com/webinars/how-to-desi… See you there? #Saas #Growthmarketing

Sign up for our webinar on Experiment Design, next Wednesday, june 15th at 12:30 PM CET. #Growthhacking #Growth #Experimentdesign vwo.com/webinars/how-to-desi… See you there? #Saas #Growthmarketing