Discussions

5 9about 1 month ago

by amir_r

Questionon A Programmable Dual-RNA–Guided DNA Endonuclease in Adaptive Bacterial Immunity

From in-vitro programmability to in-cell specificity — what's the real gap?

Bioinformatics

2 3about 1 month ago

by dortiz

Questionon The Strength of Weak Ties

How does the paper connect a network-theory idea to empirical evidence?

Sociology

1 1about 1 month ago

by dortiz

Explanationon Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

Why the endpoint definition is the whole ballgame here

Public Health

0 0about 2 months ago

by hcole

Explanationon The Oregon Health Insurance Experiment: Evidence from the First Year*

Why this is a model for evaluating policy

Economics

0 0about 2 months ago

by weizh

Critiqueon Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

What confounders / limitations matter when generalizing this RCT?

Public Health

0 0about 2 months ago

by weizh

Critiqueon The Oregon Health Insurance Experiment: Evidence from the First Year*

What external-validity concerns survive the clean design?

Economics

1 1about 2 months ago

by tbecker

Research Ideaon Highly accurate protein structure prediction with AlphaFold

Hot one: will AI structure prediction make experimental structural biology obsolete?

Bioinformatics

5 10about 2 months ago

by dortiz

Research Ideaon Lost in the Middle: How Language Models Use Long Contexts

If models are 'lost in the middle', is RAG re-ranking solving the wrong problem?

NLP

3 4about 2 months ago

by jihoon

Questionon The Oregon Health Insurance Experiment: Evidence from the First Year*

Does the lottery design fully support the causal claim?

Economics

0 0about 2 months ago

by dortiz

Questionon Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

How should we read '95% efficacy' — what is it a percentage of?

Public Health

2 1about 2 months ago

by dortiz

Explanationon Minimum Wages and Employment: A Case Study of the Fast-Food Industry in New Jersey and Pennsylvania

Why this paper is a teaching staple beyond its topic

Economics

0 0about 2 months ago

by weizh

Explanationon Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing

Explaining FDR to a non-statistician in one sentence

Statistics

0 0about 2 months ago

by weizh

Critiqueon Minimum Wages and Employment: A Case Study of the Fast-Food Industry in New Jersey and Pennsylvania

Which robustness checks matter most for a DiD like this?

Economics

0 0about 2 months ago

by tbecker

Critiqueon Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing

What assumptions does the BH procedure actually need?

Statistics

0 0about 2 months ago

by tbecker

Research Ideaon Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing

Is FDR control a band-aid for a preregistration problem?

Statistics

3 4about 2 months ago

by weizh

Questionon Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing

Why control FDR instead of family-wise error rate?

Statistics

1 1about 2 months ago

by hcole

Questionon Minimum Wages and Employment: A Case Study of the Fast-Food Industry in New Jersey and Pennsylvania

What is the parallel-trends assumption doing here, exactly?

Economics

1 1about 2 months ago

by weizh

Explanationon Estimating causal effects of treatments in randomized and nonrandomized studies.

The core idea in one line, and where it fails

Statistics

0 0about 2 months ago

by weizh

Critiqueon Estimating causal effects of treatments in randomized and nonrandomized studies.

How is this different from 'just controlling for covariates' in a regression?

Statistics

0 0about 2 months ago

by tbecker

Is measured LLM/NLP progress real, or are leaderboards just measuring contamination?

From in-vitro programmability to in-cell specificity — what's the real gap?

How does the paper connect a network-theory idea to empirical evidence?

Why the endpoint definition is the whole ballgame here

Why this is a model for evaluating policy

What confounders / limitations matter when generalizing this RCT?

What external-validity concerns survive the clean design?

Hot one: will AI structure prediction make experimental structural biology obsolete?

If models are 'lost in the middle', is RAG re-ranking solving the wrong problem?

Does the lottery design fully support the causal claim?

How should we read '95% efficacy' — what is it a percentage of?

Why this paper is a teaching staple beyond its topic

Explaining FDR to a non-statistician in one sentence

Which robustness checks matter most for a DiD like this?

What assumptions does the BH procedure actually need?

Is FDR control a band-aid for a preregistration problem?

Why control FDR instead of family-wise error rate?

What is the parallel-trends assumption doing here, exactly?

The core idea in one line, and where it fails

How is this different from 'just controlling for covariates' in a regression?