Never Met a Science

Never Met a Science explores the intersections of political communication, social science methodology, and the impact of the internet on society. It covers themes like the attention economy, technological development, social media's influence, algorithm critiques, and the evolution of digital culture, while also examining broader societal shifts and the role of information in social change.

Political Communication Social Science Methodology Internet and Society Attention Economy Technological Development Social Media's Influence Algorithm Critiques Digital Culture Information Society Quantitative Methodology

The hottest Substack posts of Never Met a Science

And their main takeaways

Describing the Journal of Quantitative Description: Digital Media

16 implied HN points • 07 Oct 20
  1. Things are moving quickly in the world of digital media
  2. The Journal of Quantitative Description: Digital Media aims to address the undersupply of quantitative descriptive knowledge
  3. Political methodology is evolving to emphasize the synthesis and aggregation of knowledge in social sciences

The Theory of the Academic Firm

16 implied HN points • 07 Aug 20
  1. The costs of doing quality causal research in the social sciences are high.
  2. Formal academic institutions are necessary to conduct research at the scale required today.
  3. There are serious transactions costs in academia, but benefits from specialization and division of labor.

Social Science Agenda Setting

16 implied HN points • 03 Aug 20
  1. Scientists should use science to decide what to study
  2. Social science agenda is currently set by personal intuitions and media reports
  3. More quantitative description and survey of researchers needed to set a rigorous social science agenda

Social Science Without Social Scientists

11 implied HN points • 21 Jan 21
  1. Social science involves processing massive data for human-scale comprehension in research.
  2. Quantitative text analysis leverages machine learning like BERT to handle large, unsupervised data processing.
  3. The philosophy of Cybernetics suggests a shift to predictive science without the need for traditional knowledge acquisition.

Where is social scientific knowledge: The .pdf or the .csv?

11 implied HN points • 06 Jan 21
  1. Social scientific knowledge is affected by changes in technology, making synthesis difficult under different technological regimes.
  2. The ultimate goal of social science is to predict human behavior by reducing knowledge dimensions through language-based theories and numerical data.
  3. There is a paradigmatic shift towards a non-human intelligence as the ultimate point of knowledge synthesis, leading to a need for radical reform in social scientific practices.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

28 Years of Boomer Presidents and the Biden Interregnum

11 implied HN points • 09 Nov 20
  1. Generations have a significant impact on government, society, and culture.
  2. The Baby Boomer generation has had a dominant influence on postwar United States.
  3. Younger generations are challenging Boomer norms with their diversity and use of technology.

Susan Wojcicki Wants You To Think That YouTube's Algorithm is All-Powerful

11 implied HN points • 26 Oct 20
  1. Public discourse on YouTube politics often overly focuses on the recommendation algorithm.
  2. The demand for YouTube content drives the importance of the platform more than the algorithm does.
  3. Emphasizing the omnipotence of tech companies' algorithms may not accurately reflect reality and can hinder effective solutions.

Temporal Validity is Distinct From External Validity

11 implied HN points • 13 Aug 20
  1. Temporal validity focuses on applying knowledge to future contexts.
  2. External validity involves factors like pre-treatment characteristics, outcomes, treatments, and contexts.
  3. Social science needs more knowledge production, descriptive knowledge, and faster research methods.

Cacophony, not conformity

5 implied HN points • 14 Sep 20
  1. Online echo chambers are not as prevalent as believed, only affecting a minority in specialized networks.
  2. Social media can create echo chambers through dense and correlated signals, leading to overconfidence and ideological polarization.
  3. Echo chambers can involve literal message repetition, such as 'copypasta,' hindering diverse voices from being heard.

Knowledge Windfalls

0 implied HN points • 03 Aug 21
  1. Running 32,000 experiments is more efficient than running just one.
  2. Collaborating with corporations can lead to valuable knowledge windfalls.
  3. Site-selection bias is a challenge in social science experiments.