About:

Jason Collins is an economist and researcher interested in behavioural science, data science, and evolutionary biology.

Website:

Specializations:

Interests:

Behavioural and data science Economics Evolutionary biology

Outgoing Links:

Subscribe to RSS:
The blog post discusses the persistence of a study on disfluency in text presentation, which claims that harder-to-read fonts can enhance analytical thinking. It reviews the original findings by Adam Alter and subsequent replicati...
The blog post discusses the concept of 'anchoring and adjustment' heuristic, illustrating how initial values (anchors) influence people's estimations, using examples involving Elvis's age and AI decision-making. It critiques the n...
The blog post discusses findings from a new study by Gemma Altinger et al. that contradicts a classic paper by Donald Redelmeier and Eldar Shafir regarding decision-making in medical contexts. The new study shows that offering mul...
The blog post discusses various resources and insights related to language models and mechanistic interpretability research. It highlights the value of ARENA’s AI Safety course for understanding LLM foundations, explores the conce...
The blog post analyzes a trial conducted by the Behavioural Insights Team (BIT) regarding loft insulation uptake. It highlights how the primary barrier to adoption was not cost, but the hassle of clearing out lofts. The trial test...
The blog post critiques a Harvard Business Review article that claims AI-generated content, termed 'workslop', is detrimental to productivity. It highlights the flawed methodology of the study, which suggests that 40% of employees...
A personal reflection on the best books read in 2025, featuring a mix of fiction and non-fiction with insights on themes like language and AI.
The author discusses the lack of methodological rigor in human-AI interaction research, comparing it to the replication crisis in psychology. He highlights the absence of pre-registration, open data, and experimental materials, as...
The article discusses the importance of collecting genetic data in longitudinal studies to understand social mobility and inequality. It highlights the Wisconsin Longitudinal Study as an example of how genetic data can enrich deca...
This reading list provides a balanced view of behavioural economics and science, offering a critique of popular books and conferences. It includes books and articles that challenge traditional behavioural economics arguments and p...
Jason Collins shares the books he read in 2024, including both fiction and non-fiction. He provides a brief review of each book and mentions that he read a total of 49 books, with an increased focus on non-fiction.
The text discusses the impact of AI on human decision-making, particularly in the context of recruitment. It highlights an experiment that found that recruiters who used high-quality AI became lazy, careless, and less skilled in t...
The text discusses the concept of megastudies in behavioral science, which involves testing many interventions in a single massive experiment. It provides examples of megastudies conducted to increase gym attendance and vaccinatio...
The text discusses the performance of AI in comparison to human performance in responding to patient questions. It highlights the brevity and curtness of human responses compared to the long, involved answers given by ChatGPT. The...
The text is a comment on the manifesto for behavioural science published in Nature Human Behaviour. The author discusses three of Hallsworth's proposals: See the system, No view from nowhere, and Data Science for Equity. The autho...
The author shares notes for an undergraduate subject in behavioural economics, including exercises and videos. They take a traditional approach but are also developing a new set of notes with a more critical lens. The notes are av...
The study examines 30 US cities that ran 73 RCTs with a national nudge unit. It found that the strength of the evidence and key city features do not strongly predict adoption of nudges. The largest predictor is whether the RCT was...
The author corrects a previous post about a chimp named Ayumu, who was thought to have superior working memory to humans. The author found that the claim was based on a flawed study and that humans can match Ayumu's performance wi...
The author discusses the use of generative AI tools in academic work, including GitHub Copilot, ChatGPT, Claude Pro, and ElevenLabs. They describe how they use these tools for coding, data analysis, data exploration, writing, lect...
The post discusses the concept of ergodicity economics and its implications for risk preferences. It includes a presentation plan, a simulation of bets, and an experimental result on shifting between additive and multiplicative en...
The text is a review of Bryan Caplan's book 'The Case Against Education'. It discusses the author's personal experience with education and the income premium for education. It also delves into Caplan's analysis of the human capita...
The text discusses the problem of p-hacking and hypothesizing after results are known in scientific research. It suggests that preregistration of analysis can help to avoid these issues. The author presents a plot of the status of...
The text discusses the state of social science, the issues with journals, the meaningless of the label 'misinformation', the Flipper Zero, and the need to clean up the list of named 'biases'. It also touches on the replication rat...
Jason Collins shares the books he enjoyed the most in 2023, including both fiction and non-fiction. He provides a brief review of each book and mentions that he read a total of 36 books, with a higher volume of cover-to-cover read...