Interview with Daniël Lakens and Smriti Mehta on the state of Psychology
Nov 18, 2023
auto_awesome
Psychologists and podcast hosts Daniël Lakens and Smriti Mehta discuss the replication crisis, responsible heterodoxy, and the role of different communication modes in shaping discourses. They also touch on AI, Lex and Elon being teenage edge lords, and controversial resignations in the field of psychology.
The replication crisis in psychology highlighted the issue of unreliable and non-reproducible findings in scientific research.
The open science movement and methodological reform aim to increase the reliability and transparency of scientific research.
The balance between attachment to theories and the importance of rigorous and unbiased research remains a challenge.
Subjective Bayesianism violates the principle of separating personal beliefs from evaluating scientific claims.
Podcasting provides a valuable platform for engaging conversations, critiquing manipulation traits, and exploring different viewpoints in academia.
Deep dives
Background of the Replication Crisis
The replication crisis in psychology emerged around 2011-2012 with the publication of studies that showed non-replicable results. It highlighted the issue of unreliable and non-reproducible findings in scientific research.
Responses to the Crisis
In response to the replication crisis, the open science movement and methodological reform gained momentum. The field started promoting practices such as registered reports, open data sharing, and improved statistical techniques. These efforts aimed to increase the reliability and transparency of scientific research.
Challenges and Criticisms
While the open science movement and methodological reform brought positive changes, there are challenges and criticisms. Some argue that it may lead to a sterilization of research, over-reliance on certain methods, and a lack of exploration of real-world phenomena. The balance between attachment to theories and the importance of rigorous and unbiased research remains a challenge.
The Future of Science
The replication crisis and the push for open science have brought changes and progress to the scientific community. It is important to recognize the historical perspective and understand that science is continuously evolving. Collaborations, distinction between theoretical and empirical scientists, and ongoing efforts to improve research practices all contribute to the advancement of science.
Critique of subjective Bayesianism and preference for objective approach
Subjective Bayesianism violates the principle of separating personal beliefs from evaluating scientific claims. It goes against the philosophy of science and introduces personal biases into the analysis. It also involves multiplying personal beliefs with data, which can be seen as making up data. The use of informative priors in subjective Bayesianism can lead to biased results. It is important to have an objective approach to evaluating scientific claims and not rely on subjective beliefs.
Misuse of Bayesian thinking and focus on personal beliefs
Bayesian thinking has been misused in some cases, with personal beliefs being inserted as priors without proper justification. This can lead to biased analyses and invalid conclusions. Claiming that Bayesian thinking is how babies think is not a strong argument for its reliability in scientific analysis. The focus should be on empirical evidence and rigorous methodology rather than subjective beliefs.
Criticism of Bayesian thinking in scientific research
Bayesian thinking should not be regarded as the gold standard in scientific research. It is often used as a buzzword without proper understanding and can lead to the misuse of statistical analysis. Bayesianism does not always provide more rigorous or accurate results than other statistical approaches. The focus should be on the quality of the research design, measures, and objective analysis rather than relying solely on Bayesian thinking.
Bayesianism and the perception of statistical rigour
Bayesianism is often perceived as a more rigorous approach to statistical analysis, but this perception is not always based on evidence. People often throw around the term Bayesian without a deep understanding of its principles and use it to sound more statistically sophisticated. However, proper statistical rigour is achieved by applying sound research design, collecting reliable data, and employing objective analysis methods rather than relying solely on Bayesian thinking.
The Importance of Podcasting
Podcasting is seen as a valuable medium of communication that provides a space for engaging conversations and different perspectives. It offers a platform for individuals to share ideas, critique common manipulation traits, and deconstruct various topics. Listeners appreciate the informal nature of podcasts, which allows for more connection and understanding. It is considered a valuable resource for those looking to explore different viewpoints, particularly in fields like academia, where podcasting can provide an alternative to traditional scholarly publications.
The Ambivalence of Academia and the Future
When it comes to academia, there is a mix of opinions regarding its future and its impact. While podcasting is seen as a valuable contribution to knowledge dissemination, some individuals express concerns about the current state of academia and its system of publishing. There is a growing recognition that intellectual contributions can be made through various mediums, such as podcasts and open textbooks, which may have a broader reach and impact than traditional academic outputs. The question of whether academia can make a positive impact and create meaningful change is discussed, with some considering alternative paths outside academia if meaningful impact cannot be achieved within the current system.
We are back with more geeky academic discussion than you can shake a stick at. This week we are doing our bit to save civilization by discussing issues in contemporary science, the replication crisis, and open science reforms with fellow psychologists/meta-scientists/podcasters, Daniël Lakens and Smriti Mehta. Both Daniël and Smriti are well known for their advocacy for methodological reform and have been hosting a (relatively) new podcast, Nullius in Verba, all about 'science—what it is and what it could be'.
We discuss a range of topics including questionable research practices, the implications of the replication crisis, responsible heterodoxy, and the role of different communication modes in shaping discourses.
Also featuring: exciting AI chat, Lex and Elon being teenage edge lords, feedback on the Huberman episode, and as always updates on Matt's succulents.
