Be good-argument-driven, not data-driven (twitchard.github.io)

Article summary

The article argues that being 'data-driven' can be misleading and that good arguments should take precedence over data. It claims that data can be misleading or incomplete, and that relying solely on data can lead to bad decisions. The article suggests that data should be used to support good arguments, but not to replace them. It also warns against the dangers of overemphasizing data, such as harming morale and ignoring important factors that cannot be measured.

Main themes

• data-driven decision-making
• role of expertise
• interpretation of data
• correlation vs causation
• limits of data
• data misuse
• theory and explanation
• scientific method

What commenters say

• Some argue that data-driven approaches can lead to better decision-making, but others claim that this approach can be flawed and overlook important factors.
• There is a debate about the role of expertise in decision-making, with some arguing that expertise is still valuable even in the presence of data.
• The importance of distinguishing between correlation and causation is highlighted, as well as the need for careful interpretation of data.
• Some commenters suggest that a combination of data and good arguments is the best approach, while others argue that data can be used to support bad arguments.
• The limitations of data in capturing complex phenomena are discussed, with some arguing that data can never fully capture the nuances of human experience.
• The potential for data to be misused or misinterpreted is also raised, with some commenters arguing that this can lead to poor decision-making.
• There is a discussion about the relationship between data, theory, and explanation, with some arguing that data alone is not enough to support a theory.
• Some commenters argue that the scientific method provides a framework for balancing data and theory, but others suggest that this method is not always followed in practice.