Unit 5.2: Mills Methods
John Stuart Mill's Methods of Causal Discovery
Overview of Mill's Five Methods
- John Stuart Mill identified five strategies for discovering causes in his 1843 work, A System of Logic:
- Method of Agreement
- Method of Difference
- Joint Method of Agreement and Difference
- Method of Residues
- Method of Concomitant Variation .
Method of Agreement
- The method states that if a single factor (X) is present in all cases where an effect occurs, then X is likely the cause.
- Example: If a group dined at a restaurant and all but one person got sick after eating seafood, it suggests seafood is the probable cause. .
Method of Difference
- This method posits that if one situation leads to an effect while another does not, and the only difference is a single factor, that factor can be inferred as the cause.
- In the dining example, since only one person did not eat chicken and everyone else who did got sick, chicken is likely the cause. .
Joint Method of Agreement and Difference
- This approach combines both previous methods to strengthen causal inference.
- For instance, if all who consumed beef fell ill while one individual who did not consume beef remained healthy, this reinforces that beef may be responsible for illness. .
Method of Residues
- This method involves identifying known causes for various effects; any remaining effect can be attributed to an unaccounted cause.
- Example: If four people donated $100 total with three contributing known amounts ($20 each), then the remaining $40 must belong to Tom as he’s the only donor left unaccounted for. .
Method of Concomitant Variation
- Here, quantitative changes in effects are linked with quantitative changes in presumed causes.
- For example: If soil samples show varying levels of radioactivity corresponding with uranium content, it suggests uranium could be causing radioactivity due to their direct relationship. .
Caveats on Mill's Methods
- Mill’s methods assume there exists a set group of possible causes already considered.