Weakness - Difficulty testing
Despite Sutherlands promise to provide a scientific, mathematical framework where future offending behaviour could be predicted, differential association proves difficult to test
Its hard to measure what pro-criminal attitudes a person has or has been exposed to
Similarly, its based on the assumption that offending behaviour will occur when pro criminal ones outweigh anti criminal ones
However, without being able to measure these it is difficult to know at what point the urge to offend is realised and the criminal career is triggered