Cause and correlation
It is dangerous to treat a correlation as conclusive evidence of a direct causal link. We need to hear good reasons for this connection. Two sorts of event may be correlated; an increase in TV viewing might take place at the same time as a rise in crime levels, for example, without there being a direct causal connection between them even if they're found together. There might be another factor which affects the rise in crime levels. To demonstrate, take the example of shoe size. You could say that people with small feet are more likely to be illiterate. Now this is because people with small feet are more likely to be very young and thus more likely to be illiterate. Having small feet isn't the reason for the illiteracy. So some correlations cause confusion and just because two things happen together doesn’t mean they cause each other.
If you say violence on the telly leads to violent crimes later on, that's known as a hypothesis - something that may be but isn't necessarily true. You really need further evidence to prove there's a direct link. A mere correlation when there are so many factors involved isn't going to prove it. You might as well say an increase in exposure to crime on the TV leads to an increase in smoking, suicide, car accidents and foreign debt just because they all occur together. So when you’re presented with apparently incontrovertible evidence suggesting X causes Y, look for alternate explanations of that correlation and check whether you can eliminate the connection.
This is a general statement based on insufficient evidence. For instance, if, on the basis of a conversation with one taxi driver we were to conclude that all taxi drivers are anti-racist we would be guilty of a rash generalisation. Even if our sample of taxi drivers consisted of all taxi drivers from the same taxi company, then to conclude that all taxi drivers are anti-racist would clearly be going far beyond the evidence. If we encountered only one racist taxi driver in all our lives, this single case would be sufficient to undermine a generalisation of this kind.
This is a caricature of your opponent’s view set up simply so that you can knock it down. Literally a straw man is a dummy made of straw used for target practice. Setting up a straw man in argument is the opposite of playing devil’s advocate. Sometimes it is a deliberate ploy, in which case it is a disreputable form of rhetoric. More often it involves a degree of wishful thinking stemming from widespread reluctance to attribute great intelligence or subtlety to someone with whom you strongly disagree. Over-confidence in your own position may lead you to treat dissenting views as easy targets when in fact they may be more complex and resistant to simple attacks.
For example, Dr Johnson made a famous attack on Bishop Berkeley’s philosophy of idealism (which claimed that we can’t be sure of the continuing existence of unperceived physical objects except on the hypothesis that God continues to perceive them) by kicking a large stone and declaring: "I refute it". His point was that it was impossible to believe that something so solid was really just composed of ideas: but Johnson was mistaken if he really thought that Berkeley’s idealism would not be able to explain the fact that Johnson’s toe hit solid rock. Only a caricature of Berkeley’s views would be vulnerable to such a point. So Johnson had to set up a straw man. Whilst it is often tempting to set up and topple easy targets this is just cheap point scoring.
Take it further
Get to grips with the rights and wrongs as Nigel and guests explore Ethics Bites
Books by Nigel Warburton:
Thinking from A to Z
Philosophy: The Basics
Philosophy: The Classics
If you think you might be interested in studying more about these subjects, find out what the Open University has to offer.