Be sure to read our best halogen oven post.
One way to distinguish between classical and operant conditioning is to look at the Law of Effect (Thorndike, 1898):
LAW OF EFFECT
Basically, the law states that learning takes place because a connection is made between the stimulus and the response and because this connection produces pleasurable consequences. Thorndike stated that the stimulus responses connection is “stamped in” if the result is pleasure and classical conditioning is the pairing of the CS and the UCS. However, in operant conditioning, the thing that produces the effect (the thing that is learnt) is what happens as a result of behaviour: the connection between behaviour and its consequences. So, the law of effect was extended to Skinner (1938) who said that behaviour is shaped and maintained by its consequences. This means that it is the consequence of the behaviour that will determine whether it is reinforced (and therefore more likely to happen again) or not.
You will occasionally see this referred to as instrumental conditioning. Operant conditioning was introduced to the world by B F Skinner (1938). He was much less interested in the theoretical side of learning than the classical conditioning theorists like Watson and Pavlov. Skinner wanted to find out how animals operate on their environment, how this operant behaves is instrumental in bringing about certain consequences and how these consequences determine the probability of this behaviour happening again. Skinner saw the learner as much more active than classical conditioning proponents.
What is an operant? It is a voluntary behaviour that is controlled by its consequences. Any consequences that makes a behaviour more likely to occur again is called a reinforcer. If the reinforcer results in something good happening then it is a positive reinforcer, if it results in getting rid of something bad then it is a negative reinforcer. The reinforcement process does to just work on desirable responses. Reinforcement increases the likelihood of any behaviour that leads to pleasant consequences, whether the behaviour is appropriate, inappropriate or neutral.
Not all consequences of behaviour are reinforcing, though. Behaviour sometimes produces effects that are unpleasant. These unpleasant consequences will make the behaviour less likely to happen again. Such consequences are called punishers. We usually think of punishment as a particular form of discipline; but to learning theorists, punishment is just like its counterpart (reinforcement) in that it is simply part of nature’s learning process that teaches organisms which responses it is wise to repeat and which it is better to avoid.
Punishment, like reinforcement, can involve positive and negative types. Positive punishers are punishing consequences that involve the presentation of something unpleasant following a behaviour. Negative punishers are punishing consequences that involve the removal of something pleasant following a behaviour. The thing that constitutes a punishing consequence can vary from child to child. This point is important because it is common for an adult to assume that they are administering a punishment when, in fact, they might be reinforcing a behaviour. For example, in the classroom, a teacher might view telling a child off for clowning around as a punishment. However, the child might see this scolding as attention and attention can be a powerful reinforcer for a child. This response of attention may make the clowning around behaviour more likely to happen again.
So far, it sounds as if whatever has been reinforced should carry on happening forever. That is not the case. If you stop reinforcing a behaviour then the behaviour will eventually stop. This process is called extinction. You might want to argue that behaviours in the real world are rarely reinforced consistently. Learning theorists agree with this and have tried out what they call different schedules of reinforcement. For some animals, reinforcement is consistent and continuous, for others partial reinforcement is used, where behaviour is reinforced on some occasions but not on others. Studies with animals (and with humans) have shown that partial reinforcement schedules result in a longer time period for learning to take place. However, once something has been learned under partial reinforcement, it takes much longer for extinction to occur.