Some weaknesses of a response-strength account of reinforcer effects