Blocking of the response-reinforcer association: Additional evidence☆