Alternatives to switch-cost scoring in the task-switching paradigm: Their reliability and increased validity