أرسل هذا في رسالة قصيرة: Utility of Crowdsourced User Experiments for Measuring the Central Tendency of User Performance: A Case of Error-Rate Model Evaluation in a Pointing Task