Courtesy bias can often affect training participants' responses to evaluation questions. Assuring participants that evaluation responses will be kept confidential or made anonymous, by leaving names off evaluation forms, may encourage participants to respond more frankly. In addition, since evaluation forms are not always the best way of evaluating a training (due to a number of factors including courtesy bias, low response rates, and the difficulty of self-reporting on the effects of training that was only just received), other methods may be used to gauge learning and improvements in performance. For example, after training people to use an information technology, trainers could observe the trainees conducting a search on their own or use an online knowledge resource to track usage patterns. This observation could be conducted several weeks after training, if possible, as a measure of whether new knowledge was retained.