When performing manual tests, it is possible to make full use of the human ability to temporarily come up with new tests when executing a certain TCD (test case definition), which is a thought divergence for testers. Automated testing does not allow the computer to do everything that the tester does, because the test script is clearly described in terms of the TCD steps and cannot take advantage of the knowledge and cognition implied by the tester. The test scripts run each time in the same content, in the same order, the exact same number of the same thing as the mobile and keyboard operations. But manual testing can be a bit different every time you run the test, which is not possible with the test script.
The results of automated tests (result cheking point) are also limited. For example, when playing the video, it can not detect whether the sound is glitch, the interface is huaping (speaking of this, there may be some people will come out to refute, said I have a way to detect whether the video screen!) Yes, we also want to use a mechanism to detect the quality of video playback time, the final conclusion is to give up, later specifically to say a video quality test means)? Maybe not. But a good tester will pay attention and report the problem. The brain of a professionally trained person is the best test tool to exceed any conceivable automated test. Testers can also identify and filter out a wide range of false alarms, which are impossible for automated testing.
Therefore, instead of manual testing compared to automated testing, it should be said that automated testing is an extension of manual testing, to complete the manual test can not complete the work, stress testing, performance testing.
Do not equate manual testing with automated testing!