What is a Round-robin test?
In: 2
Let’s say you wanted to test something, for example the purity of the air in a building. So you hire a company to perform the test and you get an answer. What problems might you have?
1. the test is a single event, air purity might change hour by hour or day by day.
2. the test is a single location, air purity might vary from place to place
3. the test can’t test itself, maybe the machine malfunctioned or the technician screwed up?.
4. the test isn’t super precise, maybe it’s only good within a + or – 25% value
5. the test is a single method, maybe no one knows that that specific test gives bad results due to some odd feature of the location – like maybe the lights are too bright or there are vibrations for a subway or something that makes a good test give poor answers, who knows.
Etc. So taking a single test, a single way, at a single time, by a single group has many possible ways to give wrong data. Let’s also say you’re testing something super important, like a potential cure for cancer. Do you want to leave it to up to those kinds of possible errors?
A “Round-Robin Test” is a method of testing something in which you try and eliminate all those problems by repeating the test over and over again; by multiple different humans, by multiple different methods, in multiple different places/times, with multiple different of everything that you think only have a single version of could be problematic.
EDIT – as an aside, it’s also good way of testing tests. Let’s say you do 10 tests on something that should all give you same answer but one of them gives you a wildly different result. What’s up? Maybe you just discovered something like example 5 where you really shouldn’t use Test F in this specific case because something screws it up.
Latest Answers