I have been churning on the idea of ‘tests as conversations’ for the last week or so. One part of this involved the skills necessary for the conversation to happen. Specifically whether knowing math and concluded that no, you do not needed to know math. Except when you do.

Huh? That’s pretty have-it-both-ways even for me.

Math is a ridiculously large field of study, so it is near impossible to say that you do not need any math, but the slice of math I think testers need to know is pretty slim.

- Basic statistics (mean, median, mode, standard deviation, sample size, etc.)
- Combinations
- Permutations

Yup. 3 things, though I’ll concede that the first one is a decent sized bucket. In over 10 years of testing I have not had to employ any math outside of those areas. Even when testing a banking application I did not have to flex my math brain as the bankers created oracles for us to verify against.

Of course, this is heavily dependent on my own experiences. I have never tested an application that is heavy in its math, like a missile guidance system for instance. I would likely argue that at that point what you want is not someone who is conversant in math, but someone who is deeply fluent in it.

I have been out of high school 12 (or 13 years; I can’t remember) which was the last time I had to do any serious math. I can still do long division, but I had to teach it to my daughter when she was in public school. Calculating the square root of something? Not a chance.

And that is ok. In no way does this lack of knowledge impact my ability to have the conversations I need to.

on 08 Sep 2008 at 5:36 pm1AnonymousIMO, knowing math may or may not be necessary depending on the field but having known math, i.e. having been able to do high-school math at one point and the ability to look up and quickly relearn whatever is necessary, is mandatory.

Given that it also falls to senior testers to check the assumptions of the designers and to do the math that they often^H^H^H^H^Hsometimes don’t do, having a ready command of math in that role is vital. (I recall a team which designed a service that would have consumed more bandwidth than the Internet backbone could have provided. They were rather indignant when I pointed this out.)