TEST SCRIPT WRITERS DON’T EXIST...

I’ve read Karen Johnson’s blog about test script writers (catch it on testingreflections.com), and found it both understandable and sad at the same time.

Generally, I like it when someone has some passion about a given topic, even when I don’t agree with it. Everyone is entitled to have an opinion and some of them are pretty interesting.

But this one, well, it was oddly slanted and it kind of tickled at me. I stuck the whole thing in my blog fodder file and read it several times.

She was interviewing candidates for her team and went ballistic over a candidate that said he read specifications and wrote test scripts. I think what she really didn’t like was the “slow, steady” response. It seems likely she hires dynamic people who can think on their feet. I understood and agreed with her statements about having some passion about what you do. But her post inferred people who write scripts are the mental equivalent of doorknobs.

She used the comparison that she bought a camera after an interview to become a journalist. She lied during her interview and said she had a camera. If you really look at that comparison, she DIDN’T HAVE THE CAMERA to start with; she had to go get one.

Someone who has never worked in an environment that either does or does not require test cases doesn’t have a camera either. You have to be able to assess whether they have or will be able to use a camera (i.e. is able to think) and that’s a really hard job.

There’s really no such thing as a “test script writer”; some automaton that magically transforms specifications into test cases with no analysis. If you review someone’s test cases, you’ll see their test ideas in writing. If someone has no test cases or nothing you can look at, you’ll either have to probe to see how they think, or just assume that if they appear to be quick on their feet and talk a good story, they’re smart enough to be a good tester. Overall, gauging talent that way is a crap shoot. I know testers that develop test cases that I wouldn’t have on my team. I know agile testers that are just as bad. Is testing methodology really a good indicator of testing ability or intelligence?

A test case, or script, is nothing more than a test idea fleshed out with some instructions. As such, the analysis skills required to translate a requirement (whether written in a spec or brought up in a meeting) into a test idea(s) are pretty much the same. I think the core issue here is “are the test ideas any good”? The format is meaningless in comparison.

I feel a little sorry for the applicant that interviewed with her, whether they were qualified for the job or not. If the rest of her company uses test cases and her area doesn’t, the applicant was offering up what was viewed as successful in other areas of the company and would have no way of knowing Things Are Done Differently Here, unless she talked about how her group operates before they got started. If she did that and he didn’t adjust his delivery accordingly, then he probably WASN’T quick enough to impress her much. I’m not convinced that IQ or personality is linked to methodologies, however.

Having test personnel that understand the company and deliverables and who are “plug and play” is, in fact, a benefit for the company. Personnel here don’t get passed around, as we have a centralized QA organization, but we pretty much reassign staff as project needs dictate. I have a preference for dynamic people who are quick on their feet as well. We’ve got a mixed bag in terms of project methodologies here; some structured work and some agile, with most falling somewhere in the middle. So what I value most in my staff is flexibility. I feel the same way about interviewees.

Karen’s blog indicated that paying “script writers” peanuts and giving them no respect was justifiable. I don’t understand how any rational person could really believe that.

Then again, maybe I’m a doorknob.