Stories
Slash Boxes
Comments
NOTE: use Perl; is on undef hiatus. You can read content, but you can't post it. More info will be forthcoming forthcomingly.

All the Perl that's Practical to Extract and Report

The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.

Why There Are No Conclusive Studies about TDD Efficacy 5 Comments More | Login | Reply /

 Full
 Abbreviated
 Hidden
More | Login | Reply
Loading... please wait.

  • Frankly, I've no problem with you using Anecdote Driven Development. I have every problem with testing advocates tell me I should be relying on their anecdotes.

    This will continue to be true until it's possible to measure the cost of creating and maintaining a piece of software throughout its lifecycle.

    What you're actually saying here is "it's very, very hard to accurately measure this cost, so ...". I do agree with the "very hard" bit (well, I might say "impossible"), but the same is true for just about any complicated research out there. It's not the case of pouring A into B and determining if C results. Studies in a variety of fields require

    • Studies in a variety of fields require acknowledging that things can't be done perfectly, but we can do enough to get a sense of how things actually work to make sane recommendations.

      You might as well quote a study that says "Berlitz German students translate the libretto of The Magic Flute twice as fast with 80% fewer errors than students of other language learning schools." ETOOMANYUNCONTROLLEDVARIABLES.

      • I can't tell what you mean due to your sarcasm. I know you're not saying "the benefits of TDD are too hard to study, so we won't even try" because that would be an idiotic statement to make and you're not an idiot. So what are you trying to say? Are you saying that the efficacy of TDD is beyond reason and we must rely on anecdote? Again, I know you're not saying this, but you're not saying much else, either :)

        In short, if you're going to make cause and effect assertions, I want to hear something backing

        • In short, if you're going to make cause and effect assertions, I want to hear something backing that up other than anecdotes.

          I believe you're in the wrong line of work then. How would you even design such a study?

          If you compare two programmers, one using TDD and one not, you have to account for productivity, knowledge, experience, and creative differences between them.

          If you compare two teams, one using TDD and one not, you have to account for the same, multiplied by at least the number of programmers on

          • If you compare two programmers, one using TDD and one not, you have to account for productivity, knowledge, experience, and creative differences between them.

            Sociologists deal with that sort of thing all the time. Their methods aren't perfect, but they do manage to get somewhere.

            The secret is that you don't compare two programmers; you compare two thousand.

Stories, comments, journals, and other submissions on use Perl; are Copyright 1998-2006, their respective owners.