WhatIsEffectiveness

Michele McCarthy has always - unconsciously or consciously - pursued the goal of increasing her own personal effectiveness. To Michele, effectiveness is the ratio of effort to results

As I read Michele's statement, I had a strong negative reaction, which is always a good sign that I'm about to learn something.

A high ratio of effort to results may help identify personal management that is inefficient and ineffective. But, I have a strong hunch that a low ratio of effort to results tells me little about effectivenss. For instance, I spend time writing but I have few published words, which gives me a high ratio of effort to results. Although I may be inefficient, I feel I am highly effective because I'm engaged in an activity that is important to me. So, at the momement, effectiveness seems to me to be about producing the results that matter to me.

What does personal effectiveness mean to you?

I think the word to define above is results. If results = output volume, then your nervousness is justified. If results = requirements, then the ratio makes good sense.

I would worry that short-term requirements may lead to long term negative consequences (the ecological view) but that means that requirements need scrutinizing for hidden assumptions, and someone has to advocate the long term viewpoint. That sounds like an architect to me.

BobLee 2002.04.18

I wonder. I view my writing as a desire rather than a requirement. I get less fulfillment from the results when I must write to satisfy a deadline, which I see as someone else's desire. Note, I am talking about my own personal effectiveness. The most effective results for me are the ones that bring me the most joy.

SteveSmith 2002.04.18

I think Steve's hit on a key point, but I don't think he'd agree with the statement, "If it feels good, do it." I believe effectiveness to be elusive and very context dependent. Here my thoughts parallel Jerry's look at quality in QSMI. As a tester, I frequently struggle with the ideas that non-testers have about testing. To some of them, if I run lots of tests and find no anomalies, they feel very comfortable about the item being tested. So for them, the testing has been effective. For me, it's very frustrating as I have no idea what the quality of the item is. For me, testing effectiveness, is running a smallish number of tests and finding numerous significant anomalies that form a coherent pattern. Yet doing so puts me at risk of a judgement of low effectiveness on the non-testers part. This is a relatively obvious example, but most of the environments, I've worked in, favor the first definition of effectiveness. That definition then gets extended to run the largest number of tests in the shortest possible time, preferably overnight, and still find no anomalies. Having said all that, I also believe that the above non-tester definition is appropriate in the context of regression testing. MikeMelendez 2002.04.18

Well, Quality has its "ilities", so does effectiveness have its "nesses"? I think that creative laziness is one key to effectiveness. Like "lazy evaluation" in algorithms, it avoids work that doesn't produce, and looks for better/faster/easier ways to do things.

Without creative laziness, when your wife asks you what you want for dinner, you would have to answer, "I don't know, I haven't killed it yet!"

BobLee 2002.04.18

Tonight, I discovered the problem I was having with the idea that effectiveness is the ratio of effort to results.

I was in a shopping mall when I heard a child crying. From what I observed across the mall, the child was using the standard child threat of I'll scream unless you go into the store I want. I've faced this threat many times and I'm sure many of you have too. The parents were retaliating with the powerful parent threat of I'll leave you here by yourself unless you come with us. I watched as the parents turned their backs to the child and started walking off. The little boy started crying and struggled to catch back up with his parents. The whole scene broke my heart.

If I measured the effectiveness of the parent's tactics by using the ratio of effort to results, I would conclude that the parents' actions were highly effective. The little boy quickly did what they desired. But, we must look at more than just that moment in time. Incidents, like this one, may leave the boy feeling that his parents may abandon him. If you, like me, feel that a child should never have to fear that his parents would ever abandon him, than we must conclude the parents' actions were highly ineffective.

I'm sure that there is a long list of actions that are momentarily effective but when viewed over a longer period, are seen as highly ineffective. Please add some of your own examples to this thread.

Let me return the example of writing that I used earlier in the thread. Although my ratio of effort to results is high at the moment, over the long term, for many reasons, I believe my efforts to my results are just right.

SteveSmith 2002.04.20

I don't see that effectiveness has anything to do with effort. Effectiveness is about the value of results. If you and I get the same result, but you put in twice the effort I do, you are half as efficient, but just as effective.

I'm scratching my head as I think about comparison points for judging efficacy. Comparisons seem important and dangerous. SteveSmith 2002.04.21

I've been aware of the concept and term of "bebugging" for almost 20 years. I didn't know, but I am not surprised that Jerry coined it. The way I learned it was as a measure of your test coverage. You randomly add known bugs and run your tests. The percentage of the known bugs not-found gives you a rough measure of the real bugs that remain. The accuracy of the measure depends on how realistic the known bugs are in composition and distribution, hence, as Jerry notes, the measure is very rough. "Fault injection" is a current buzzword where I work, as in "We need more fault injection." As I understand local usage, the term refers to simplifying negative testing. Build in the capacity to introduce an error in the code being developed, so the test can check if the code properly handles the fault without having to introduce a much larger context. In making tests simpler it allows you to increase your coverage, by allowing time to write more tests. These are related but distinct concepts. Keith, was your use of "fault injection" close to one of these?

See BeBuggingForTesters discussion.

That said, I have never worked anywhere where time was available for bebugging, though I have suggested it on a number of occassions to my management. In most cases, they much prefered that I simply run more tests or (sigh) the same ones repeatedly. Though "fault injection" gets talked about here, most developers seem to believe that they do not have the time to code it. I have hear of "fault injection" software, but I have difficulty imagining that it could be generalized, so have assumed it to be domain-specific. MikeMelendez 2002.4.22

I have never worked anywhere where time was available for________, though I have suggested it on a number of occassions to my management.

Mike, I've generalized your statement. Fill in the blank and it's a common theme for individuals in too many organizations. I started the NoTimeAvailForDifferentMethods thread to discuss the problem.

SteveSmith 2002.04.22

Moved BeBuggingForTesters discussion to a separate page. BobLee 2002.04.22

Updated: Monday, April 22, 2002