Distinguishing Between the Quality of Data and Its Quantity With Stoichiometry

Let me jump right into this discussion without a prelude explaining any of it. The explanations would take too long. And no one who's interested in this wants to read too much. Everyone else who doesn't get it at first glance will just skip it.

+++++++++++++++++

It is essential to know both the quality and quantity of a given set of data. Any measure of unit counts as data for this conversation. Know only the quantity or amount of data and not its essential quality is foolish, as only the wise know.

Many folks I work with in Quality Assurance care only about the amount of defects they have in pre-production. As if a body count alone can tell you if you're winning or losing the war. Vietnam should have taught us that much!

I tell them that the number of defects matters nothing. The requirements that the defects impact is all that matters. No more, and no less.

If it is required that a software application work in such a way, it should matter that it doesn't work in that way. Not working as required is a defect. Therefore, the best way to present defects is to tie them to their associated requirements. Surprisingly, this is a tough sell.

So let me put my argument another way. Let's use something as objective and impersonal as anything: chemistry's simple equation for stoicheometry. The most basic equation is that an acid combined with a base will form water and salt.

acid + base = water + salt

I had my "Aha!" moment in Qualitative Chemistry with this realization. Whenever we got bitten by a mosquito, my mother would tell us to apply ammonia on it because it takes away the itch. Windex window cleaner works as well as straight ammonia because it is made up of ammonia - so it has the same active ingredient.

The night before this enlightening lecture, I watched a documentary (while studying chemistry of all things) on why bug bites sting. It also explained why some stings hurt more than others. It turns out that bug bites are acidic and the acid burns our skin as it dissolves it.

The molarity or concentration of the acid is what makes more acidic bug bites sting worse than less acidic bug bites. Gnats sting the least, mosquitoes more, wasps a hell of a lot, and scorpions the worst of all. Increasing acidity hurts more like more concentrated vinegar (acetic acid) tastes more bitter and sour lemons (ascorbic acid) make you pucker.

A mosquito bite has the same acidity as vinegar. In terms of molarity, the concentration of the solution, vinegar has an equivalent molarity as ammonia. The base and acid combines to form water and salt.

That's why dabbing ammonia on a mosquito bite takes away the sting. And why after you dab the ammonia on the bite, you get a droplet of water forming over the spot. When the water evaporates, you flake off the salt. After all of this, the sting is gone!

Qualitative Chemistry:
acid (mosquito bite) + base (ammonia application) = water (droplet bubble) + salt (resulting crust)

What I described above can be understood as Requirements. According to the laws of chemistry, mixing an acid and base yields water and salt. When that doesn't happen, it is a defect.

You must understand stoicheometry in order to detect a defect in the equation. Only then can you qualify the defect. But understanding only the defect without knowing the requirements is stupid.

Quantitative Chemistry:
HCl (hydrochloric acid) + NaOH (sodium hydroxide - base) = H2O (water) + NaCl (sodium chloride - salt)

People who focus solely on the defects and not on the total equation with requirements itself focus on the horse's ass and not on the horse itself. Defects are the horse's ass and
Requirements are the horse. Recognizing something does not mean you see everything.

It's like seeing a confusing equation above and thinking that just because you recognize the elements on the Periodic Table means you understand chemistry itself. As if that makes you an expert at Stoichiometry when it doesn't because you fail to grasp the significance of its elegant equation.

No comments:

Post a Comment