Sunday 5 December 2021

The Nervous System

I was confronted recently with an incredulous response to the expression "to satisfy a demand". You might be able to satisfy a person, but a demand? An abstraction used in economics, like "need" (not unlike need, can anything be like need?), it comes via desire. In the pseudoscience of marketing, this goes from talk of market research into consumer preferences, whereby the claim is made that a need is recognized and can be fulfilled, to anticipating what a potential customer might want and, given the right circumstances, might think they want, whereby the market is identified for the target that it is and, for want of that want, desire itself is constructed solely for the purpose of creating a demand that can be lucratively satisfied.
 So many questions
This might seem counter to what's taught in school. If one wants to conduct reliable market research, biases must be held in check, which entails how questions should (not) be constructed as well as how questioners should (not) pose the questions. Important to note here is that a coherent art-form has been developed in which manipulation of the target is couched within a framework intended to appear unbiased. I'm not talking so much the deliberate attempt to conceal the corruption of a professional process, which certainly exists, as I am the natural tendency for humans to strive for survival of an assumed self interest. We all must have at some point in our lives chosen our words carefully, improvising a sort of cost-benefit analysis in service of what we think we want, or don't.
This little trip we're on is a key constituent of the woo world of economics. They call it "the business cycle". It's a reputed self-correcting, demonstrably Darwinian system that always leads back to financial hardship with better times on the horizon, at intervals defined in such a way as to fuel the expectations that do the cycling. It's sooo circular!
If you believe the survivors are the ones who in good times prepare themselves for the inevitable decline and then avail themselves of prudent efficiency should the former preparation not suffice, you've been a good student. BBA's in the mail. Yet if you also believe in the alchemy of advertising, there must be a suicidal inability to acknowledge the difference between what you need and what you'd better do without. Is there anyone who hasn't yet heard the story of the lab animal who chose the cocaine lever over the food lever until it starved to death? I hear a voice that says "likely apocryphal". Well, it either is or isn't, idn't it? The statistician must love the unknown, spinning plausibilities out of their client's market's threadbare imagination.
The market here is the so-named consumer. But if you anticipate the actions of each enterprise based on self-interest, each actor has its own market with its own specific demand to satisfy. 

  Revolutionary gaming console
Widget builder bigwigs engage mad men who liaise with people pollers to see which widgets have the potential to be moved most in the coming season. Theoretically, the success of the former relies on each thereafter. It's foremost in the interest of the people pollers that the mad men believe they've got an accurate look into the people's heads. It's foremost in the interest of the mad men that the bigwigs believe their funny teevee spots will move widgets like nobody's business. And, in other words, if widgets do move, it's critical that each can take due credit based upon their confidence in the others.
But what if the creative types have their own ideas, as creative types tend to do? They will after all be analysing the data for themselves and not just following some mathematical formula self-evident in the numbers. A mad man engaged to Big Widget might be secretly married to an idea. He'll nod his head at what the "numbers say" or take the people pollers aside and have them rephrase their questions. The mad man might be right about what will best make the market trample over itself to get this one particular widget. And the probability of that happening is tied to that one mad man's guttural understanding of human psychology. Or luck.
There goes the likelihood.
The accuracy of a random sample survey appears to confirm the results of an earlier questionnaire that had indicated a strong public preference for things asked about in surveys. In the original survey, 73 percent of respondents had indicated that they preferred something or another, with "pretty much absolutely" being the majority at 42 percent, 19 percent responding with "more or less kind of" and "How should I know?" coming in last at 12 percent, which was placed in the affirmative category because it seemed more affirmative than the most common response of those who declined to participate.
In the abracadabra of the opinion poll, a sample size represents everyone and ignores everyone else. The latter are on the wrong side of the response rate. A low response rate is problematic. The non-respondents are the problem. Statisticians worth their money know how to explain how they correct for this problem should it come to the need for an explanation. A statistician eager to please knows you can't please everyone and hence regards apparitions as unrelated to matter, let alone the matter at hand. The key is a high enough response rate. Where that fails, a correspondingly larger sample size can obscure what's missing. If a sample size of repsondents can be representative, would not then non-repsondents be representative as well? Whatever that says, it doesn't say. Interesting as regards what's missing is that therefrom all kinds of plausibility can be manufactured. Simply use your imagination.
After the flood
And it's not just with the tools in legacy media anymore. Ironically it's their decades-long demonstration of the craft of fact fudging and fealty first to state/corporate ownership that underpins the most incredibly broad banded array of creative guesswork that, for those lapping it up, is supposed to represent some kind of grass roots alternative to establishment journalism. Theoretically, the dam has broken and information can no longer be suppressed. But because of the precedent that is counter-intelligence, every crackpot has a built-in excuse whenever they're tainted by association with inconvenient contradictions. Anyone and everyone could be an agent of some kind, either spinning their fantastical disinformation themselves, or simply sponsoring some muddy narrator by cooptation. You can never really know. And the most difficult part to get around, or not, depending on your bias, is that the liar does not become a truth-teller just because someone else is telling the most absurd lies about them. Yet somehow the relative plausibility rules.
And how can the long-established precedent of state secret keeping do anything but fuel speculation into all sorts of silliness of diminishing plausibility? An independent audit of a bookkeeping endeavor is impossible whenever patriotic lawfare wields the built-in excuse that "we don't want the enemy to know what we're doing". This cannot eliminate the likelihood, with respect to whoever it is we think we are (or "we" is), that anyone who doesn't know whence & why these moneyed efforts flow, is an assumed enemy. Or a target market.

 United in common currency
A warning from your own agency about an outside threat (economic, organic, or otherwise) conditioned upon a failure of those acting in your stead from taking the agency's recommended action is a threat in the form of an IOU the agency will present to you for payment should yours have chosen not to borrow, or deliver its verdict whether or not you decide to answer any questions in your defense.
It's easy to discount the laughably low turnout in the streets of an absurd faction one'd otherwise become nervous about. Insert a smirk of relief because, anyway, look at the misspelling on that protest sign. Alternatively, insert your shock because the headline says you're shocked. Or at least a metonym that represents you is said to be. Certainly this small sample, too small to be considered ragtag, does not represent the 27 percent. Who would? They're just the ones who dared to answer in the negative on the questionnaire, but won't show their faces. Then there's everyone else. Anybody got a plausible response rate?