Home » Uncategorized

Preambles and Future Directions

I will be using this blog to assemble a number of different concepts that I introduced over many years in previous blogs (indicated in bold); then I will explain where all of this will be going in the future.  I am turning 50 years old in a couple of weeks, and I find that I habitually take inventory of my belongings these days before beginning any lengthy mission or journey.  I recently acquired a fairly expensive device called a CPAP machine.  It resembles a small stereo with a long hose at the end of which there is an attachment for my nose.  I am delighted that my work benefits and the province subsidized the cost.  As I understand the situation, most people periodically stop breathing in their sleep although it normally happens only a few times per hour.  But for some individuals, such incidents occur much more often.  A doctor prescribed me the machine to help me sleep more soundly.  Having successfully slept using this device, I find that it certainly helps.  But on some days, the reading is exceptionally good – e.g. nearly no incidents per hour.  On other days, the reading although still much better than pre-CPAP seems unusually higher.  Therefore, I wondered how the reading can be excellent on some days and significantly worse on others.  How might a person improve the situation?

I’m using this personal experience in lieu of a common business scenario:  why are sales great on some days while terrible on others?  In order to engage this type of question, it is necessary to distinguish between descriptive data and prescriptive data.  A blog that I can remember writing many years ago is about this difference.  Prescriptive data is about controlling processes and ensuring conformance.  An example of prescriptive data can be found on the image below on the upper frame:  beats per minute from the left hand at 70; beats per minute from the right hand at 68.5; non-breathing incidents of 3.3 per hour.  The purpose of this data is to delineate the problem using externally imposed metrics – usually a scale – in order to establish if the levels are desirable or undesirable.  However, if the non-breathing incidents jump from 3.3 per hour to, say, to 6 or 9 per hour, prescriptive data fails to provide any explanation.  The only assertions that can be made using metrics alone would be in terms of trends, distributions, and correlations.  In other words, quantitative analysis using statistics and technicals is possible through the use of prescriptive data.  The prescriptive data below is from a mass data object from a system called Elmira.


On the other hand, the lower frame on the image contains descriptive data, which I have also called structural data.  It is structural because it indicates its relationship to me and therefore how I might alter my behaviours to affect outcomes.  One might think of it as the deconstruction of outcomes.  Conceptually, descriptive data is made up of the events that give rise to that which the prescriptive data is meant to measure.  A quick review of the events reveals that, among other things, I am interested in whether the foods that I eat and the medications that I take influence the number of non-breathing incidents.  How does a person determine what events to include?  This is a critical question likely to trigger deep and meandering journeys; it is the philosophy behind the numbers.  Theoretically, the events chosen for descriptive data should lead to the outcomes apparent in the prescriptive data.  I wonder if anybody can determine the event-recognition or -distribution methodology that I chose.  If I only had medication among the events, I think many people would suggest that I am listing the medications that I take.  On a more abstract level, I am using “notable quotidian life events,” which I would describe as a type of codified narrative.  It is possible to use types of paths or narrative for ontological patterns in order to invoke descriptive events.

Structural data is important in relation to an unconvinced audience.  More often than not, this is an argument about metrics.  Some scientists accept that climate change is inevitable – and they are offering advice on how organizations and nations can adapt to rapid developments.  Other scientists, if they can be described as such, argue about the horrors of climate change in terms of the likely losses and damages, costs, and loss of life.  I guess one should not necessarily expect a great deal from people “good with numbers” but who might not actually have the foggiest idea about how to deal with climate change.  Consequently, there is a structural and non-structural approach to engage climate change.  Structural data begets a structural response: e.g. where to build bridges; locate floodplains; construct buildings and highways.  In the context of collecting structural data, it can be worthwhile applying life narrative in order to generate data:  how real people are being affected in their day-to-day lives by climate developments; how are people adapting to and coping with their harsh realities.  While it is possible to apply a lens focused on metrics, there are limits to legitimacy that can be gained by a person hollering wolf without an actual understanding of the underlying problem.  A call for a structural response necessitates structural data.

I have said that using metrics leads an analyst to only half of the story, but actually the metrics only represent the tip of the iceberg.  I have written about the pathology of disassociation between metrics and events.  I know that data scientists have been indoctrinated into the world of business metrics.  There might be a great deal of discussion and debate on what metrics are important.  However, even before this level of analysis, much can be said on how organizations might 1) insulate events from the realities of the market; 2) at times substitute those events; and then for full adulteration 3) promote events that are alienated.  In effect, even if metrics were perfect, the extent to which they can be connected to the organizational body must be questioned.  An organization might be metrics-progressive but phenomena-insensitive.  This is a disease of the organization comparable to intoxication or drug addiction.  Consequently, I have often written about aspects of organizational pathology associated with data alienation.  A component of systems theory as applied to ecology is the biofeedback mechanism.  Within an organizational context, it is possible sever the mechanism such that production becomes disconnected from its actions, the market, and possibly its own employees.

When Ford was producing the Model T, at some point probably due to declining sales he had to accept the reality that people no longer wished to purchase it.  It became necessary for him to initiate a process of transformation not just of the production environment in order to create new cars but also deployment of human resources.  Imagine how much more difficult the undertaking would be if he had a lot of big plans but little understanding of his business market and production environment.  However, at times there is an absence of data to guide the process; and decision-makers must return to the question of what data to collect in order to carry the organization forward.  This is the non-pathological process of embracing change through methodic thinking and deliberation.  I have suggested, however, that the data revolution makes it possible to essentially weaponize data – to power a process of exploration through brute strength.  I am sure this seems fairly empowering except that it likely impossible using metrics on the absence of event data.  It is necessary to expand data from its edges.  I explained how edge theory found in ecology can be applied to data.  Yet it remains necessary to also make effective use of metrics.  Recognition is a process of detection and delineation.

Future Directions

I find that business people often emphasize the importance of common sense and keeping things simple.  Simplicity can be the outcome of extremely complex processes.  (This is me rephrasing somebody quoting Steve Jobs.)  If the idea is for management to make use of only common sense and to keep things only simple, many managers could be easily replaced or simply removed from the process.  Much of the world’s population is fairly plain and poorly educated, after all.  With common sense, there would hardly be a need for computers, big data, or analysts.  I would therefore argue that my original argument remains intact:  it is necessary to put some effort ascertaining what data – specifically, event data – to collect.  I thought of an analogy.  I recently renewed my hunting licence although I don’t really hunt.  It is the expectation of moving north after I retire that caused me to do the hunting exam perhaps about 20 years ago.  I have certainly fished.  Whether it is hunting or fishing, there is the following sort of question:  where should and what should I do in order to catch the fish (turkey, deer, or moose)?  When I was a child, I used to wonder if I could think like a fish; and if so, where might I as a fish choose to go?  I would ask myself whether I would find my bait appealing; whether its movement might induce me to bite; whether I might be hungry or full during certain times of the day.

In the absence of data, first of all it is important not to panic; secondly, I feel that it can be worthwhile to approach the problem as a hunter.  Hunter will instinctively know about edge theory, biofeedback mechanisms, and the use of narrative.  “If I were a terrorist, how might I evade capture; where might I go; how can I use local patterns of behaviour to confuse the hunter?”  However, such processes are not necessarily common in relation to data when it is necessary to adapt a detection regime to compile events.  The future as I expect it for me involves the strategic use of ontological patterns in order to compile event data.  But it is going to be fun.  It will likely involve trapping, aiming for moving targets, and selling skins.  At the same time, I recognize that I have been blogging about event data; and there might not be much of a market given that so many people don’t even know what it is.  But that’s fine.  The things we do in solitude sometimes help to build the character of a person rather than satisfy the needs of the market.  Nonetheless, I hope that people eventually reflect on the largely unexplored edge and come to conclude that future of civilization might be found there.