We have never discovered a subatomic particle without first identifying the possibility of its presence mathematically. In fact, no new particles have been discovered incidentally since the basic subatomic particles decades ago; anything discovered more recently has been defined mathematically and then observed.
But if we observe something mathematically and then verify its existence through observation, did we really discover it, or did we simply prove what we already knew?
In order words, how strong is the observer’s bias in particle physics, when everything we discover through a particle field or accelerator, has already been calculated mathematically?
Observer bias (also called experimenter bias or research bias) is the tendency to see what we expect to see, or what we want to see.
And to most answering the two questions, the response would be a high degree of observer bias. But that is because we are on the outside looking in – meaning we display our own observer bias when we look into the world of particle physics.
But what if somebody were to look into our world, a world ravaged by COVID-19? What observer bias would they see?
Projections of COVID-19 have shown for months an increase in death rates or mortality as we move into the Fall and enter the Flu Season. But because we know the next few months will be brutal, will we somehow negate the effects, or at least the severity, because we are aware?
I am not suggesting we can wish this pandemic away – but will our collective awareness of what has come affect our response to what happens? For example, will we continue to stop caring and stop implementing the necessary behavioral precautions necessary to minimize the impact of the virus?
We have already seen signs of fatigue from months of restricted activity, of political unrest as the pandemic evolves into a rally-cry for independent rights – culminating into a dynamic blend of medical, legal, and economic chaos manifesting before our very own eyes.
Yet despite all the chaos, we continue to attempt to make sense of what we see. Much of this analysis helps us understand what is going on around us – whether right or wrong, the exercise serves to relieve the existential angst that arises from the unknown. We create narratives about the world around us – whether it is a mythological origin story, or a scientific conclusion, in the end, it is all a narrative. And narratives are critical in our understanding of COVID-19 – critical to how we observe this pandemic.
Therefore, in the process of creating the narrative, we should be keen not to succumb to the narrative we create for ourselves. A narrative filled with mental anxiety, sadness, depression only worsening during the time of COVID-19. Yet for far too many, this is the narrative we have seen.
And this narrative creates a counternarrative of indifference, of escapism in which people seek refuge through everyday activities that served as escapes that have worsened the pandemic.
Narratives give us perspective, as sociologist Studs Terkel noted decades ago when he chronicled the lives of hardworking Americans, sharing in their, “daily humiliations”, putting a face to a job title, and shedding light to the inner angst each felt. And narratives also help us develop new ways of thinking, which should be the goal of each patient encounter – to build the clinical relationship and add to the growing narrative. In that sense, the patient encounter is nothing more than a stream of narratives, built over time, encounter by encounter. The thought patterns form from the different interpretations and perceptions the provider and patient develop drive the course of the conversation. Eventually these patterns define the conversation and define how the patient relationship changes over time.
These changes, applied to an entire provider’s practice, explains how different providers develop different patient mixes over time. A provider that sees a lot of elderly patients will tend to get more referrals for similar patients. A hospital that sees patients from an insurance mix will likely see more patients from that insurance mix. These trends emerge from the thought patterns that create them as much as they reinforce the thought patterns. And even region biases grow to be ingrained as thought patterns.
A cough and rash in urban Baltimore will lead to a different clinical work up than similar symptoms presenting in southern Indiana, just north of the Ohio River. With the differences arising from different diseases that are more common in one area versus the other – in Baltimore, a sexually transmitted disease, and in southern Indiana, a fungal disease.
Data then, should be not be viewed outside of its appropriate context, which together comprises the narrative through which we understand what transpires around us. The narrative through which the data is built around is how we understand what is perceived and subsequently interpreted. And rightfully so – we form associations with words that are then reinforced by speech patterns that come from our thought patterns. Or, as Albert Ellis quipped – “how we think, we talk” – as the thought patterns create our interpretations, which influences how we communicate, and subsequently the stories we tell ourselves.
Yet the channel flows bidirectionally. And perhaps more importantly, the flow of conversation mirrors the flow of thoughts in reverse as well, which is the genesis of observation bias.
So if the government and community leaders are serious about stopping the pandemic, and encouraging safe behavior to mitigate the most harmful effects of the pandemic, we should do away with the data and starting telling the right stories.
For the stories we tell influence the thoughts we think – and what we think determines how we act.
Heatmaps showing opioid sub-epidemics by demography and urbanicity
Total number of deaths in each category from 1999 through 2016 are shown in the upper left corner of each plot. The colors indicate age-adjusted mortality rates per 100,000 people. (Synth Opioids OTM: synthetic opioids other than methadone. This category includes fentanyl and its analogs.)