Making risky decisions with imperfect data and why white coats can harm you
? Gary Nuttall MBCS CITP
Helping organisations to extract more value from data | Business Analysis & Project Management | Online training, workshops and emerging technology webinars and masterclasses | Keynote speaker
How do you and your organisation make decisions ??Typically it involves gathering data, reviewing options – looking at benefits, risks, probabilities etc.?It’s all data driven isn’t it ??Well, in a perfect world it is.?But what happens if a decision is needed urgently and you don’t have all of the data to make a properly informed choice ??This is where psychology comes into play and you hear comments like “gut feel” or you’ll defer to a choice recommended by an expert.
When you explore the deferral to authority approach you become aware of two significant risks it introduces.?Firstly it’s amazing just how much trust we will put in experts or those in authority.??Sometimes that trust is actually obedience and Social psychologist Stanley Milgram’s experiment on the “white coat effect” ?is worth reading (https://nature.berkeley.edu/ucce50/ag-labor/7article/article35.htm).?Secondly, the experts may actually be no better informed and are themselves influenced by their superiors (intellectually or organisationally). This is where safety critical industries rely on tests, trials and regulators to ensure safety and efficacy of decision making.
Let’s try this out with our own experiment and a few scenarios…….
There’s an emergency and you need to make a decision IMMEDIATELY!?You’ve swallowed something poisonous and the doctors say unless you act quickly and take an antidote then you may die or experience significant harm.?What would you do ? ??In these circumstances you will rely on the expertise and judgement in the doctors.?You wouldn’t start asking for the data would you ?
Let’s play with the parameters a little.?You’re now told that you have swallowed something that MAY be harmful.?The antidote is experimental and may itself cause harm.?Now you’ll start asking for more data.?Such as the likely risk/harm.?You’ll probably still listen to the doctors though.
Now, in our third scenario you discover that you’ve swallowed something that, as long as you’re fairly healthy, won’t harm you.?And the antidote may have side effects.?What would you do now ??Many people will still defer to the experts.?But what if the experts don’t know either ??They’ll defer to their regulators and experts and assume that they’ve done the research to ensure that the antidote is safe and that it is effective. ??This is why drugs have to complete rigorous phase 1, 2 and 3 clinical trials, the results of which are independently peer reviewed. Except when there's an emergency.
领英推荐
So we can see how our judgement, particularly in critical situation is heavily influenced by others.
Fascinating therefore to see that a study has just been published (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9428332/) which shows that in a recent emergency situation the harm of the antidote consistently outweighed the reduced risk. In the case of one of the products examined, the excess risk of serious AESIs 10.1 per 10,000. That's one serious adverse reaction per 910 injections.
However, we should always cross-validate our data if we can!?It’s reassuring therefore to see that the white coated experts have done something similar and in October 2022 the UK Joint Committee on Vaccination and Immunisation (JCVI) was briefed ?by the UK Health Security Agency (HSA) about what is known as an NNV.?This is the Number Needed to Vaccinate in order to prevent a single severe hospitalization.?As can seen from the table, in order to prevent one severe hospitalization, 7,500 people in the 70+ age group need to be injected. ??If you’re younger than that I’ve included the rest of the table in the image. Remember that the other report identified a serious adverse reaction in one in every 910 injections.
To conclude.?Be careful of making risky decisions by delegating the decisions to those who may be no better informed than you are.
UK HSA briefing: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1131409/appendix-1-of-jcvi-statement-on-2023-covid-19-vaccination-programme-8-november-2022.pdf
Done-For-You Client Acquisition Machine - We help ambitious Coaches and Consultants 3x their revenue with our DFY system.
2 年Gary, thanks for sharing!
Helping you sell more stuff to more people and businesses
2 年My dad took his beloved elderly cat to the vet recently, concerned she might have arthritis. Costly blood-tests ensued… nothing for creaky joints, but indicators that she *may* get renal failure. Solution - madly expensive cat food that she hates. He is miserable about it too, watching her resist eating until starving. He has decided that one year of happy cat eating is better than two years of unhappy cat hating food! I agree. Be careful what you wish for when asking for an opinion!
Principal at William Garrity Associates Ltd Management Consultant FinTech
2 年Human Factors Science focuses on human and errors extensively used in aviation - James Reason always a good source, author of the Swiss Cheese model.
I help people find more time to do more of the value add work the customer wants them to do
2 年So are you saying asking total strangers on Facebook isn't the best idea? ??