Please enable JavaScript.
Coggle requires JavaScript to display documents.
Public Trust in Statistics (Philosophy of trust (Who do we trust?, Why do…
Public Trust in Statistics
Trust in publishing institutions
Goverment agencies
Universities
Pirvate companies
Interpretations by the media
Misuse of statistics
Overgeneralization
Biased samples
Data manipulation
Axis manipulation
Cherry picking
Some fallacies and paradox
Prosecutor's fallacy
Gambler's fallacy
Simpson's paradox
Base rate fallacy
Multiple comparisons fallacy
Proof of null hypothesis
Faulty polling
e.g. in case of a poll seeking tax opinions ('Do you believe that you should be taxed so other citizens don't have to work?')
Interpretation of data
Mathematical literacy impacts trust on statistics/science
How does the public interpret statistics and what impacts their ability to interpret it?
Ethical issues surrounding class - lower class have worse access to good education so therefore less mathematical literacy. Are they less fairly equipt to make informed decisions regarding stuff like voting. Like lower class voting for Brexit not understanding that the economic repercussions would impact them the most.
Confirmation bias
Selection bias
Correlation =/=> causation
If you have enough data you can find correlation between anything
Ethics of lying
Is it ever ethical to lie?
Under what circumstances?
Utilitarian v Deotological arguments
Examples
Nixon
Censorship during war
Intent (does intent matter?)
i.e. misrepresenting climate change statistics to boost action
How do we judge intent?
Philosophy of trust
Who do we trust?
Why do we trust?
What is trust?
What happens when trust is abused?
How do we form trust as individuals and as a general public?
What affects our trust?
Under what circumstances do we accept statistical manipulation and to what extent? (Dispute distrust)
Case studies/questions
The failure of Long-Term Capital Management, a hedge fund managed by Nobel laureates that blindly followed flawed financial models, leading to a loss of several billion dollars
When the data comes from a trustworthy source, who do we trust with analysing it?
The red bus with the 350mil NHS lie during brexit, immigration statistics
conformation bias
statistics scandal that cost the UK billions (link to article in whatsapp)
Scientific perspectives
Generally, in the science community we want to represent the most accurate and truthful result.
What are common best practices for reporting and publishing stats?
How well are these implemented?
How effective are they?
History
What are the origins of public interactions with statistics?
What are the origins of public opinion in general?
Who are some important theorists regarding this topic?
social perspectives
Utilitarianism (that a course of action should be taken by considering the most positive outcome)
Deontology (judges whether an action is right or wrong based on a moral code)
Is it ever ethical to manipulate statistics? What impact does that have on trust?
To what extent should we trust government statistics?
stakeholders
General public
The public may blame greed causing companies to manipulate statistics leading to distrust
Companies
Companies need to turn a profit to survive and compete against a lot of competition they may argue that capitalism (government) is to blame and manipulation of statistics is necessary to stay afloat
Government
politicians/government want to maintain order and would justify data manipulation by blaming public for being unstable (needing control) or preventing an outbreak of war/financial instability (for the greater good)
solutions
To prevent manipulation of statistics (again leading to distrust of statistics in general) from companies, the government may suggest checking/supervising company produced statistics. But governments have their own agendas, so the public and companies may not agree with this solution.
To solve lack of mathematical literacy (leading to distrust in statistics) the general public may want equal access to education for all, but the government may say that is not feasible
(that stakeholders may suggest)