comment: on making statistics valuable to the public

16
Comment: On Making Comment: On Making Statistics Valuable Statistics Valuable to the Public to the Public Theodore M. Porter Theodore M. Porter Professor, History of Science Professor, History of Science Department of History, UCLA Department of History, UCLA [email protected] [email protected]

Upload: lester-hunt

Post on 30-Dec-2015

24 views

Category:

Documents


0 download

DESCRIPTION

Comment: On Making Statistics Valuable to the Public. Theodore M. Porter Professor, History of Science Department of History, UCLA [email protected]. The birth of statistics. Numerical records are as old as writing. Censuses and government accounts go back to ancient times. - PowerPoint PPT Presentation

TRANSCRIPT

Comment: On Making Comment: On Making Statistics Valuable to the Statistics Valuable to the

PublicPublic

Theodore M. PorterTheodore M. PorterProfessor, History of ScienceProfessor, History of ScienceDepartment of History, UCLADepartment of History, UCLA

[email protected]@history.ucla.edu

The birth of statisticsThe birth of statistics

Numerical records are as old as Numerical records are as old as writing. Censuses and government writing. Censuses and government accounts go back to ancient times.accounts go back to ancient times.

The origin of statistics, around 1800, The origin of statistics, around 1800, came with ambitions to make a came with ambitions to make a science of the state, of statecraft…science of the state, of statecraft…

… … And with an ideal of public And with an ideal of public knowledge. Census records were no knowledge. Census records were no longer secrets of state.longer secrets of state.

The science of statisticsThe science of statistics

Statistics was at first a social science Statistics was at first a social science and not a mathematical one.and not a mathematical one.

It was championed by people who It was championed by people who hoped the numbers could speak for hoped the numbers could speak for themselves….themselves….

… … And would quickly identify the And would quickly identify the causes of social problems such as causes of social problems such as crime, suicide, and epidemic disease. crime, suicide, and epidemic disease.

PierreCharlesDupin’scartogramof illiteracy(1819)

Thanks to Michael Friendly: www.math.yorku.ca/SCS/gallery/

Guerry andBalbi use statisticalmaps to tryto establishconnectionbetweeninstructionand crime(1829)

Ambitions for statisticsAmbitions for statistics

Adolphe Quetelet (1796-1874)wanted statistics to be an experimental science of legislation.

The power of statisticsThe power of statistics

““A well made statistic is an impassible A well made statistic is an impassible testimony, above intimidation and testimony, above intimidation and seduction alike” (Michel Chevalier, seduction alike” (Michel Chevalier, 1860)1860)

“…“…Wherever the struggle resurfaces Wherever the struggle resurfaces between the champions of the general between the champions of the general interest and that of private interest, interest and that of private interest, you will find (statisticians) at our post, you will find (statisticians) at our post, armed and ready to march.” (Alfred armed and ready to march.” (Alfred de Foville, 1892)de Foville, 1892)

How powerful are statistics really?How powerful are statistics really?

But (Foville added) the governments But (Foville added) the governments never listen …never listen …

And it seemed that the message of And it seemed that the message of statistics was often not so statistics was often not so straightforward as statisticians had straightforward as statisticians had hoped. The charts showed no simple hoped. The charts showed no simple relationship between instruction and relationship between instruction and crime, for example.crime, for example.

Statistics and InformationStatistics and Information

The question of public knowledge of The question of public knowledge of statistics presumes that statistics are a statistics presumes that statistics are a form of information.form of information.

The idea of information, argues Yaron The idea of information, argues Yaron Ezrahi, usually assumes that the facts Ezrahi, usually assumes that the facts (numbers) are self-explanatory, readily (numbers) are self-explanatory, readily available for direct use by anyone.available for direct use by anyone.

Are statistics really transparent? Are their Are statistics really transparent? Are their meanings superficially evident?meanings superficially evident?

The Public and ScienceThe Public and Science

From “public understanding of From “public understanding of science” to “public engagement with science” to “public engagement with science.”science.”

A scientific spirit may be more A scientific spirit may be more important than knowledge of important than knowledge of particular facts.particular facts.

Often (notably in the US) members of Often (notably in the US) members of the public reject what they know to the public reject what they know to be accepted science (e.g. evolution).be accepted science (e.g. evolution).

Public knowledge of statisticsPublic knowledge of statistics

One of the papers regrets that the One of the papers regrets that the public is ignorant of basic numbers public is ignorant of basic numbers (population of Italy, euro/dollar (population of Italy, euro/dollar exchange rate, trends of CO2 exchange rate, trends of CO2 emissions, etc.)emissions, etc.)

The other takes the economists’ view The other takes the economists’ view that people will know what’s in their that people will know what’s in their interest to know, based on costs and interest to know, based on costs and benefits to them.benefits to them.

What good is knowledge of What good is knowledge of statistics?statistics?

It may be less a matter of private It may be less a matter of private advantage than a public good.advantage than a public good.

We might hope citizens will know We might hope citizens will know more than private interest dictates.more than private interest dictates.

What numbers are important for the What numbers are important for the public? How do they (we) determine public? How do they (we) determine which numbers are important?which numbers are important?

Is it good that they think of numbers Is it good that they think of numbers as straightforward “information”?as straightforward “information”?

A little knowledge could be a A little knowledge could be a dangerous thingdangerous thing

Numbers circulate widely during Numbers circulate widely during elections. They also are manipulated elections. They also are manipulated in various ways.in various ways.

Hypothetically, a president or prime Hypothetically, a president or prime minister might take measures to minister might take measures to improve some number in the short improve some number in the short term (just before an election) term (just before an election) knowing quite well that this will knowing quite well that this will create problems in the longer term.create problems in the longer term.

Statistics that matterStatistics that matter

The “misery The “misery index” became index” became famous in the famous in the Reagan-Bush1 Reagan-Bush1 years. This table years. This table is meant to is meant to defend Bush2. defend Bush2. Should it be Should it be taken seriously?taken seriously?

It’s good if the public knows more It’s good if the public knows more about statisticsabout statistics

But it may be just as But it may be just as important for them to important for them to learn to take numbers learn to take numbers with a grain of salt…with a grain of salt…

……And to know as And to know as much as possible much as possible about what the about what the statisics mean.statisics mean.

Thank youThank you