SlideShare ist ein Scribd-Unternehmen logo
1 von 27
Analyzing bias in data
Jonathan Stray
Columbia Journalism School
IRE 2019
Institute for the Future’s “unintended harms of technology”, ethicalos.org
Part I: Quantitative Fairness
What does this mean?
What would fair mean here?
Same number of white/minority drivers ticketed?
White/minority drivers ticketed in same ratio as local resident
demographics?
White/minority drivers ticketed in same ratio as local driver
demographics?
White/minority drivers ticketed for driving at the same speeds?
Legal concept: “similarly situated”
Similarly situated. Alike in all relevant ways for purposes of a particular
decision or issue. This term is often used in discrimination cases, in
which the plaintiff may seek to show that he or she was treated
differently from others who are similarly situated except for the alleged
basis of discrimination. For example, a plaintiff who claims that she was
not promoted because she is a woman would seek to show that similarly
situated men -- that is, men with similar qualifications, experience, and
tenure with the company -- were promoted.
Wex’s law dictionary, Legal Information Institute, Cornell
Florida sentencing analysis adjusted for “points”
Bias on the Bench, Michael Braga, Herald Tribune
Containing 1.4 million entries, the DOC database notes the exact number of points assigned to defendants
convicted of felonies. The points are based on the nature and severity of the crime committed, as well as
other factors such as past criminal history, use of a weapon and whether anyone got hurt. The more points a
defendant gets, the longer the minimum sentence required by law.
Florida legislators created the point system to ensure defendants committing the same crime are treated
equally by judges. But that is not what happens.


The Herald-Tribune established this by grouping defendants who committed the same crimes according to
the points they scored at sentencing. Anyone who scored from 30 to 30.9 would go into one group, while
anyone who scored from 31 to 31.9 would go in another, and so on.
We then evaluated how judges sentenced black and white defendants within each point range, assigning a
weighted average based on the sentencing gap.
If a judge wound up with a weighted average of 45 percent, it meant that judge sentenced black defendants
to 45 percent more time behind bars than white defendants.
Bias on the Bench: How We Did It, Michael Braga, Herald Tribune
For a brief period, Massachusetts recorded “warnings” as well as tickets,
allowing us to directly compare who got off easy and who didn’t.
Calibration
The idea: a prediction means the same thing for each group.
Same percentage of re-arrest among black and white defendants who were scored as high
risk. Same percentage of equally qualified men and women hired. Whether you will get a
loan depends only on your probability of repayment.
Mathematically:
Equal positive predictive value (“precision”) for each group.
A classifier with this property: most standard machine learning algorithms.
Drawbacks: Disparate impacts may exacerbate existing disparities. Error rates may differ
between groups in unfair ways.
Legal principle: similarly situated
Moral principle: equality of opportunity
Legal concept: “disparate impact”
D. Adverse impact and the "four-fifths rule."
A selection rate for any race, sex, or ethnic group which is less than four-
fifths (4/5) (or eighty percent) of the rate for the group with the highest
rate will generally be regarded by the Federal enforcement agencies as
evidence of adverse impact, while a greater than four-fifths rate will
generally not be regarded by Federal enforcement agencies as evidence
of adverse impact.
29 CFR § 1607.4
Uniform Guidelines on Employee Selection Procedures,
Information on impact
Demographic Parity
The idea: the prediction should not depend on the group.
Same percentage of black and white defendants scored as high risk. Same percentage of
men and women hired. Same percentage of rich and poor students admitted.
Mathematically:
Equal rate of true/false prediction for all groups.
A classifier with this property: choose the 10 best scoring applicants in each group.
Drawbacks: Doesn’t measure who we accept, as long as we accept equal numbers in
each group. The “perfect” predictor, which always guesses correctly, is considered unfair if
the base rates are different.
Legal principle: disparate impact
Moral principle: equality of outcome
ProPublica argument: fairness as error rates
Equal error rates
The idea: Don’t let a classifier make most of its mistakes on one group.
Same percentage of black and white defendants who are not re-arrested are scored as
high risk. Same percentage of qualified men and women mistakenly turned down. If you
would have repaid a loan, you will be turned down at the same rate regardless of your
income.
Mathematically:
Equal false positive rate, true positive rate between groups.
A classifier with this property: use different thresholds for each group.
Drawbacks: Classifier must use group membership explicitly. Calibration is not possible
(the same score will mean different things for different groups.)
Legal principle: disparate treatment
Moral principle: equality of opportunity
Part II: Fairness In the Real World
Image by Craig Froehle
With different base rates, calibration, demographic, and error rate fairness are
mutually exclusive.
This can be proved with a little arithmetic, but the intuition is:
- Can’t have demographic parity and calibration if different groups have
different qualifications.
- If risk really predicts outcome (calibration), then one group will have higher
risk scores, which means more positives and therefore more more false
positives.
Impossibility theorem
False Positive Rate can be gamed
ï»żA second misconception is that the false positive rate is a reasonable proxy of a
group’s aggregate well- being, loosely defined.


Suppose, hypothetically, that prosecutors start enforcing low-level drug crimes that
disproportionately involve black individuals, a policy that arguably hurts the black
community. Further suppose that the newly arrested individuals have low risk of
violent recidivism, and thus are released pending trial.


As a result, the false positive rate for blacks would decrease. To see this, recall
that the numerator of the false positive rate (the number of detained defendants
who do not reoffend) remains unchanged while the denominator (the number of
defendants who do not reoffend) increases.
Corbett-Davies and Goel, The Measure and Mismeasure of Fairness, 2018
Megan Stevenson, Assessing Risk Assessment in Action, 2018
Real-world results from Virginia
Algorithmic output may be ignored anyway
First, it is still unclear whether risk-assessment tools actually have a great
impact on the daily proceedings in courtrooms. During my days of
observation, I found that risk-assessment tools are often actively resisted in
criminal courts. Most judges and prosecutors do not trust the algorithms.
They do not know the companies they come from, they do not understand
their methods, and they often find them useless. Consequently, risk-
assessment tools often go unused: social workers complete the software
programs’ questionnaires, print out the score sheets, add them to the
defendants’ files
 after which the scores seem to disappear and are rarely
mentioned during hearings or plea bargaining negotiations.
AngĂšle Christin, The Mistrials of Algorithmic Sentencing
Challenges to determining fairness through data
Groups never differ by just race / gender / class alone.
There are several plausible definitions of “fair,” and they are both
controversial and mutually exclusive.
Every analysis method has potential false negatives and false
positives. Causality is a particular problem.
Humans may follow or ignore algorithmic recommendations
Part III: Reframing the Problem
When considering an algorithmic system, what do you compare it to?
Absolute fairness – We don’t have perfect prediction or perfect data, and there
may not be agreement over which definition of fairness to use.
As fair as possible given the data – It may be possible to achieve this, given a
particular definition of fairness, if we understand very well what the limitations of
the input data are.
An improvement over current processes and human decision-makers – It’s
possible to evaluate existing institutions by the same standards as algorithms,
and the results do not always favor humans.
An improvement over other possible reforms – If the humans are biased and
the algorithms are biased, is there some other approach?
Fairness by Comparison
Sandra Mayson, Bias In, Bias Out
Prediction in an Unequal World
(2) AUTOMATED DECISION SYSTEM IMPACT ASSESSMENT.
The term ‘‘automated decision system impact assessment’’ means a study
evaluating an automated decision system and the automated decision system’s
development process, including the design and training data of the automated
decision system, for impacts on accuracy, fairness, bias, discrimination, privacy,
and security
Algorithmic Accountability Act of 2019 (proposed)
Proposed algorithmic fairness legislation
doesn’t define “fairness”
Bias In, Bias Out, Sandra Mayson
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3257004
Assessing Risk Assessment in Action, Megan Stevenson
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3016088
Open Policing Project – Findings
https://openpolicing.stanford.edu/findings/
Open Policing Project – Workbench Tutorial
https://app.workbenchdata.com/workflows/18232/
21 Definitions of Fairness and Their Politics, Arvind Narayanan
https://www.youtube.com/watch?v=jIXIuYdnyyk
Resources

Weitere Àhnliche Inhalte

Was ist angesagt?

Issues in Policing SAR 1
Issues in Policing SAR 1Issues in Policing SAR 1
Issues in Policing SAR 1Marcos Corley
 
Rational choice theory
Rational choice theoryRational choice theory
Rational choice theoryShaista Mariam
 
81-220-1 - Chapter4
81-220-1 - Chapter481-220-1 - Chapter4
81-220-1 - Chapter4mpalaro
 
Routine activity & rational choice theory final project
Routine activity & rational choice theory final projectRoutine activity & rational choice theory final project
Routine activity & rational choice theory final projectLisa Shelby
 
Issues Midterm
Issues MidtermIssues Midterm
Issues MidtermMarcos Corley
 
SociologyExchange.co.uk Shared Resource
SociologyExchange.co.uk Shared ResourceSociologyExchange.co.uk Shared Resource
SociologyExchange.co.uk Shared Resourcesociologyexchange.co.uk
 
Chapter 1
Chapter 1Chapter 1
Chapter 1glickauf
 
Issues in Policing SAR 2
Issues in Policing SAR 2Issues in Policing SAR 2
Issues in Policing SAR 2Marcos Corley
 
Chapter 5
Chapter 5Chapter 5
Chapter 5glickauf
 
01 basic concepts
01 basic concepts01 basic concepts
01 basic conceptsJim Gilmer
 

Was ist angesagt? (14)

Issues in Policing SAR 1
Issues in Policing SAR 1Issues in Policing SAR 1
Issues in Policing SAR 1
 
Capstone paper
Capstone paperCapstone paper
Capstone paper
 
Rational choice theory
Rational choice theoryRational choice theory
Rational choice theory
 
Issues SAR 3
Issues SAR 3Issues SAR 3
Issues SAR 3
 
81-220-1 - Chapter4
81-220-1 - Chapter481-220-1 - Chapter4
81-220-1 - Chapter4
 
Routine activity & rational choice theory final project
Routine activity & rational choice theory final projectRoutine activity & rational choice theory final project
Routine activity & rational choice theory final project
 
Issues Midterm
Issues MidtermIssues Midterm
Issues Midterm
 
SociologyExchange.co.uk Shared Resource
SociologyExchange.co.uk Shared ResourceSociologyExchange.co.uk Shared Resource
SociologyExchange.co.uk Shared Resource
 
Chapter 1
Chapter 1Chapter 1
Chapter 1
 
Issues in Policing SAR 2
Issues in Policing SAR 2Issues in Policing SAR 2
Issues in Policing SAR 2
 
Chapter 5
Chapter 5Chapter 5
Chapter 5
 
01 basic concepts
01 basic concepts01 basic concepts
01 basic concepts
 
Issues Final
Issues FinalIssues Final
Issues Final
 
Discrimination Discovery
Discrimination DiscoveryDiscrimination Discovery
Discrimination Discovery
 

Ähnlich wie Analyzing Bias in Data - IRE 2019

Frameworks for Algorithmic Bias
Frameworks for Algorithmic BiasFrameworks for Algorithmic Bias
Frameworks for Algorithmic BiasJonathan Stray
 
Frontiers of Computational Journalism week 6 - Quantitative Fairness
Frontiers of Computational Journalism week 6 - Quantitative FairnessFrontiers of Computational Journalism week 6 - Quantitative Fairness
Frontiers of Computational Journalism week 6 - Quantitative FairnessJonathan Stray
 
algorithmic-bias.pptx
algorithmic-bias.pptxalgorithmic-bias.pptx
algorithmic-bias.pptxTewodrosEshete1
 
Great Writing 3 From Great Paragraphs To Great Essay
Great Writing 3 From Great Paragraphs To Great EssayGreat Writing 3 From Great Paragraphs To Great Essay
Great Writing 3 From Great Paragraphs To Great EssayTara Smith
 
300 words agree or disagree to each questionsQ1.My working.docx
300 words agree or disagree to each questionsQ1.My working.docx300 words agree or disagree to each questionsQ1.My working.docx
300 words agree or disagree to each questionsQ1.My working.docxdomenicacullison
 
Telephone Essay In Tamil. Online assignment writing service.
Telephone Essay In Tamil. Online assignment writing service.Telephone Essay In Tamil. Online assignment writing service.
Telephone Essay In Tamil. Online assignment writing service.Jill Johnson
 
In this module, you will learn about the controversies surrounding.docx
In this module, you will learn about the controversies surrounding.docxIn this module, you will learn about the controversies surrounding.docx
In this module, you will learn about the controversies surrounding.docxjaggernaoma
 
Argumentative Essay Outline. Online assignment writing service.
Argumentative Essay Outline. Online assignment writing service.Argumentative Essay Outline. Online assignment writing service.
Argumentative Essay Outline. Online assignment writing service.Jeanne Hall
 
Measuring Human Perception to Defend Democracy
Measuring Human Perceptionto Defend DemocracyMeasuring Human Perceptionto Defend Democracy
Measuring Human Perception to Defend DemocracyElissa Redmiles
 
Pick one aspect of the criminal justice system that was discussed in.docx
Pick one aspect of the criminal justice system that was discussed in.docxPick one aspect of the criminal justice system that was discussed in.docx
Pick one aspect of the criminal justice system that was discussed in.docxJUST36
 
Choose Essay Writing Help Wisely To Succeed At The F
Choose Essay Writing Help Wisely To Succeed At The FChoose Essay Writing Help Wisely To Succeed At The F
Choose Essay Writing Help Wisely To Succeed At The FBrenda Thomas
 
Nothing Works; Disproportionate Minority Confinement
Nothing Works; Disproportionate Minority ConfinementNothing Works; Disproportionate Minority Confinement
Nothing Works; Disproportionate Minority ConfinementClyde Knight Jr. Criminologist
 
Consequentialism Theory Essay
Consequentialism Theory EssayConsequentialism Theory Essay
Consequentialism Theory EssayNina Vazquez
 

Ähnlich wie Analyzing Bias in Data - IRE 2019 (15)

Frameworks for Algorithmic Bias
Frameworks for Algorithmic BiasFrameworks for Algorithmic Bias
Frameworks for Algorithmic Bias
 
Frontiers of Computational Journalism week 6 - Quantitative Fairness
Frontiers of Computational Journalism week 6 - Quantitative FairnessFrontiers of Computational Journalism week 6 - Quantitative Fairness
Frontiers of Computational Journalism week 6 - Quantitative Fairness
 
algorithmic-bias.pptx
algorithmic-bias.pptxalgorithmic-bias.pptx
algorithmic-bias.pptx
 
Great Writing 3 From Great Paragraphs To Great Essay
Great Writing 3 From Great Paragraphs To Great EssayGreat Writing 3 From Great Paragraphs To Great Essay
Great Writing 3 From Great Paragraphs To Great Essay
 
300 words agree or disagree to each questionsQ1.My working.docx
300 words agree or disagree to each questionsQ1.My working.docx300 words agree or disagree to each questionsQ1.My working.docx
300 words agree or disagree to each questionsQ1.My working.docx
 
Telephone Essay In Tamil. Online assignment writing service.
Telephone Essay In Tamil. Online assignment writing service.Telephone Essay In Tamil. Online assignment writing service.
Telephone Essay In Tamil. Online assignment writing service.
 
In this module, you will learn about the controversies surrounding.docx
In this module, you will learn about the controversies surrounding.docxIn this module, you will learn about the controversies surrounding.docx
In this module, you will learn about the controversies surrounding.docx
 
Argumentative Essay Outline. Online assignment writing service.
Argumentative Essay Outline. Online assignment writing service.Argumentative Essay Outline. Online assignment writing service.
Argumentative Essay Outline. Online assignment writing service.
 
Measuring Human Perception to Defend Democracy
Measuring Human Perceptionto Defend DemocracyMeasuring Human Perceptionto Defend Democracy
Measuring Human Perception to Defend Democracy
 
Pick one aspect of the criminal justice system that was discussed in.docx
Pick one aspect of the criminal justice system that was discussed in.docxPick one aspect of the criminal justice system that was discussed in.docx
Pick one aspect of the criminal justice system that was discussed in.docx
 
Chapter 9
Chapter 9Chapter 9
Chapter 9
 
Choose Essay Writing Help Wisely To Succeed At The F
Choose Essay Writing Help Wisely To Succeed At The FChoose Essay Writing Help Wisely To Succeed At The F
Choose Essay Writing Help Wisely To Succeed At The F
 
Nothing Works; Disproportionate Minority Confinement
Nothing Works; Disproportionate Minority ConfinementNothing Works; Disproportionate Minority Confinement
Nothing Works; Disproportionate Minority Confinement
 
Consequentialism Theory Essay
Consequentialism Theory EssayConsequentialism Theory Essay
Consequentialism Theory Essay
 
the paper
the paperthe paper
the paper
 

Mehr von Jonathan Stray

Frontiers of Computational Journalism week 11 - Privacy and Security
Frontiers of Computational Journalism week 11 - Privacy and SecurityFrontiers of Computational Journalism week 11 - Privacy and Security
Frontiers of Computational Journalism week 11 - Privacy and SecurityJonathan Stray
 
Frontiers of Computational Journalism week 10 - Truth and Trust
Frontiers of Computational Journalism week 10 - Truth and TrustFrontiers of Computational Journalism week 10 - Truth and Trust
Frontiers of Computational Journalism week 10 - Truth and TrustJonathan Stray
 
Frontiers of Computational Journalism week 9 - Knowledge representation
Frontiers of Computational Journalism week 9 - Knowledge representationFrontiers of Computational Journalism week 9 - Knowledge representation
Frontiers of Computational Journalism week 9 - Knowledge representationJonathan Stray
 
Frontiers of Computational Journalism week 8 - Visualization and Network Anal...
Frontiers of Computational Journalism week 8 - Visualization and Network Anal...Frontiers of Computational Journalism week 8 - Visualization and Network Anal...
Frontiers of Computational Journalism week 8 - Visualization and Network Anal...Jonathan Stray
 
Frontiers of Computational Journalism week 5 - Algorithmic Accountability and...
Frontiers of Computational Journalism week 5 - Algorithmic Accountability and...Frontiers of Computational Journalism week 5 - Algorithmic Accountability and...
Frontiers of Computational Journalism week 5 - Algorithmic Accountability and...Jonathan Stray
 
Frontiers of Computational Journalism - Final project suggestions
Frontiers of Computational Journalism - Final project suggestionsFrontiers of Computational Journalism - Final project suggestions
Frontiers of Computational Journalism - Final project suggestionsJonathan Stray
 
Frontiers of Computational Journalism week 4 - Statistical Inference
Frontiers of Computational Journalism week 4 - Statistical InferenceFrontiers of Computational Journalism week 4 - Statistical Inference
Frontiers of Computational Journalism week 4 - Statistical InferenceJonathan Stray
 
Frontiers of Computational Journalism week 3 - Information Filter Design
Frontiers of Computational Journalism week 3 - Information Filter DesignFrontiers of Computational Journalism week 3 - Information Filter Design
Frontiers of Computational Journalism week 3 - Information Filter DesignJonathan Stray
 
Frontiers of Computational Journalism week 2 - Text Analysis
Frontiers of Computational Journalism week 2 - Text AnalysisFrontiers of Computational Journalism week 2 - Text Analysis
Frontiers of Computational Journalism week 2 - Text AnalysisJonathan Stray
 
Frontiers of Computational Journalism week 1 - Introduction and High Dimensio...
Frontiers of Computational Journalism week 1 - Introduction and High Dimensio...Frontiers of Computational Journalism week 1 - Introduction and High Dimensio...
Frontiers of Computational Journalism week 1 - Introduction and High Dimensio...Jonathan Stray
 

Mehr von Jonathan Stray (10)

Frontiers of Computational Journalism week 11 - Privacy and Security
Frontiers of Computational Journalism week 11 - Privacy and SecurityFrontiers of Computational Journalism week 11 - Privacy and Security
Frontiers of Computational Journalism week 11 - Privacy and Security
 
Frontiers of Computational Journalism week 10 - Truth and Trust
Frontiers of Computational Journalism week 10 - Truth and TrustFrontiers of Computational Journalism week 10 - Truth and Trust
Frontiers of Computational Journalism week 10 - Truth and Trust
 
Frontiers of Computational Journalism week 9 - Knowledge representation
Frontiers of Computational Journalism week 9 - Knowledge representationFrontiers of Computational Journalism week 9 - Knowledge representation
Frontiers of Computational Journalism week 9 - Knowledge representation
 
Frontiers of Computational Journalism week 8 - Visualization and Network Anal...
Frontiers of Computational Journalism week 8 - Visualization and Network Anal...Frontiers of Computational Journalism week 8 - Visualization and Network Anal...
Frontiers of Computational Journalism week 8 - Visualization and Network Anal...
 
Frontiers of Computational Journalism week 5 - Algorithmic Accountability and...
Frontiers of Computational Journalism week 5 - Algorithmic Accountability and...Frontiers of Computational Journalism week 5 - Algorithmic Accountability and...
Frontiers of Computational Journalism week 5 - Algorithmic Accountability and...
 
Frontiers of Computational Journalism - Final project suggestions
Frontiers of Computational Journalism - Final project suggestionsFrontiers of Computational Journalism - Final project suggestions
Frontiers of Computational Journalism - Final project suggestions
 
Frontiers of Computational Journalism week 4 - Statistical Inference
Frontiers of Computational Journalism week 4 - Statistical InferenceFrontiers of Computational Journalism week 4 - Statistical Inference
Frontiers of Computational Journalism week 4 - Statistical Inference
 
Frontiers of Computational Journalism week 3 - Information Filter Design
Frontiers of Computational Journalism week 3 - Information Filter DesignFrontiers of Computational Journalism week 3 - Information Filter Design
Frontiers of Computational Journalism week 3 - Information Filter Design
 
Frontiers of Computational Journalism week 2 - Text Analysis
Frontiers of Computational Journalism week 2 - Text AnalysisFrontiers of Computational Journalism week 2 - Text Analysis
Frontiers of Computational Journalism week 2 - Text Analysis
 
Frontiers of Computational Journalism week 1 - Introduction and High Dimensio...
Frontiers of Computational Journalism week 1 - Introduction and High Dimensio...Frontiers of Computational Journalism week 1 - Introduction and High Dimensio...
Frontiers of Computational Journalism week 1 - Introduction and High Dimensio...
 

KĂŒrzlich hochgeladen

JAPAN: ORGANISATION OF PMDA, PHARMACEUTICAL LAWS & REGULATIONS, TYPES OF REGI...
JAPAN: ORGANISATION OF PMDA, PHARMACEUTICAL LAWS & REGULATIONS, TYPES OF REGI...JAPAN: ORGANISATION OF PMDA, PHARMACEUTICAL LAWS & REGULATIONS, TYPES OF REGI...
JAPAN: ORGANISATION OF PMDA, PHARMACEUTICAL LAWS & REGULATIONS, TYPES OF REGI...anjaliyadav012327
 
Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3JemimahLaneBuaron
 
Advance Mobile Application Development class 07
Advance Mobile Application Development class 07Advance Mobile Application Development class 07
Advance Mobile Application Development class 07Dr. Mazin Mohamed alkathiri
 
Disha NEET Physics Guide for classes 11 and 12.pdf
Disha NEET Physics Guide for classes 11 and 12.pdfDisha NEET Physics Guide for classes 11 and 12.pdf
Disha NEET Physics Guide for classes 11 and 12.pdfchloefrazer622
 
1029-Danh muc Sach Giao Khoa khoi 6.pdf
1029-Danh muc Sach Giao Khoa khoi  6.pdf1029-Danh muc Sach Giao Khoa khoi  6.pdf
1029-Danh muc Sach Giao Khoa khoi 6.pdfQucHHunhnh
 
CARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxCARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxGaneshChakor2
 
APM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAPM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAssociation for Project Management
 
Student login on Anyboli platform.helpin
Student login on Anyboli platform.helpinStudent login on Anyboli platform.helpin
Student login on Anyboli platform.helpinRaunakKeshri1
 
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...fonyou31
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdfSoniaTolstoy
 
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...Krashi Coaching
 
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...Sapna Thakur
 
Mastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionMastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionSafetyChain Software
 
Arihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfArihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfchloefrazer622
 
Russian Call Girls in Andheri Airport Mumbai WhatsApp 9167673311 💞 Full Nigh...
Russian Call Girls in Andheri Airport Mumbai WhatsApp  9167673311 💞 Full Nigh...Russian Call Girls in Andheri Airport Mumbai WhatsApp  9167673311 💞 Full Nigh...
Russian Call Girls in Andheri Airport Mumbai WhatsApp 9167673311 💞 Full Nigh...Pooja Nehwal
 
The byproduct of sericulture in different industries.pptx
The byproduct of sericulture in different industries.pptxThe byproduct of sericulture in different industries.pptx
The byproduct of sericulture in different industries.pptxShobhayan Kirtania
 
Web & Social Media Analytics Previous Year Question Paper.pdf
Web & Social Media Analytics Previous Year Question Paper.pdfWeb & Social Media Analytics Previous Year Question Paper.pdf
Web & Social Media Analytics Previous Year Question Paper.pdfJayanti Pande
 
Activity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdfActivity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdfciinovamais
 
The basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptxThe basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptxheathfieldcps1
 

KĂŒrzlich hochgeladen (20)

JAPAN: ORGANISATION OF PMDA, PHARMACEUTICAL LAWS & REGULATIONS, TYPES OF REGI...
JAPAN: ORGANISATION OF PMDA, PHARMACEUTICAL LAWS & REGULATIONS, TYPES OF REGI...JAPAN: ORGANISATION OF PMDA, PHARMACEUTICAL LAWS & REGULATIONS, TYPES OF REGI...
JAPAN: ORGANISATION OF PMDA, PHARMACEUTICAL LAWS & REGULATIONS, TYPES OF REGI...
 
Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3
 
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptxINDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
 
Advance Mobile Application Development class 07
Advance Mobile Application Development class 07Advance Mobile Application Development class 07
Advance Mobile Application Development class 07
 
Disha NEET Physics Guide for classes 11 and 12.pdf
Disha NEET Physics Guide for classes 11 and 12.pdfDisha NEET Physics Guide for classes 11 and 12.pdf
Disha NEET Physics Guide for classes 11 and 12.pdf
 
1029-Danh muc Sach Giao Khoa khoi 6.pdf
1029-Danh muc Sach Giao Khoa khoi  6.pdf1029-Danh muc Sach Giao Khoa khoi  6.pdf
1029-Danh muc Sach Giao Khoa khoi 6.pdf
 
CARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxCARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptx
 
APM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAPM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across Sectors
 
Student login on Anyboli platform.helpin
Student login on Anyboli platform.helpinStudent login on Anyboli platform.helpin
Student login on Anyboli platform.helpin
 
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
 
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
 
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
 
Mastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionMastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory Inspection
 
Arihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfArihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdf
 
Russian Call Girls in Andheri Airport Mumbai WhatsApp 9167673311 💞 Full Nigh...
Russian Call Girls in Andheri Airport Mumbai WhatsApp  9167673311 💞 Full Nigh...Russian Call Girls in Andheri Airport Mumbai WhatsApp  9167673311 💞 Full Nigh...
Russian Call Girls in Andheri Airport Mumbai WhatsApp 9167673311 💞 Full Nigh...
 
The byproduct of sericulture in different industries.pptx
The byproduct of sericulture in different industries.pptxThe byproduct of sericulture in different industries.pptx
The byproduct of sericulture in different industries.pptx
 
Web & Social Media Analytics Previous Year Question Paper.pdf
Web & Social Media Analytics Previous Year Question Paper.pdfWeb & Social Media Analytics Previous Year Question Paper.pdf
Web & Social Media Analytics Previous Year Question Paper.pdf
 
Activity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdfActivity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdf
 
The basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptxThe basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptx
 

Analyzing Bias in Data - IRE 2019

  • 1. Analyzing bias in data Jonathan Stray Columbia Journalism School IRE 2019
  • 2. Institute for the Future’s “unintended harms of technology”, ethicalos.org
  • 5. What would fair mean here? Same number of white/minority drivers ticketed? White/minority drivers ticketed in same ratio as local resident demographics? White/minority drivers ticketed in same ratio as local driver demographics? White/minority drivers ticketed for driving at the same speeds?
  • 6. Legal concept: “similarly situated” Similarly situated. Alike in all relevant ways for purposes of a particular decision or issue. This term is often used in discrimination cases, in which the plaintiff may seek to show that he or she was treated differently from others who are similarly situated except for the alleged basis of discrimination. For example, a plaintiff who claims that she was not promoted because she is a woman would seek to show that similarly situated men -- that is, men with similar qualifications, experience, and tenure with the company -- were promoted. Wex’s law dictionary, Legal Information Institute, Cornell
  • 7. Florida sentencing analysis adjusted for “points” Bias on the Bench, Michael Braga, Herald Tribune
  • 8. Containing 1.4 million entries, the DOC database notes the exact number of points assigned to defendants convicted of felonies. The points are based on the nature and severity of the crime committed, as well as other factors such as past criminal history, use of a weapon and whether anyone got hurt. The more points a defendant gets, the longer the minimum sentence required by law. Florida legislators created the point system to ensure defendants committing the same crime are treated equally by judges. But that is not what happens. 
 The Herald-Tribune established this by grouping defendants who committed the same crimes according to the points they scored at sentencing. Anyone who scored from 30 to 30.9 would go into one group, while anyone who scored from 31 to 31.9 would go in another, and so on. We then evaluated how judges sentenced black and white defendants within each point range, assigning a weighted average based on the sentencing gap. If a judge wound up with a weighted average of 45 percent, it meant that judge sentenced black defendants to 45 percent more time behind bars than white defendants. Bias on the Bench: How We Did It, Michael Braga, Herald Tribune
  • 9. For a brief period, Massachusetts recorded “warnings” as well as tickets, allowing us to directly compare who got off easy and who didn’t.
  • 10. Calibration The idea: a prediction means the same thing for each group. Same percentage of re-arrest among black and white defendants who were scored as high risk. Same percentage of equally qualified men and women hired. Whether you will get a loan depends only on your probability of repayment. Mathematically: Equal positive predictive value (“precision”) for each group. A classifier with this property: most standard machine learning algorithms. Drawbacks: Disparate impacts may exacerbate existing disparities. Error rates may differ between groups in unfair ways. Legal principle: similarly situated Moral principle: equality of opportunity
  • 11.
  • 12. Legal concept: “disparate impact” D. Adverse impact and the "four-fifths rule." A selection rate for any race, sex, or ethnic group which is less than four- fifths (4/5) (or eighty percent) of the rate for the group with the highest rate will generally be regarded by the Federal enforcement agencies as evidence of adverse impact, while a greater than four-fifths rate will generally not be regarded by Federal enforcement agencies as evidence of adverse impact. 29 CFR § 1607.4 Uniform Guidelines on Employee Selection Procedures, Information on impact
  • 13. Demographic Parity The idea: the prediction should not depend on the group. Same percentage of black and white defendants scored as high risk. Same percentage of men and women hired. Same percentage of rich and poor students admitted. Mathematically: Equal rate of true/false prediction for all groups. A classifier with this property: choose the 10 best scoring applicants in each group. Drawbacks: Doesn’t measure who we accept, as long as we accept equal numbers in each group. The “perfect” predictor, which always guesses correctly, is considered unfair if the base rates are different. Legal principle: disparate impact Moral principle: equality of outcome
  • 15. Equal error rates The idea: Don’t let a classifier make most of its mistakes on one group. Same percentage of black and white defendants who are not re-arrested are scored as high risk. Same percentage of qualified men and women mistakenly turned down. If you would have repaid a loan, you will be turned down at the same rate regardless of your income. Mathematically: Equal false positive rate, true positive rate between groups. A classifier with this property: use different thresholds for each group. Drawbacks: Classifier must use group membership explicitly. Calibration is not possible (the same score will mean different things for different groups.) Legal principle: disparate treatment Moral principle: equality of opportunity
  • 16. Part II: Fairness In the Real World
  • 17. Image by Craig Froehle
  • 18. With different base rates, calibration, demographic, and error rate fairness are mutually exclusive. This can be proved with a little arithmetic, but the intuition is: - Can’t have demographic parity and calibration if different groups have different qualifications. - If risk really predicts outcome (calibration), then one group will have higher risk scores, which means more positives and therefore more more false positives. Impossibility theorem
  • 19. False Positive Rate can be gamed ï»żA second misconception is that the false positive rate is a reasonable proxy of a group’s aggregate well- being, loosely defined. 
 Suppose, hypothetically, that prosecutors start enforcing low-level drug crimes that disproportionately involve black individuals, a policy that arguably hurts the black community. Further suppose that the newly arrested individuals have low risk of violent recidivism, and thus are released pending trial. 
 As a result, the false positive rate for blacks would decrease. To see this, recall that the numerator of the false positive rate (the number of detained defendants who do not reoffend) remains unchanged while the denominator (the number of defendants who do not reoffend) increases. Corbett-Davies and Goel, The Measure and Mismeasure of Fairness, 2018
  • 20. Megan Stevenson, Assessing Risk Assessment in Action, 2018 Real-world results from Virginia
  • 21. Algorithmic output may be ignored anyway First, it is still unclear whether risk-assessment tools actually have a great impact on the daily proceedings in courtrooms. During my days of observation, I found that risk-assessment tools are often actively resisted in criminal courts. Most judges and prosecutors do not trust the algorithms. They do not know the companies they come from, they do not understand their methods, and they often find them useless. Consequently, risk- assessment tools often go unused: social workers complete the software programs’ questionnaires, print out the score sheets, add them to the defendants’ files
 after which the scores seem to disappear and are rarely mentioned during hearings or plea bargaining negotiations. AngĂšle Christin, The Mistrials of Algorithmic Sentencing
  • 22. Challenges to determining fairness through data Groups never differ by just race / gender / class alone. There are several plausible definitions of “fair,” and they are both controversial and mutually exclusive. Every analysis method has potential false negatives and false positives. Causality is a particular problem. Humans may follow or ignore algorithmic recommendations
  • 23. Part III: Reframing the Problem
  • 24. When considering an algorithmic system, what do you compare it to? Absolute fairness – We don’t have perfect prediction or perfect data, and there may not be agreement over which definition of fairness to use. As fair as possible given the data – It may be possible to achieve this, given a particular definition of fairness, if we understand very well what the limitations of the input data are. An improvement over current processes and human decision-makers – It’s possible to evaluate existing institutions by the same standards as algorithms, and the results do not always favor humans. An improvement over other possible reforms – If the humans are biased and the algorithms are biased, is there some other approach? Fairness by Comparison
  • 25. Sandra Mayson, Bias In, Bias Out Prediction in an Unequal World
  • 26. (2) AUTOMATED DECISION SYSTEM IMPACT ASSESSMENT. The term ‘‘automated decision system impact assessment’’ means a study evaluating an automated decision system and the automated decision system’s development process, including the design and training data of the automated decision system, for impacts on accuracy, fairness, bias, discrimination, privacy, and security Algorithmic Accountability Act of 2019 (proposed) Proposed algorithmic fairness legislation doesn’t define “fairness”
  • 27. Bias In, Bias Out, Sandra Mayson https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3257004 Assessing Risk Assessment in Action, Megan Stevenson https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3016088 Open Policing Project – Findings https://openpolicing.stanford.edu/findings/ Open Policing Project – Workbench Tutorial https://app.workbenchdata.com/workflows/18232/ 21 Definitions of Fairness and Their Politics, Arvind Narayanan https://www.youtube.com/watch?v=jIXIuYdnyyk Resources