Ethics in Statistics : Opportunities and Challenges / edited by Hassan Doosti.

Format:

Book

Contributor:

Doosti, Hassan, editor.

Language:

English

Subjects (All):

Data mining.

Physical Description:

1 online resource (597 pages)

Edition:

First edition.

Place of Publication:

London, England : Ethics International Press Ltd, [2024]

Summary:

Data plays a vital role in different parts of our lives. In the world of big data, and policy determined by a variety of statistical artifacts, discussions around the ethics of data gathering, manipulation and presentation are increasingly important. Ethics in Statistics aims to make a significant contribution to that debate. The processes of gathering data through sampling, summarising of the findings, and extending results to a population, need to be checked via an ethical prospective, as well as a statistical one. Statistical learning without ethics can be harmful for mankind.This edited collection brings together contributors in the field of data science, data analytics and statistics, to share their thoughts about the role of ethics in different aspects of statistical learning.

Contents:

Intro

Preface

Setting the Tone:

Expressing Gratitude:

Explaining the Motivation:

Insights into the Editing Process:

Chapter 1 Ethics in Statistics: Opportunities and Challenges

Dr. Andrea Vicini0F(

Introduction

Privacy, profiling and personal data: where is the limit?

Methods and transparency: why is adequate attention not paid to the methodology adopted in preliminary foresight studies?

Governance of statistical institutions: what is the role of statistical institutions in society, and how must their relationship with the government be regulated in a democracy?

References

Chapter 2 Learning Outcomes Supporting the Integration of Ethical Reasoning into Quantitative Courses: Three Tasks for Use in Three General Contexts

Rochelle E. Tractenberg1F(

Abstract

Getting "ethics" into quantitative courses: neither simple nor straightforward

Background: Education Sciences

What is a learning outcome (LO)?

Role of LOs in curriculum and instructional design

Formulating LOs

Scaffolding and Blooms

Achieving psychometric validity through instructional design

How to use Messick's three questions:

Contexts in quantitative courses for three activities or tasks that comprise Ethical Reasoning

Stakeholder Analysis: a template for considering the impact of decisions in mathematical and quantitative practice.

Learning outcomes featuring Stakeholder Analysis, codes/ guidelines, and ethical reasoning at three Bloom's levels of complexity

Learning outcomes early in course (low Blooms)

Learning outcomes midway through course (middle Blooms)

Learning outcomes at end of course (high Blooms)

Discussion

Chapter 3 Ethics in Image Processing: Opportunities and Challenges

Helia Farhood4F(, Fariba Lotfi*5F(, Matineh Pooshideh*, Amin Beheshti*, and Samuel Muller*.

Abstract

Consent for the Use of Image Data

Consent and Ethical Obligation

Consent for Academic Purpose

Consent for Commercial Purpose

Image Manipulation

Facial Image Manipulation

Detecting Fake Images

Ethical Items in Image Collection and Datasets

Medical Imaging

Non-medical Imaging (Biometric)

Conclusion

Bibliography

Chapter 4 What is "Ethical AI"? Leading or Participating on an Ethical Team and/or Working in Statistics, Data Science, and Artificial Intelligence

Rochelle E. Tractenberg26F(

What is AI

What does "ethical AI" mean?

Association of Computing Machinery (ACM) Code of Ethics and Professional Conduct (2018)

ASA Ethical Guidelines for Statistical Practice (2022)

Conclusions

Chapter 5 Data Governance and Quality: Involving Data Sources' Owners

Irena Križman and Bruno Tissot29F(

The impact of the data revolution

Substituting or complementing official statistics?

Quality principles: from official statistics to the private sector

Towards a set of self-commitments for providers of alternative data?

Coordination between data providers

Partnership between data providers and official statisticians - good practice examples from Slovenia

Chapter 6 Facilitating the Integration of Ethical Reasoning into Quantitative Courses: Stakeholder Analysis, Ethical Practice Standards, and Case Studies

Rochelle E. Tractenberg39F(, Suzanne Thornton40F((

Options for "teaching ethics"

Case studies

Teaching Ethical Reasoning instead of "ethics"

Stakeholder analysis: a relevant subset of Ethical Reasoning

Using the ASA Ethical Guidelines explicitly: a relevant subset of Ethical Reasoning.

Integrating ethical content in a Mathematical Statistics course

Syllabus outline

Notes on syllabus elements: Unit One. Motivate and contextualize

Notes on syllabus elements: Unit Two. From examples to practice

Notes on syllabus elements: Unit Three. Independent and collaborative critical thinking

Discussion and Conclusions

APPENDIX: Syllabus for a First Course in Mathematical Statistics

Learning objectives

Topics outline for the semester

Unit One - Thoughtful use of estimation techniques

Unit Two - Statistical inference and stewardship

Unit Three - Disciplinary best practices for common study designs and analyses

Chapter 7 Ethical Decision-Making in Data Analysis: Navigating Challenges and Ensuring Integrity

Dr. Noah Mutai48F(

Ethical Considerations in Data Collection

Protecting Sensitive Data and Maintaining Anonymity

Balancing Benefits and Risks

Ensuring Representation and Inclusivity

Navigating Bias and Fairness

Selection Bias

Sampling Bias

Non-Response Bias

Measurement Bias

Confirmation Bias

Reporting Bias

Response Bias

Experimenter Bias

Sampling Frame Bias

Strategies for Bias Mitigation in Data Analysis

Random Sampling

Matching and Pairing

Blind Analysis

Variable Transformation

Sensitivity Analysis

External Validation

Transparency and Openness

Clear Communication of Methods in Data Analysis

Comprehensive Method Descriptions

Technical Detail and Jargon

Step-by-Step Breakdown

Rationale Behind Choices

Assumptions and Limitations

Avoiding Ambiguity

Transparency in Code and Scripts

Feedback and Peer Review

Documentation of Procedures in Data Analysis

Comprehensive Record Keeping

Timestamps and Dates

Data Cleaning and Transformation.

Software and Tools

Step-by-Step Descriptions

Variables and Definitions

Code and Scripts

Assumptions and Choices

Changes and Corrections

Feedback Incorporation

Archival and Storage

Open Data Practices in Data Analysis

Data Sharing

Data Depositories

Documentation

Metadata

Licensing and Terms of Use

Reproducibility and Replicability

Version Control

Data Privacy and Anonymization

Data Access Restrictions

Communication Channels

Engaging with the Community

Pre-Analysis Plans in Data Analysis

Defining Research Objectives

Hypothesis Formulation

Data Collection Procedures

Variable Selection and Transformation

Analysis Methods

Dealing with Multiple Comparisons

Controlling for Confounding Variables

Publication and Reporting Intentions

Transparency in Changes

Peer Review and Validation

Disclosure of Limitations in Data Analysis

Types of Limitations

Data Limitations

Methodological Limitations

Sampling Limitations

Scope and Generalizability

Assumption and Uncertainty

External Factors

Handling Uncertainty and Interpretation

Communicating Uncertainty Responsibly

Avoiding Overinterpretation

Contextualizing Results

Recognizing Implications and Limitations

Avoiding Sensationalism

Balancing Transparency with Actionability

Addressing Conflicts of Interest

Identifying Conflicts of Interest

Maintaining Objectivity and Independence

Disclosing Conflicts

Managing Conflicts

Balancing Ethical Considerations

Ethical Reporting

Impact on Society and Decision-Making

Responsible Communication of Results

Social, Economic, and Political Consequences

Mitigating Unintended Negative Consequences

References.

Chapter 8 Teaching at the Intersection of Social Justice, Ethics, and the ASA Ethical Guidelines for Statistical Practice

Rochelle E. Tractenberg49F(

Curriculum Development Guidelines

The Statistics and Data Science Pipeline

The ASA Ethical Guidelines for Statistical Practice

The Stakeholder Analysis

Ethical Reasoning Paradigm

Synthesizing these tools to integrate ethics and social justice

Dimension 1: practice ethically

Dimension 2: Identify &amp

respond to unethical act/request

Chapter 9 Validation of Statistical Results - An Ethical Perspective

G. S. Dissanayake and T. M. R. P. Yatigammana.

Data and models in statistics requiring validation for an ethical outcome.

Examples for validated data and models.

Chapter 10 Ethical Considerations for Data Involving Human Gender and Sex Variables

Suzanne Thornton50F(, Rochelle E. Tractenberg51F((

History of gender and sex data

Gender and sex minorities in (modern) ethical research design

Moving from connotative categorizations towards descriptive ones

Two-step questioning method

Direct questioning method

Reflections on these methods

Ethical considerations throughout the statistics and data science pipeline

Tasks 1-2: Planning/Designing and Data collection/munging/ wrangling

Tasks 3-4: Analysis (perform or program to perform) and Interpretation

Tasks 5-6: Documenting your work and Reporting your results/communication

Discussion and conclusion

Acknowledgements

Chapter 11 Empowering Minds to Nurture Ethical Awareness on Infographic Integrity Among Students and Educators.

Salma Banu Nazeer Khan53F(, Ayse Aysin Bilgin54F((, Deborah Richards* and Paul Formosa55F(((.

Notes:

Description based on publisher supplied metadata and other sources.

Description based on print version record.

Includes bibliographical references.

ISBN:

1-871891-66-3

1-871891-65-5

OCLC:

1432590675