Auditing the AI auditors: A framework for evaluating fairness and bias in high stakes AI predictive models

Researchers, governments, ethics watchdogs, and the public are increasingly voicing concerns about unfairness and bias in artificial intelligence (AI)-based decision tools. Psychology's more-than-a-century of research on the measurement of psychological traits and the prediction of human behavi...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:The American psychologist Ročník 78; číslo 1; s. 36
Hlavní autori: Landers, Richard N, Behrend, Tara S
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: United States 01.01.2023
Predmet:
ISSN:1935-990X, 1935-990X
On-line prístup:Zistit podrobnosti o prístupe
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract Researchers, governments, ethics watchdogs, and the public are increasingly voicing concerns about unfairness and bias in artificial intelligence (AI)-based decision tools. Psychology's more-than-a-century of research on the measurement of psychological traits and the prediction of human behavior can benefit such conversations, yet psychological researchers often find themselves excluded due to mismatches in terminology, values, and goals across disciplines. In the present paper, we begin to build a shared interdisciplinary understanding of AI fairness and bias by first presenting three major lenses, which vary in focus and prototypicality by discipline, from which to consider relevant issues: (a) individual attitudes, (b) legality, ethicality, and morality, and (c) embedded meanings within technical domains. Using these lenses, we next present as a standardized approach for evaluating the fairness and bias of AI systems that make predictions about humans across disciplinary perspectives. We present 12 crucial components to audits across three categories: (a) components related to AI models in terms of their source data, design, development, features, processes, and outputs, (b) components related to how information about models and their applications are presented, discussed, and understood from the perspectives of those employing the algorithm, those affected by decisions made using its predictions, and third-party observers, and (c) meta-components that must be considered across all other auditing components, including cultural context, respect for persons, and the integrity of individual research designs used to support all model developer claims. (PsycInfo Database Record (c) 2023 APA, all rights reserved).
AbstractList Researchers, governments, ethics watchdogs, and the public are increasingly voicing concerns about unfairness and bias in artificial intelligence (AI)-based decision tools. Psychology's more-than-a-century of research on the measurement of psychological traits and the prediction of human behavior can benefit such conversations, yet psychological researchers often find themselves excluded due to mismatches in terminology, values, and goals across disciplines. In the present paper, we begin to build a shared interdisciplinary understanding of AI fairness and bias by first presenting three major lenses, which vary in focus and prototypicality by discipline, from which to consider relevant issues: (a) individual attitudes, (b) legality, ethicality, and morality, and (c) embedded meanings within technical domains. Using these lenses, we next present as a standardized approach for evaluating the fairness and bias of AI systems that make predictions about humans across disciplinary perspectives. We present 12 crucial components to audits across three categories: (a) components related to AI models in terms of their source data, design, development, features, processes, and outputs, (b) components related to how information about models and their applications are presented, discussed, and understood from the perspectives of those employing the algorithm, those affected by decisions made using its predictions, and third-party observers, and (c) meta-components that must be considered across all other auditing components, including cultural context, respect for persons, and the integrity of individual research designs used to support all model developer claims. (PsycInfo Database Record (c) 2023 APA, all rights reserved).
Researchers, governments, ethics watchdogs, and the public are increasingly voicing concerns about unfairness and bias in artificial intelligence (AI)-based decision tools. Psychology's more-than-a-century of research on the measurement of psychological traits and the prediction of human behavior can benefit such conversations, yet psychological researchers often find themselves excluded due to mismatches in terminology, values, and goals across disciplines. In the present paper, we begin to build a shared interdisciplinary understanding of AI fairness and bias by first presenting three major lenses, which vary in focus and prototypicality by discipline, from which to consider relevant issues: (a) individual attitudes, (b) legality, ethicality, and morality, and (c) embedded meanings within technical domains. Using these lenses, we next present psychological audits as a standardized approach for evaluating the fairness and bias of AI systems that make predictions about humans across disciplinary perspectives. We present 12 crucial components to audits across three categories: (a) components related to AI models in terms of their source data, design, development, features, processes, and outputs, (b) components related to how information about models and their applications are presented, discussed, and understood from the perspectives of those employing the algorithm, those affected by decisions made using its predictions, and third-party observers, and (c) meta-components that must be considered across all other auditing components, including cultural context, respect for persons, and the integrity of individual research designs used to support all model developer claims. (PsycInfo Database Record (c) 2023 APA, all rights reserved).Researchers, governments, ethics watchdogs, and the public are increasingly voicing concerns about unfairness and bias in artificial intelligence (AI)-based decision tools. Psychology's more-than-a-century of research on the measurement of psychological traits and the prediction of human behavior can benefit such conversations, yet psychological researchers often find themselves excluded due to mismatches in terminology, values, and goals across disciplines. In the present paper, we begin to build a shared interdisciplinary understanding of AI fairness and bias by first presenting three major lenses, which vary in focus and prototypicality by discipline, from which to consider relevant issues: (a) individual attitudes, (b) legality, ethicality, and morality, and (c) embedded meanings within technical domains. Using these lenses, we next present psychological audits as a standardized approach for evaluating the fairness and bias of AI systems that make predictions about humans across disciplinary perspectives. We present 12 crucial components to audits across three categories: (a) components related to AI models in terms of their source data, design, development, features, processes, and outputs, (b) components related to how information about models and their applications are presented, discussed, and understood from the perspectives of those employing the algorithm, those affected by decisions made using its predictions, and third-party observers, and (c) meta-components that must be considered across all other auditing components, including cultural context, respect for persons, and the integrity of individual research designs used to support all model developer claims. (PsycInfo Database Record (c) 2023 APA, all rights reserved).
Author Landers, Richard N
Behrend, Tara S
Author_xml – sequence: 1
  givenname: Richard N
  orcidid: 0000-0001-5611-2923
  surname: Landers
  fullname: Landers, Richard N
  organization: Department of Psychology
– sequence: 2
  givenname: Tara S
  orcidid: 0000-0002-7943-5298
  surname: Behrend
  fullname: Behrend, Tara S
  organization: Department of Psychological Sciences
BackLink https://www.ncbi.nlm.nih.gov/pubmed/35157476$$D View this record in MEDLINE/PubMed
BookMark eNpNkMtOwzAURC1URB-w4QOQl2wCsRPHNruq4lEJiQ1I7KIb-6Z1mxd2UsTf00KRmM3MSEezmCkZNW2DhFyy-IbFibyFuov30pKfkAnTiYi0jt9H__KYTEPY7BmhNDsj40QwIVOZTchmPljXu2ZF-zXS-ZLCobc-3NE5LT3U-Nn6LS1bT3EH1QA_bAnONxgChcbSwkGgrqFrt1rT0MMWw2Go82id6d0Oad1arMI5OS2hCnhx9Bl5e7h_XTxFzy-Py8X8OYKUpX0khC5EqkwmTCpKFaOUZWK5zZgtCp1mhhkpYmERjEqsEiVXwuiEaSnBSq74jFz_7na-_Rgw9HntgsGqggbbIeQ84zoWWskDenVEh6JGm3fe1eC_8r9_-DeLrGia
CitedBy_id crossref_primary_10_1080_00207543_2024_2432463
crossref_primary_10_1007_s12528_025_09467_z
crossref_primary_10_3389_feduc_2025_1565938
crossref_primary_10_1016_j_accinf_2025_100739
crossref_primary_10_1080_00049530_2024_2419682
crossref_primary_10_1177_20539517241290217
crossref_primary_10_1007_s10869_025_10035_6
crossref_primary_10_1017_iop_2022_44
crossref_primary_10_1016_j_copsyc_2024_101836
crossref_primary_10_1109_TAFFC_2023_3296695
crossref_primary_10_3390_diagnostics14151594
crossref_primary_10_1111_peps_12643
crossref_primary_10_1007_s10869_022_09829_9
crossref_primary_10_1007_s13748_024_00345_w
crossref_primary_10_1080_2331186X_2025_2490425
crossref_primary_10_7717_peerj_cs_1630
crossref_primary_10_1007_s43681_025_00728_2
crossref_primary_10_1371_journal_pone_0318486
crossref_primary_10_2196_69007
crossref_primary_10_1080_10494820_2025_2546630
crossref_primary_10_1080_2331186X_2025_2526436
crossref_primary_10_1080_08874417_2024_2417672
crossref_primary_10_1016_j_jrt_2025_100129
crossref_primary_10_1016_j_inffus_2024_102760
crossref_primary_10_1080_10508422_2024_2332699
crossref_primary_10_1177_08404704241291226
crossref_primary_10_1007_s12027_024_00785_w
crossref_primary_10_1007_s43681_024_00590_8
crossref_primary_10_3390_jcdd10120485
crossref_primary_10_20525_ijrbs_v12i9_2737
crossref_primary_10_1111_ijsa_12396
crossref_primary_10_1111_peps_12593
crossref_primary_10_1007_s10869_025_10022_x
crossref_primary_10_1016_j_copsyc_2024_101829
crossref_primary_10_36096_ijbes_v7i1_641
crossref_primary_10_1111_emip_12593
crossref_primary_10_2174_0113816128287941231206050340
crossref_primary_10_1007_s43681_023_00266_9
crossref_primary_10_1016_j_jflm_2025_102895
crossref_primary_10_1007_s43681_024_00529_z
crossref_primary_10_1097_ACM_0000000000005980
crossref_primary_10_1080_1359432X_2025_2517592
crossref_primary_10_2196_48633
crossref_primary_10_4102_sajip_v48i0_1988
crossref_primary_10_1007_s43681_024_00450_5
crossref_primary_10_1177_20539517241299732
crossref_primary_10_32604_cmes_2023_029451
crossref_primary_10_7759_cureus_49252
crossref_primary_10_3390_info16050400
crossref_primary_10_1109_ACCESS_2024_3521279
crossref_primary_10_1109_ACCESS_2024_3375763
crossref_primary_10_3390_cancers14122897
crossref_primary_10_1080_00036846_2024_2321835
crossref_primary_10_1080_10447318_2023_2232982
crossref_primary_10_1080_23311975_2024_2344743
crossref_primary_10_1007_s43681_024_00445_2
crossref_primary_10_3389_fpsyg_2024_1353022
crossref_primary_10_1177_10422587241304676
crossref_primary_10_1016_j_sciaf_2024_e02281
crossref_primary_10_1109_JPROC_2024_3403898
crossref_primary_10_1111_ijsa_12459
crossref_primary_10_1057_s41262_025_00383_2
crossref_primary_10_1111_ijsa_12456
crossref_primary_10_1111_ijsa_12499
crossref_primary_10_1007_s10639_024_12605_2
crossref_primary_10_1111_ijsa_12376
crossref_primary_10_1111_jedm_12330
crossref_primary_10_3389_feduc_2025_1499497
crossref_primary_10_1145_3710908
crossref_primary_10_3917_inno_pr2_0153
crossref_primary_10_1080_10447318_2024_2331858
crossref_primary_10_1007_s10869_025_10067_y
crossref_primary_10_3389_fpsyg_2024_1390182
crossref_primary_10_1080_00140139_2025_2512426
crossref_primary_10_1108_IJOES_02_2024_0060
crossref_primary_10_1111_peps_12578
crossref_primary_10_1108_RAF_01_2025_0006
crossref_primary_10_1016_j_clsr_2024_106053
crossref_primary_10_1016_j_genhosppsych_2025_02_018
crossref_primary_10_1007_s44206_023_00074_y
ContentType Journal Article
DBID CGR
CUY
CVF
ECM
EIF
NPM
7X8
DOI 10.1037/amp0000972
DatabaseName Medline
MEDLINE
MEDLINE (Ovid)
MEDLINE
MEDLINE
PubMed
MEDLINE - Academic
DatabaseTitle MEDLINE
Medline Complete
MEDLINE with Full Text
PubMed
MEDLINE (Ovid)
MEDLINE - Academic
DatabaseTitleList MEDLINE
MEDLINE - Academic
Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod no_fulltext_linktorsrc
Discipline Psychology
EISSN 1935-990X
ExternalDocumentID 35157476
Genre Journal Article
GroupedDBID ---
--Z
-DZ
-ET
-~X
.GJ
07C
0R~
186
23M
2KS
354
3EH
53G
5GY
5RE
5VS
6TJ
7RZ
85S
9M8
AAAHA
AAIKC
AAMNW
AAWTL
AAYOK
ABCQX
ABIVO
ABNCP
ABPPZ
ACGFO
ACHQT
ACKIV
ACNCT
ACPQG
ACTDY
ADMHG
AEHFB
AFFNX
AGNAY
AI.
ALEEW
ALMA_UNASSIGNED_HOLDINGS
AS9
ASUFR
AWKKM
AZXWR
BKOMP
CGNQK
CGR
CS3
CUY
CVF
D~A
ECM
EIF
EPA
F5P
FTD
HVGLF
HZ~
ISO
L7B
LPU
LW5
MS~
MVM
NHB
NPM
O9-
OHT
OMK
OPA
OVD
P2P
PQQKQ
ROL
RXW
SES
SKT
SPA
TAE
TEORI
TN5
TWZ
UBC
UCV
UHB
UHS
ULE
UPT
URZ
VH1
VQA
VQP
WH7
X6Y
XIH
XJT
XOL
XZL
YCJ
YCV
YIF
YIN
YQI
YR5
YRY
YYP
YYQ
YZZ
ZCA
ZCG
ZGI
ZHY
ZKG
ZPI
~02
~A~
3KI
7X8
ABVOZ
PHGZT
PUEGO
ID FETCH-LOGICAL-a414t-559b548c65c45f80e77f3d2d61dbb946c1c7505deac83d85f285c931977ad7282
IEDL.DBID 7X8
ISICitedReferencesCount 103
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000754045100001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1935-990X
IngestDate Thu Oct 02 10:28:06 EDT 2025
Wed Feb 19 02:25:49 EST 2025
IsDoiOpenAccess false
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 1
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-a414t-559b548c65c45f80e77f3d2d61dbb946c1c7505deac83d85f285c931977ad7282
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ORCID 0000-0001-5611-2923
0000-0002-7943-5298
OpenAccessLink https://doi.org/10.1037/amp0000972
PMID 35157476
PQID 2629059878
PQPubID 23479
ParticipantIDs proquest_miscellaneous_2629059878
pubmed_primary_35157476
PublicationCentury 2000
PublicationDate 2023-01-01
PublicationDateYYYYMMDD 2023-01-01
PublicationDate_xml – month: 01
  year: 2023
  text: 2023-01-01
  day: 01
PublicationDecade 2020
PublicationPlace United States
PublicationPlace_xml – name: United States
PublicationTitle The American psychologist
PublicationTitleAlternate Am Psychol
PublicationYear 2023
SSID ssj0005891
Score 2.6852164
Snippet Researchers, governments, ethics watchdogs, and the public are increasingly voicing concerns about unfairness and bias in artificial intelligence (AI)-based...
SourceID proquest
pubmed
SourceType Aggregation Database
Index Database
StartPage 36
SubjectTerms Artificial Intelligence
Humans
Title Auditing the AI auditors: A framework for evaluating fairness and bias in high stakes AI predictive models
URI https://www.ncbi.nlm.nih.gov/pubmed/35157476
https://www.proquest.com/docview/2629059878
Volume 78
WOSCitedRecordID wos000754045100001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1JS8NAFB7UeujFfakbI3gdmm2WeJEgFj1YelDILcwWqGISmyr4732TxZ4EwUsgkIQwee-b782bfB9CVz6TQaCYIVpSRoD_e0T5ghFuhBEq9qmhqjGb4NOpSNN41i241d22yh4TG6A2pXZr5OOABTFQAcHFTfVOnGuU6652FhrraBAClXGJydOVWnjnmAcchRJA3bSXJw35WL5VDT1qlIF_oZbNFDPZ_u_L7aCtjlzipI2GXbRmiz00_MG4r330kri_MGC6wkD8cPKApTsvF_U1TnDeb9TCwGRxrwMO17qmj4NELAuD1VzWeF5gJ3SMgVu-2to9qFq4lo8DT9y469QH6Hly93R7Tzq7BSIjP1oSqC0U1C-aUR3RXHiW8zw0gWG-USqOmPY10AtqAKpFaATNA0F1DCnMuTQcSrdDtFGUhT1G2PqB4dwz0ioZWcliX-ee8ZhVgoWRoiN02Y9jBuHsehSysOVHna1GcoSO2o-RVa3uRhYC94Lqh5384e5TNHTG8O1iyRka5JDM9hxt6s_lvF5cNHECx-ns8RubiskN
linkProvider ProQuest
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Auditing+the+AI+auditors%3A+A+framework+for+evaluating+fairness+and+bias+in+high+stakes+AI+predictive+models&rft.jtitle=The+American+psychologist&rft.au=Landers%2C+Richard+N&rft.au=Behrend%2C+Tara+S&rft.date=2023-01-01&rft.eissn=1935-990X&rft.volume=78&rft.issue=1&rft.spage=36&rft_id=info:doi/10.1037%2Famp0000972&rft_id=info%3Apmid%2F35157476&rft_id=info%3Apmid%2F35157476&rft.externalDocID=35157476
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1935-990X&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1935-990X&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1935-990X&client=summon