McGill PSYCH 306 Research Methods in Psychology Chapter 4 PDF

PSYCH 306 Research Methods in Psychology Chapter 4: Research Ethics 1 Outline Canadianresearch guidelines CPA Issues & human participants – Historical context – CPA guidelines Animal subjects – Historical context – APA guidelines Scient...

PSYCH 306 Research Methods in Psychology Chapter 4: Research Ethics 1 Outline Canadianresearch guidelines CPA Issues & human participants – Historical context – CPA guidelines Animal subjects – Historical context – APA guidelines Scientific integrity 2  Research Ethics Help scientists to define what is legitimate to do or not What a ‘moral’ research procedure involves Includes the concerns, dilemmas, and conflicts that arise over the proper way to conduct research Applies to both participants and experimenters 3  Research Ethics Encompasses everything: – Measurement techniques – Participant selection – Which research designs and strategies can be used with certain populations – How data is analyzed – How results are reported 4  Research Ethics Acting ethically requires that a researcher balance the value of advancing knowledge against the value of non-interference in the lives of others Protecting research Gaining knowledge participants and and finding a clear upholding human answer to a research rights. question. 5 Historical Roots After WWII: Nuremburg Code (1947) as a result of atrocious Nazi experiments – Foundation of ethical guidelines used today for psychological and medical research 6 Nuremberg Code 10 important guidelines: 1. Consent sometime animal can giveimplicit 2. Benefit of knowledge for society consent 3. Knowledge of anticipated results (animal studies) 4. No unnecessary physical/mental suffering 5. No risk of death 6. Risk must be lower than importance of problem riskbenefit 7. Adequate facilities researchfacilities tradof underthebestconditionspossible 8. Competence of researchers troffs guidelines 9. Participants' withdrawal allowed 10. Termination of study by researchers 7 Belmont Report (US, 1979) from Milgrim'sreport Outlines three core principles: issueof obédience 1. Respect of persons – Individuals should consent; those cannot need to be protected parent otheradult 2. Beneficence (act of mercy, kindness; doing good for others) – No harm, minimize risks, maximize benefits 3. Justice – Fairness in procedures for selecting participants 8  Tri-Council Policy Statement (Canada, 2007) Three core principles: Respect for Persons: recognize intrinsic value of participants as people; informed and ongoing consent; accountability and transparency. Concern for Welfare: ensure the privacy and control of information about Ps (confidentiality); do not put in harm’s way; advise of any and all risks. Justice: equal respect and concern for all Ps; no segment of population unduly burdened by harms of research; selection according to inclusion criteria that is justified by questions. justifying youractions 9  Canadian Code of Ethics for Psychologists (CPA, 2017) Respect for dignity of persons (most important): – Do no harm; informed consent; protection of privacy; protection of vulnerable populations and individuals factsaboutparticipationwhichtheyshouldagree Responsible caring: – Competence; maximize benefit, minimize harm ofresearcher forpopulation ofparticipants Integrity in relationships:specifictoCanadiancode – Accurate and honest (no fraud, plagiarism, suppression of findings); minimize deception & debrief afterexperiment Responsibility to society: – Contribute to discipline of psychology and to the good of society 10 Informed Consent Three principles: 1. Participants must be informed of what will be done to them & why (purpose of the study) in standard consent form – Sometimes complete disclosure is impossible beforehand biaisesinfluencebehavioronstudy Especially the “why” of the study; participants might alter their behaviour with that knowledge – In that case, tell exactly what will be done, but not why until the end debrief 11 Informed Consent Special/vulnerable populations: – Individuals lacking cognitive capacity or full freedom to give true informed consent Additional protection: Effieinformed consent relativecangiveconsent – Children, developmental disability – Also those susceptible to undue influence Economically disadvantaged, lack health care, etc. implicitcoercion ifoffer amount sameamountof fortime sonocoercion In the case of children & vulnerable populations, researchers must seek assent from participants & consent from guardians 12 Informed Consent 2. Participants must have complete understanding: – Informing is not enough; does not mean they understand details to consentforms somejustsign you can make them paraphaseparts of consentform 13 Informed Consent ex offering a lot of compared to 3. Participation must be voluntary and not coerced otherstudies – Right to withdraw at any time, without consequences 14 Observation & Consent Consent is not required for the observation of people in public places where: a) There is no intervention staged by the researcher, or direct interaction with the individuals or groups b) There is no reasonable expectation of privacy expublicmetrostop classroom phffpafemission c) Dissemination of research results does not allow identification of specific individuals if inpublicspace participant identitycan'tbelearnedfromdata 15 Deception When participants are not given complete and accurate information Two types of deception: – Passive: leave out information presfftsftp.ttaistherefarch – Active: alter information (give false feedback) or use confederates (who pose as participants) jusEymet benefitsmustoutweightrisk fakseifthticipant Youhandle Must be justified (benefits must outweigh risks) – Consider all alternatives (and justify rejection) to REB (Research Ethics Board) for approval Cannot conceal information about physical pain or severe emotional distress ex stress pighhyfaffffhtheforeandithastobejustified Must immediately debrief with a complete explanation 16  hee I be pa allyver w a lyread Confidentiality & Anonymity Practice of keeping all individual information obtained during a study in a private & secure location Anonymity: Names not associated with data Code numbers rather than participant names are used theybecomeanonymised codenumbers onlynamesforconsentforms Only group data is reported – Example: averages for thegroupstandarddeviation – No single participant examined unlesscasestudy Exceptions: medical/neuroscience cases in which the participants are special cases that are repeated across studies example: study of HM’s memory loss 17 Confidentiality & Anonymity – Access to the data is limited to the research team members listed on the ethics protocol – Password protected files, encryption, hard copies kept in locked file cabinets – Traveling with laptops: make sure you do not leave research files open where others can view them! 18  Animal Research Ethics can't test on humans Why? – To learn more about societies – To learn more about humans – To learn more about a behaviour – To conduct research that is impossible to do with humans 19 Historical Roots Animal Welfare Act (1966): – General standards for animal care in research and exhibition – minimal standard – Last amended in 2008 Canadian Council for Animal Care (1968): – Standards for care, treatment, and use of animals in science in Canada – To prevent harm to animals 20  Canadian Council on Animal Care no consent important to justify Governs all animal research in Canada Laws and guidelines for animal housing, feeding, cleanliness, health Guidelines require no to little pain or stress Each university has an Animal Care Committee that reviews ethics proposals with animals 3 principles for reviewing animal research: couldyoureplaceanimalandstilldoresearch Replacement - replacing animals with an alternative (such as AI model) Reduction- minimizing number of animals used Refinement- modifying procedures to minimize distress or stress 21 Majority of Animal Research is not for comparison with humans Keepingthemhealthy foodchainhealthy informhumansociety 22 University Research Ethics Boards: Decision Process Any research team must apply to a research ethics committee for permission and approval before research begins These committees exist in all universities and funding agencies At McGill: 4-5 committees (behavioral-adult, behavioral- minors/vulnerable populations, medical, animal) YES: Conduct research Assess Do Benefits Assess Potential Outweigh Risks? Benefits Risks NO: Study Protocol must Be modified 23  Scientific Ethics / Integrity: Publication Issues erratumretractionsftp.ffuseisn honestmistake Mistake versus fraud (erratum / retraction) Data fabrication & falsification of findings changesdatato beabletopublishit Plagiarism of sources easyto tell whenfake Ghost-writing & fake peer reviewspâpersheyhavetoacknowledge hiresomeonetowritepapersfamiliarwithscientific ghostwriters Safeguards: replication, peer review & watchdogs journalwebsite peerscientist theyidentifyfraud in publishfer 24  Scientific Integrity Ethics violations can occur at the data analysis and publication phase Scientific misconduct: – Violating basic and generally accepted standards of honest scientific research – Such as research fraud, plagiarism, or suppressed findings datff.EE intendedto find nqyyt 3timespublishwithoutfirst2 25  Scientific Integrity Distinction between unintended mistakes and fraud: Research mistakes: - Reporting values or analyses mistakenly that are caught after publication - Authors can publish an “erratum” for errors that were inadvertently created report errors on initial publication Research fraud: ofresearch adjustpart – To invent, falsify or distort study data or to lie about how a study was conducted – Editors can publish a “retraction” for important offenses (fraud, plagiarism, duplicate publication, etc) 26  Research Fraud Suppressed Findings moreindustrysettings less in uni – Sometimes, studies try to conceal their findings e.g., poor safety in some industry records e.g., lottery commission asking a sociologist to remove from her report section on negative social effects from gambling 27  Plagiarism Using another person’s words or ideas without giving them proper credit – Instead passing them off as your own More easily detected nowadays honestmistake read somuch you think it's your idea 28  Safeguards Replication, peer review, & watchdogs can take 1 year 29  Safeguards for Scientific Integrity – Retraction Watch Blog started in 2010 by health journalists Ivan Oransky and Adam Marcus started to keep track of bad science Founded the Center for Science Integrity Created a searchable database with over 18,000 retractions 50% of retractions are due to fabrication, falsification, or plagiarism 10% due to forged authorship, fake peer reviews, and failure to obtain ethics approval 30 Research Fraud 31 Research Fraud Stapel: a Dutch former professor of social psychology, suspended by Tilburg University for fabricating and manipulating data in his research publications. By 2015, 58 of Stapel’s publications had been retracted. 32 Pressures for Unethical Research? Publish or perish: – Career building pressure, need to publish, gain prestige, tenure clock, etc. Must obtain significant findings to publish – Ethical research takes a long time to complete – Costs more money Need for success, admiration 33  Ethics in Social Media Truth or Falsehood? Nokia is installing internet A university banned the use of capital letters on the moon. in email to avoid scaring students. true UK Independent, UK, Oct 2020 true The Guardian, 2021 34  Truth or Falsehood? Gorilla learns to knit Spinach learns to send email CNN, April 1, 2009 story Euronews, April 2016 story false true 35  sou a s Why is Fake News Important? United Nations: Information pollution affects citizens’ capacity to make informed decisions. Disinformation, misinformation, and mal-information are creating mistrust in public institutions. www.undp.org/eurasia/dis/misinformation 36  Defining Fake News DISINFORMATION Information that is false and deliberately created to harm a person, social group, organization or country. Sometimes called “fake news” Example: UFOs are the result of invasion by space aliens MISINFORMATION Information that is false, but not deliberately created with the intention of misleading readers or causing harm. Example: COVID is primarily transmitted through surfaces (later found to be airborne) MALINFORMATION (hardest to discriminate from true news) Information that is based on real facts, but deliberately manipulated to inflict harm on a person, organization or country. true but turned into something not so true to harm 37  Why is Disinformation Effective? Cherry-picking: presenting only facts that support a certain view. Often used to support any contrary or equivocal evidence, presented as a challenge to the scientific consensus: what want to see yousee you confirmation biais Example: tobacco companies paid for research that uncovered genuine links between lung cancer and other (non-tobacco) factors and promoted those as evidence against the stronger link between lung cancer and smoking 38  Why is Disinformation Effective? Double standard: holding evidence that supports the scientific consensus to a higher standard than the evidence that challenges it. Example: anti-vaccination arguments that consider the small number of studies that imply a link between vaccination and conditions such as autism as having the same weight as the many studies that show no such link. Certain outcomes to a higher standard Reliance on false experts: Amplifying the opinions of a few scientists who challenge the consensus. Example: Dr. David Smith, PhD, a chiropractor and naturopath, has given his opinion on the effect of fluoride on schizophrenia. 39  Conspiracy Theories: Attempt to explain harmful or tragic events as the result of the actions of a small powerful group. Reject the accepted evidence or explanation; often based on easily falsifiable evidence. Driven by a desire to make sense of social forces that are self-relevant, important, and threatening Examples: UFOs seen near an Air Force base as evidence of extraterrestrial life Nevada base that tested several planes (U-2) Encyclopedia Britannica, 40 2023  Pizzagate Conspiracy Theory 2016: Staff member of Hilary Clinton (US presidential candidate) had their emails hacked and published online Conspiracy theorists claimed the emails contained coded messages that connected several Democratic Party officials and U.S. restaurants with human trafficking rings. Named after Comet Ping Pong Pizzaria, one of the restaurants in Washington, D.C The Pizzaria was threatened by hundreds of people LA Times, 2017; for several weeks. Washington Post, 2017 41  Psychological causes of conspiracy theories: Attribution error: people overestimate causes that arise from human motives, while underestimating causes related to situation factors (social context, random chance) Confirmation bias: tendency to attend to, focus on, and give greater credence to evidence that fits with our existing beliefs. All Our Evidence evidence beliefs we ignore Evidence we believe https://www.scribbr.com/research-bias/confirmation-bias/ 42  Experimental Evidence for Confirmation Bias People are more likely to believe evidence that supports their pre-existing views 2 groups of students who believed either that: - capital punishment reduces crime rates - capital punishment does not reduce crime Lord et al, 1979 Each student read 2 stories: A) story supporting the idea that CP reduces crime B) story supporting the idea that CP does not reduce crime Then rated their belief in whether CP reduces crime after reading the stories: Students who believed CP reduces crime thought story A) was credible and story B) was not Students who believed CP does not reduce crime said that story B) was credible and A) was not Both groups were more committed to their original perspective than earlier 43  Analysis of Confirmation Bias Analysis of 126,000 verified false and true news stories tweeted (Twitter) from 2006-2017. False news spread more quickly by Twitter readers than true news. 3 measures of how news is spread on Twitter: in cascades of # individuals (size) over several iterations (depth) and minutes (time) # minutes elapsed 3 Iterations (Depth) Start (first tweet) Shared with 3 individuals (Size) Vosoughi, Roy, & Aral (2018) Science 44 False News spreads more quickly than Truth Size: False news spreads to more people % of total tweets News is spread in cascades of # individuals (size) over several iterations (depth) and minutes (time) Depth: False news spreads over more iterations Time: False news spreads in fewer minutes slower % of total tweets faster Vosoughi et al, Science, 2018 45 dept ended sooner mi true s false news went faster Disinformation Spreads in Echo Chambers Why does false news spread more quickly? One reason: echo chambers (blogs, forums) Echo chamber = groups of users who share a strong opinion and align themselves in a group where they are exposed to content similar to their beliefs Beliefs become strengthened by repeated interactions with sources who share the same bias Online algorithms learn to select content that matches what a reader wants to hear Study of echo chambers in Reddit and Twitter Cinelli, 2021, PNAS examples: Death penalty, abortion rights, climate change 46 How Echo Chambers Spread News More important is who said it than what is said High-hate posts do not spread more quickly than low-hate posts: Goel, 2023, PNAS What is said doesn’t matter Posts by High-hate users spread more quickly than posts by lower-hate users: Who says it matters Source attribution: Users disseminate news more often from a trusted source than from a trusted fact 47  Disinformation through Spoofing Spoofing: disguising a communication from an unknown source as originating from a known, trusted source (reliance on false experts) Caller ID spoofing: deliberately falsifying the information shown on a phone’s caller ID display to disguise an identity. Email spoofing: email sent from a false sender address, asking the recipient to provide sensitive data. Example: Dear ones, I am a recent widow and I need your help to access my $5million account. All you need to do is share your bank information with me Online news spoofing: Posting a fake news story on a fraudulent website designed to look legitimate Spoofing content uses urgent and emotional language to convince people. How to check the validity of a website: https://www.factcheck.org 48 Disinformation through Image Editing 1930s: Nikolai Yezhov was removed from a group photograph with Joseph Stalin by government censors after he fell out of favor with the Soviet government. lot more now with AI https://cerrocoso.libguides.com/fakenews/images 49 Disinformation through Image Editing https://cerrocoso.libguides.com/fakenews/images Left image: Activist Emma Right image: edited by outside Gonzalez (2018, Florida) was party to show her tearing up photographed tearing up a target the US Constitution, shared on poster. social media, sparked outrage 50 Disinformation through Image Editing US Presidential candidate John Kerry was photoshopped into picture with famous anti-war activist Jane Fonda When to test whether an image has been unedited? When the photo triggers a strong emotional reaction or sends a very strong message 51 Testing whether an Image has been edited Google images: copy an image into the search bar to find where it originated (compare entered photographs against copies online). Tools also show who created the photo and where else it has appeared online. Use your eyes: reflective surfaces should not change (mirrors); check for size changes in objects that appear smaller or larger than they should; check for shadow inconsistencies. Check the metadata: a cache of information stored in an image’s file for clues (right-click on image, then Properties / Details). Reveals details like where the picture was taken, when it was last saved, and copyright. 52  Whether an Image has been edited Image Cloning: Whether an image was superimposed on the original image (change in # pixels) Lighting: inconsistencies in shadows, Queen Elizabeth and Prince William, London, England reflections, light sources Facial features: Whether the irises of a person’s eyes match Daily Mail, UK, July 2013 53  When are scientific findings false news? Some journals pretend to be real scientific journals by hijacking (taking without consent) the journal name, publishers’ and editors’ names Example: predatory journals: They accept articles for publication, along with authors’ fees They do not perform usual quality checks for plagiarism or ethical approval They claim to follow a regular peer-review process but do not do so They are often named like a bona fide peer-reviewed journal Copycats Bona Fide Frontier Scientific Publishing Frontiers in Psychology Naturepub Nature Academy of Science and Social Science Proceedings of National Academy of Sciences To check whether a journal is bona fide or predatory: enter its name in https://beallslist.net/#update What is a vanity journal? One that does not hide the fact that it publishes for a fee with no review 54  Retractions of Science Papers Do MMR vaccines cause autism in children? Wakefield et al (1998) published a paper in the Lancet - a peer-reviewed scientific journal - retracted because n=12 and no direct evidence - epidemiological studies with large numbers yielded different outcomes Year of retraction #Citations #Citations before 2010 after 2010 2010 643 940 It takes a long time for a retracted article to be recognized as “false news” Why? Startling findings are talked about more than retractions (less exciting) Followup news on old topics is less interesting to readers (does not sell) Some scientists did not read the retraction (continued to cite original) 55  How to Evaluate Disinformation www.factcheck.org: Public awareness program to help spot false News Is this a credible Source? Compare with other credible sources (www.factcheck.org/scicheck/) to determine validity of the scientific claims made in the media Is the Perspective biased? Are the Experts bona fide? Are other credible sources reporting the same story? 56  NEXT TIME: CHAPTER 13 THE DESCRIPTIVE RESEARCH STRATEGY 57