A Medley of Potpourri

Wednesday, September 4, 2024

Academic authorship

From Wikipedia, the free encyclopedia
https://en.wikipedia.org/wiki/Academic_authorship

Academic authorship of journal articles, books, and other original works is a means by which academics communicate the results of their scholarly work, establish priority for their discoveries, and build their reputation among their peers.

Authorship is a primary basis that employers use to evaluate academic personnel for employment, promotion, and tenure. In academic publishing, authorship of a work is claimed by those making intellectual contributions to the completion of the research described in the work. In simple cases, a solitary scholar carries out a research project and writes the subsequent article or book. In many disciplines, however, collaboration is the norm and issues of authorship can be controversial. In these contexts, authorship can encompass activities other than writing the article; a researcher who comes up with an experimental design and analyzes the data may be considered an author, even if she or he had little role in composing the text describing the results. According to some standards, even writing the entire article would not constitute authorship unless the writer was also involved in at least one other phase of the project.

Definition

Guidelines for assigning authorship vary between institutions and disciplines. They may be formally defined or simply cultural norms. Incorrect assignment of authorship occasionally leads to charges of academic misconduct and sanctions for the violator. A 2002 survey of a large sample of researchers who had received funding from the U.S. National Institutes of Health revealed that 10% of respondents claimed to have inappropriately assigned authorship credit within the last three years. This was the first large scale survey concerning such issues. In other fields only limited or no empirical data is available.

Authorship in the natural sciences

The natural sciences have no universal standard for authorship, but some major multi-disciplinary journals and institutions have established guidelines for work that they publish. The journal Proceedings of the National Academy of Sciences of the United States of America (PNAS) has an editorial policy that specifies "authorship should be limited to those who have contributed substantially to the work" and furthermore, "authors are strongly encouraged to indicate their specific contributions" as a footnote. The American Chemical Society further specifies that authors are those who also "share responsibility and accountability for the results" and the U.S. National Academies specify "an author who is willing to take credit for a paper must also bear responsibility for its contents. Thus, unless a footnote or the text of the paper explicitly assigns responsibility for different parts of the paper to different authors, the authors whose names appear on a paper must share responsibility for all of it."

Authorship in mathematics

In mathematics, the authors are usually listed in alphabetical order (the so-called Hardy-Littlewood Rule).

Authorship in medicine

The medical field defines authorship very narrowly. According to the Uniform Requirements for Manuscripts Submitted to Biomedical Journals, designation as an author must satisfy four conditions. The author must have:

Contributed substantially to the conception and design of the study, the acquisition of data, or the analysis and interpretation
Drafted or provided critical revision of the article
Provided final approval of the version to publish
Agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved

Acquisition of funding, or general supervision of the research group alone does not constitute authorship. Biomedical authorship is prone to various misconducts and disputes. Many authors – especially those in the middle of the byline – do not fulfill these authorship criteria. Some medical journals have abandoned the strict notion of author, with the flexible notion of contributor.

Authorship in the social sciences

The American Psychological Association (APA) has similar guidelines as medicine for authorship. The APA acknowledge that authorship is not limited to the writing of manuscripts, but must include those who have made substantial contributions to a study such as "formulating the problem or hypothesis, structuring the experimental design, organizing and conducting the statistical analysis, interpreting the results, or writing a major portion of the paper". While the APA guidelines list many other forms of contributions to a study that do not constitute authorship, it does state that combinations of these and other tasks may justify authorship. Like medicine, the APA considers institutional position, such as Department Chair, insufficient for attributing authorship.

Authorship in the humanities

Neither the Modern Languages Association nor the Chicago Manual of Style define requirements for authorship (because usually humanities works are single-authored and the author is responsible for the entire work).

Growing number of authors per paper

From the late 17th century to the 1920s, sole authorship was the norm, and the one-paper-one-author model worked well for distributing credit. Today, shared authorship is common in most academic disciplines, with the exception of the humanities, where sole authorship is still the predominant model. Between about 1980-2010 the average number of authors in medical papers increased, and perhaps tripled. One survey found that in mathematics journals over the first decade of the 2000's, "the number of papers with 2, 3 and 4+ authors increased by approximately 50%, 100% and 200%, respectively, while single author papers decreased slightly."

In particular types of research, including particle physics, genome sequencing and clinical trials, a paper's author list can run into the hundreds. In 1998, the Collider Detector at Fermilab (CDF) adopted a (at that time) highly unorthodox policy for assigning authorship. CDF maintains a standard author list. All scientists and engineers working at CDF are added to the standard author list after one year of full-time work; names stay on the list until one year after the worker leaves CDF. Every publication coming out of CDF uses the entire standard author list, in alphabetical order. Other big collaborations, including most particle physics experiments, followed this model.

In large, multi-center clinical trials authorship is often used as a reward for recruiting patients. A paper published in the New England Journal of Medicine in 1993 reported on a clinical trial conducted in 1,081 hospitals in 15 different countries, involving a total of 41,021 patients. There were 972 authors listed in an appendix and authorship was assigned to a group. In 2015, an article in high-energy physics was published describing the measurement of the mass of the Higgs boson based on collisions in the Large Hadron Collider; the article boasted 5,154 authors, the printed author list needed 24 pages.

Large authors lists have attracted some criticism. They strain guidelines that insist that each author's role be described and that each author is responsible for the validity of the whole work. Such a system treats authorship more as credit for scientific service at the facility in general rather that as an identification of specific contributions. One commentator wrote, "In more than 25 years working as a scientific editor ... I have not been aware of any valid argument for more than three authors per paper, although I recognize that this may not be true for every field." The rise of shared authorship has been attributed to Big Science—scientific experiments that require collaboration and specialization of many individuals.

Alternatively, the increase in multi-authorship is according to a game-theoretic analysis a consequence of the way scientists are evaluated. Scientists are judged by the number of papers they publish, and by the impact of those papers. Both measures are integrated into the most popular single value measure $h$ -index. The $h$ -index correlates with winning the Nobel Prize, being accepted for research fellowships and holding positions at top universities. When each author claims each paper and each citation as his/her own, papers and citations are multiplied by the number of authors. Since it is common and rational to cite own papers more than others, a high number of coauthors increases not only the number of own papers, but also their impact. As result, game rules set by $h$ -index being a decision criterion for success create a zero-sum $h$ -index ranking game, where the rational strategy includes maximizing the number of coauthors up to the majority of the researchers in a field. Data of 189 thousand publications showed that the coauthors' number is strongly correlated with $h$ -index. Hence, the system rewards heavily multi-authored papers. This problem is openly acknowledged, and it could easily be "corrected" by dividing each paper and its citations by the number of authors, though this practice has not been widely adopted.

Finally, the rise in shared authorship may also reflect increased acknowledgment of the contributions of lower level workers, including graduate students and technicians, as well as honorary authorship, while allowing for such collaborations to make an independent statement about the quality and integrity of a scientific work.

Order of authors in a list

Rules for the order of multiple authors in a list have historically varied significantly between fields of research. Some fields list authors in order of their degree of involvement in the work, with the most active contributors listed first; other fields, such as mathematics or engineering, sometimes list them alphabetically. Historically, biologists tended to place a principal investigator (supervisor or lab head) last in an author list whereas organic chemists might have put him or her first. Research articles in high energy physics, where the author lists can number in the tens to hundreds, often list authors alphabetically. In the academic fields of economics, business, finance or particle physics, it is also usual to sort the authors alphabetically.

Although listing authors in order of the involvement in the project seems straightforward, it often leads to conflict. A study in the Canadian Medical Association Journal found that more than two-thirds of 919 corresponding authors disagreed with their coauthors regarding contributions of each author.

Responsibilities of authors

Authors' reputations can be damaged if their names appear on a paper that they do not completely understand or with which they were not intimately involved. Numerous guidelines and customs specify that all co-authors must be able to understand and support a paper's major points.

In a notable case, American stem-cell researcher Gerald Schatten had his name listed on a paper co-authored with Hwang Woo-suk. The paper was later exposed as fraudulent and, though Schatten was not accused of participating in the fraud, a panel at his university found that "his failure to more closely oversee research with his name on it does make him guilty of 'research misbehavior.'"

All authors, including co-authors, are usually expected to have made reasonable attempts to check findings submitted for publication. In some cases, co-authors of faked research have been accused of inappropriate behavior or research misconduct for failing to verify reports authored by others or by a commercial sponsor. Examples include the case of Professor Geoffrey Chamberlain named as guest author of papers fabricated by Malcolm Pearce, (Chamberlain was exonerated from collusion in Pearce's deception) and the co-authors of Jan Hendrik Schön at Bell Laboratories. More recent cases include Charles Nemeroff, former editor-in-chief of Neuropsychopharmacology, and the so-called Sheffield Actonel affair.

Additionally, authors are expected to keep all study data for later examination even after publication. Both scientific and academic censure can result from a failure to keep primary data; the case of Ranjit Chandra of Memorial University of Newfoundland provides an example of this. Many scientific journals also require that authors provide information to allow readers to determine whether the authors may have commercial or non-commercial conflicts of interest. Outlined in the author disclosure statement for the American Journal of Human Biology, this is a policy more common in scientific fields where funding often comes from corporate sources. Authors are also commonly required to provide information about ethical aspects of research, particularly where research involves human or animal participants or use of biological material. Provision of incorrect information to journals may be regarded as misconduct. Financial pressures on universities have encouraged this type of misconduct. The majority of recent cases of alleged misconduct involving undisclosed conflicts of interest or failure of the authors to have seen scientific data involve collaborative research between scientists and biotechnology companies.

Unconventional types of authorship

Honorary authorship

Honorary authorship is sometimes granted to those who played no significant role in the work, for a variety of reasons. Until recently, it was standard to list the head of a German department or institution as an author on a paper regardless of input. The United States National Academy of Sciences, however, warns that such practices "dilute the credit due the people who actually did the work, inflate the credentials of those so 'honored,' and make the proper attribution of credit more difficult." The extent to which honorary authorship still occurs is not empirically known. However, it is plausible to expect that it is still widespread, because senior scientists leading large research groups can receive much of their reputation from a long publication list and thus have little motivation to give up honorary authorships.

A possible measure against honorary authorships has been implemented by some scientific journals, in particular by the Nature journals. They demand that each new manuscript must include a statement of responsibility that specifies the contribution of every author. The level of detail varies between the disciplines. Senior persons may still make some vague claim to have "supervised the project", for example, even if they were only in the formal position of a supervisor without having delivered concrete contributions. (The truth content of such statements is usually not checked by independent persons.) However, the need to describe contributions can at least be expected to somewhat reduce honorary authorships. In addition, it may help to identify the perpetrator in a case of scientific fraud.

Gift, guest and rolling authorship

More specific types of honorary authorship are gift, guest and rolling authorship. Gift authorship consists of authorship obtained by the offer of another author (honorary or not) with objectives that are beyond the research article itself or are ulterior, as promotion or favor. Guest authors are those that are included with the specific objective to increase the probability that it becomes accepted by a journal. A rolling authorship is a special case of gift authorship in which the honor is granted on the basis of previous research papers (published or not) and collaborations within the same research group. The "rolled" author may (or may not) be imposed by a superior employee for reasons that range from the research group's strategic interests, personal career interests, camaraderie or (professional) concession. For instance, a post-doc researcher in the same research group where his PhD was awarded, may be willing to roll his authorship into any subsequent paper from other researchers in that same group, overseeing the criteria for authorship. Per se, this would not cause authorship issues unless the collaboration was imposed by a third party, like a supervisor or department manager, in which case it is called a coercive authorship. Still, omitting the authorship criteria by prioritizing hierarchy arguments, is an unethical practice. This kind of practices may hinder free-thinking and professional independence, and thus should be tackled by research managers, clear research guidelines and authors agreements.

Ghost authorship

Ghost authorship occurs when an individual makes a substantial contribution to the research or the writing of the report, but is not listed as an author. Researchers, statisticians and writers (e.g. medical writers or technical writers) become ghost authors when they meet authorship criteria but are not named as an author. Writers who work in this capacity are called ghostwriters.

Ghost authorship has been linked to partnerships between industry and higher education. Two-thirds of industry-initiated randomized trials may have evidence of ghost authorship. Ghost authorship is considered problematic because it may be used to obscure the participation of researchers with conflicts of interest.

Litigation against the pharmaceutical company, Merck over health concerns related to use of their drug, Rofecoxib (brand name Vioxx), revealed examples of ghost authorship. Merck routinely paid medical writing companies to prepare journal manuscripts, and subsequently recruited external, academically affiliated researchers to pose as the authors.

Authors are sometimes included in a list without their permission. Even if this is done with the benign intention to acknowledge some contributions, it is problematic since authors carry responsibility for correctness and thus need to have the opportunity to check the manuscript and possibly demand changes.

Fraudulent paid-for authorship

Researchers can pay to intentionally and dishonestly list themselves as authors on papers they have not contributed to, usually by using an academic paper mill which specializes in authorship sales.

Anonymous and unclaimed authorship

Authors occasionally forgo claiming authorship, for a number of reasons. Historically some authors have published anonymously to shield themselves when presenting controversial claims. A key example is Robert Chambers' anonymous publication of Vestiges of the Natural History of Creation, a speculative, pre-Darwinian work on the origins of life and the cosmos. The book argued for an evolutionary view of life in the same spirit as the late Frenchman Jean-Baptiste Lamarck. Lamarck had long been discredited among intellectuals by this time and evolutionary (or development) theories were exceedingly unpopular, except among the political radicals, materialists, and atheists – Chambers hoped to avoid Lamarck's fate.

In the 18th century, Émilie du Châtelet began her career as a scientific author by submitting a paper in an annual competition held by the French Academy of Sciences; papers in this competition were submitted anonymously. Initially presenting her work without claiming authorship allowed her to have her work judged by established scientists while avoiding the bias against women in the sciences. She did not win the competition, but eventually her paper was published alongside the winning submissions, under her real name.

Scientists and engineers working in corporate and military organizations are often restricted from publishing and claiming authorship of their work because their results are considered secret property of the organization that employs them. One notable example is that of William Sealy Gosset, who was forced to publish his work in statistics under the pseudonym "Student" due to his employment at the Guinness brewery. Another account describes the frustration of physicists working in nuclear weapons programs at the Lawrence Livermore Laboratory – years after making a discovery they would read of the same phenomenon being "discovered" by a physicist unaware of the original, secret discovery of the phenomenon.

Satoshi Nakamoto is a pseudonym of a still unknown author or authors' group behind a white paper about bitcoin.

Non-human authorship

Artificial intelligence systems have been credited with authorship on a handful of academic publications, however, many publishers disallow this on the grounds that "they cannot take responsibility for the content and integrity of scientific papers".

Conflicts of interest in academic publishing

From Wikipedia, the free encyclopedia

Conflicts of interest (COIs) often arise in academic publishing. Such conflicts may cause wrongdoing and make it more likely. Ethical standards in academic publishing exist to avoid and deal with conflicts of interest, and the field continues to develop new standards. Standards vary between journals and are unevenly applied. According to the International Committee of Medical Journal Editors, "[a]uthors have a responsibility to evaluate the integrity, history, practices and reputation of the journals to which they submit manuscripts".

Conflicts of interest increase the likelihood of biases arising; they can harm the quality of research and the public good (even if disclosed). Conflicts of interest can involve research sponsors, authors, journals, journal staff, publishers, and peer reviewers.

Avoidance, disclosure, and tracking

The avoidance of conflicts of interest and the changing of the structure of institutions to make them easier to avoid are frequently advocated for. Some institutional ethics policies ban academics from entering into specific types of COIs, for instance by prohibiting them from accepting gifts from companies connected with their work. Education in ethical COI management is also a tool for avoiding COI problems.

Disclosure of COIs has been debated since the 1980s; there is a general consensus favouring disclosure. There is also a view that COI concerns and some of the measures taken to reduce them are excessive.

Criticisms of disclosure policies include:

authors disclosing COIs may feel pressured to present their research in a more biased manner to compensate;
disclosure may discourage beneficial academic–industrial collaboration;
disclosure may decrease public trust in research;
researchers who have disclosed their COIs may feel license to behave immorally;
disclosure may be taken as a sign of honesty or expertise and thus increase trust;
some types of COI may be more likely than others to go unnoticed or unreported;
awareness of a COI does not make people immune to being influenced by bias; generally, people do not sufficiently discount biased advice;
disclosure discourages the judging of work purely on its merits;
disclosure causes more intense scrutiny for wrongdoing.

While disclosure is widely favoured, other COI management measures have narrower support. Some publications hold the opinion that certain COIs disqualify people from certain research roles; for instance, that the testing of medicines should be done only by people who neither develop medicines nor are funded by their manufacturers.

Conflicts of interest have also been considered as a statistical factor confounding evidence, which must therefore be measured as accurately as possible and analysed, requiring machine-readable disclosure.

Codes of conduct

Journals have individual ethics policies and codes of conduct; there are also some cross-journal voluntary standards.

The International Committee of Medical Journal Editors (ICMJE) publishes Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly work in Medical Journals, and a list of journals that pledge to follow it. The guideline lays down detailed rules for conflict-of-interest declaration by authors. It also says; "All participants in the peer-review and publication process—not only authors but also peer reviewers, editors, and editorial board members of journals—must consider their conflicts of interest when fulfilling their roles in the process of article review and publication and must disclose all relationships that could be viewed as potential conflicts of interest". These recommendations have been criticized and revised to remove loopholes allowing the non-disclosure of conflicts of interest.

The Council of Science Editors publishes a White Paper on publication ethics. Citing the ICMJE that "all participants in the peer-review and publication process must disclose all relationships that could be viewed as potential conflicts of interest", it highly recommends COI disclosure for sponsors, authors, reviewers, journals, and editorial staff.

The Good Publication Practice (GPP) guidelines, covering industry-sponsored medical research, are published by the International Society of Medical Publication Professionals.

The Committee on Publication Ethics (COPE) publishes a code of conduct stating, "[t]here must be clear definitions of conflicts of interest and processes for handling conflicts of interest of authors, reviewers, editors, journals and publishers, whether identified before or after publication".

The Open Access Scholarly Publishers Association's Principles of Transparency and Best Practice in Scholarly Publishing is intended to separate legitimate journals from predatory publishers and defines a minimal standard; clear and clearly stated COI policies.

A 2009 US Institute of Medicine report on medical COIs states that conflict-of-interest policies should be judged on their proportionality, transparency, accountability, and fairness; they should be effective, efficient, and targeted, known and understood, clearly identify who is responsible for monitoring, enforcement, and amendment, and apply equally to everyone involved. Review by conflict-of-interest committees is also recommended, and the lack of transparency and COI declaration in developing COI guidelines criticized.

As of 2015, journal COI policies often have no enforcement provisions. COI disclosure obligations have been legislated; one example of such legislation is the US Physician Payments Sunshine Act, but these laws do not apply specifically to journals.

COIs by agent

COIs of journals

Journals are often not transparent about their institutional COIs, and do not apply the same disclosure standards to themselves as they do to their authors. Four out of six major general medical journals that were contacted for a 2010 COI study refused to provide information about the proportion of their income that derived from advertisements, reprints, and industry-supported supplements, citing policies on non-disclosure of financial information.

Owners and governing bodies

The owner of an academic journal has ultimate power over the hiring and firing of editorial staff; editors' interests in pleasing their employers conflict with some of their other editorial interests.Journals are also more likely to accept papers by authors who work for the journals' hosting institutions.

Some journals are owned by publishers. When journals print reviews of books published by their own publishers, they rarely (as of 2013) add COI disclosures. The publishers' interest in maximizing profit will often conflict with academic interests or ethical standards. In the case of closed-access publications, publishers' desire for high subscription income may conflict with an editorial desire for broader access and readership. There have been multiple mass resignations of editorial boards over such conflicts, which are often followed by the editorial board founding a new, non-profit journal to compete with their former one.

Some journals are owned by academic societies and professional organisations. Leading journals can be very profitable and there is often friction about revenue between the journal and the member society that owns it. Some academic societies and professional organisations are themselves funded by membership fees and/or donations. If the owners benefit financially from donations, the journal has a conflict between its financial interest in satisfying the donors—and therefore the owners—and its journalistic interests. Such COIs with industry donors have drawn criticism.

Reprints

A reprint is a copy of an individual article that is printed and sold as a separate product by the journal or its publisher or agent. Reprints are often used in pharmaceutical marketing and other medical marketing of products to doctors. This gives journals an incentive to produce good marketing material.Journals sell reprints at very high profit margins, often around 70%, as of 2010. A journal may sell a million dollars worth of reprints of a single article if, for example, it is a large industry-funded clinical trial. The selling of reprints can bring in over 40% of a journal's income.

Impact factors, reputation, and subscriptions

If a journal is accused of managing COIs badly, its reputation is harmed.

The impact factor of a journal is often used to rate it, although this practice is widely criticized. A journal will generally want to increase its impact factor in hope of gaining more subscriptions, better submissions, and more prestige. As of 2010, industry-funded papers generally get cited more than others; this is probably due in part to industry-paid publicity.

Some journals engage in coercive citation, in which an editor forces an author to add extraneous citations to an article to inflate the impact factor of the journal in which the extraneous papers were published. A survey found that 86% of academics consider coercive citation unethical but 20% have experienced it. Journals appear to preferentially target younger authors and authors from non-English-speaking countries. Journals published by for-profit companies used coercive citation more than those published by university presses.

Journals may find it difficult to correct and retract erroneous papers after publication because of legal threats.

Advertising

Many academic journals contain advertising. The portion of a journal's revenue coming from advertising varies widely, according to one small study, from over 50% to 1%. As of 2010, advertising revenues for academic journals are generally falling. A 1995 survey of North American journal editors found that 57% felt responsible for the honesty of the pharmaceutical advertisements they ran and 40% supported peer-review of such advertisements. An interest in increasing advertising revenue can conflict with interests in journalistic independence and truthfulness.

Publishers

Publishers may not be strongly motivated to ensure the quality of their journals. In the Australasian Journal of Bone & Joint Medicine case, the printer Elsevier Australia put out six journal-like publications containing articles about drugs made by the Merck Group, which paid for and controlled the publications.

COIs of journal staff

Personal conflicts of interest faced by journal staff are individual. If a person leaves the journal—unlike the COIs of journals as institutions—their personal COIs will go with them.

As of 2015, COIs of journal staff are less commonly reported than those of authors. For instance, one 2009 World Association of Medical Editors (WAME) policy document states, "Some journals list editors' competing interests on their website but this is not a standard practice". The ICMJE, however, requires that the COIs of editors and journal staff be regularly declared and published.

One 2017 Open Payments study of influential US medical journals found half of the editors received payments from industry; another study that used a different sample of editors reported two-thirds. As of 2002, systems for reporting wrongdoing by editors often do not exist.

Many journals have policies limiting COIs staff can enter into; for instance, accepting gifts of travel, accommodation, or hospitality may be prohibited. As of 2016, such policies are rarely published. Most journals do not offer COI training; as of 2015, many journals report a desire for better guidance on COI policy.

COIs of peer reviewers

The ICJME recommendations require peer reviewers to disclose conflicts of interest. Half to two-thirds of journals, depending on subject area, did not follow this recommendation in the first two decades of the 21st century. As of 2017, if a peer reviewer fails to disclose a conflict of interest, the paper will generally not be withdrawn, corrected, or re-reviewed; the reviews, however, may be reassessed.

If peer reviewers are anonymous, their COIs cannot be published. Some experiments with publishing the names of reviewers have been undertaken; in others, the identities of reviewers were disclosed to authors, allowing authors to identify COIs. Some journals now have an open review process in which everything, including the peer reviews and the names of the reviewers, and editor and author comment, is published transparently online.

The duties of peer review may conflict with social interests or institutional loyalties; to avoid such COIs, reviewers may be excluded if they have some forms of COI, such as having collaborated with the author.

Readers of academic papers may spot errors, informally or as part of formal post-publication peer review. Academics submitting corrections to papers are often asked by the publishers to pay over 1,000 US dollars for the publication of their corrections.

COIs of article authors

Authors of individual papers may face conflicts with their duty to report truthfully and impartially. Financial, career, political, and social interests are all sources of conflict. Authors' institutional interests become sources of conflict when the research might harm the institution's finances or offend the author's superiors.

Many journals require authors to self-declare their conflicts of interest when submitting a paper; they also ask specific questions about conflicts of interest. The questions vary substantially between journals. Author declarations, however, are rarely verified by the journal. As of 2018, "most editors say it's not their job to make sure authors reveal financial conflicts, and there are no repercussions for those who don't". Even if a conflict of interest is reported by a reader after publication, COPE does not suggest independent investigation, as of 2017.

As a result, as of 2018, authors often fail to declare their conflicts of interest. Rates of nondisclosure vary widely in reported studies.

The COPE retraction guidelines state, "Retractions are also used to alert readers to ... failure to disclose a major competing interest likely to influence interpretations or recommendations". As of 2018, however, if an author fails to disclose a COI, the paper will usually be corrected; it will not usually be retracted. Paper retractions, notifications to superiors, and publication bans are possible. Non-disclosure incidents harm academic careers. Authors are held to have collective responsibility for the contents of an article; if one author fails to declare a conflict of interest, the peer review process may be deemed compromised and the whole paper retracted.

The publisher may charge authors substantial fees for retracting papers, even in cases of honest error, giving them a financial disincentive to correct the record.

Public registries of author COIs have been suggested. Authors face administrative burdens in declaring COIs; standardized declarations or a registry could reduce these.

Ghost authors and non-contributing authors

Ghost authorship, where a writer contributes but is not credited, has been estimated to affect a significant proportion of the research literature. Honorary authorship, where an author is credited but did not contribute, is more common. Being named as an author on many papers is good for an academic's career. Failure to adhere to authorship standards is rarely punished. To avoid misreported authorship, a requirement that all authors describe the contribution they made to the study ("movie-style credits") has been advocated for. Ghostwriters may be legally liable for fraud.

The ICMJE criteria for authorship require that authors contribute:

Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; and
Drafting the work or revising it critically for important intellectual content; and
Final approval of the version to be published; and
Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

The ICMJE requires that "All those designated as authors should meet all four criteria for authorship, and all who meet the four criteria should be identified as authors. Those who do not meet all four criteria should be acknowledged." Academics who have had publication ethics training and those who are aware of the ICMJE authorship criteria are more stringent in their concepts of authorship and are more likely to consider breaches of authorship as misconduct, as are more junior researchers. Awareness is low; one study found only about half of researchers had read the ICJME criteria.

COIs of study sponsors

If a study requires outside funding, this can be a major source of conflicting interests; for instance in cases where the manufacturer of a drug is funding a study into its safety and efficacy or where the sponsor hopes to use the research to defend itself in litigation. Sponsors of a study may involve themselves in the design, execution, analysis, and write-up of a study. In extreme cases, they may carry out the research and ghostwrite the article with almost no involvement from the nominal author.Movie-style credits are advocated as a way to avoid this.

There are many opportunities for bias in trial design and trial reporting. For instance, a trial that compares a drug against the wrong dose of a competing drug may produce spuriously positive results.

In some cases, a contract with a sponsor may mean those named as investigators and authors on the papers may not have access to the trial data, control over the publication text, or the freedom to talk about their work. While authors and institutions have an interest in avoiding such contracts, it conflicts with their interest in competing for funding from potential study sponsors. Institutions that set stricter ethical standards for sponsor contracts lose contracts and funding when sponsors go elsewhere.

Sponsors have required contractual promises that the study is not reported without the sponsor's approval (gag clauses) and some have sued authors over compliance. Trials may go unpublished to keep commercial information secret or because the trial results were unfavourable. Some journals require that human trials be registered to be considered for publication; some require the declaration of any gag clauses as a conflict of interest; since 2001, some also require a statement that the authors have not agreed to a gag clause. Some journals require a promise to provide access to the original data to researchers intending to replicate the work. Some research ethics boards, universities, and national laws prohibit gag clauses. Gag clauses may not be legally enforceable if compliance would cause sufficient public harm. Non-publication has been found to be more common in industry-funded trials, contributing to publication bias.

It has been suggested that having many sponsors with different interests protects against COI-induced bias. As of 2006, there was no evidence for or against this hypothesis.

Effect on conclusions of research

There is evidence that industry funding of studies of medical devices and drugs results in these studies having more positive conclusions regarding efficacy (funding bias). A similar relationship has been found in clinical trials of surgical interventions, where industry funding leads to researchers exaggerating the positive nature of their findings. Not all studies have found a statistically significant relationship between industry funding and the study outcome.

Interests of research participants

Chronically ill medical research participants report expectation of being told about COIs and some report they would not participate if the researcher had some sorts of COIs. With few exceptions, multiple ethical guidelines forbid researchers with a financial interest in the outcome from being involved in human trials.

The consent agreements entered into with study participants may be legally binding on the academics but not on the sponsor, unless the sponsor has a contractual commitment saying otherwise.

Ethical rules, including the Declaration of Helsinki, require the publication of results of human trials. participants in which are often motivated by a desire to improve medical knowledge. Patients may be harmed if safety data, such risks to patients, are kept secret. Duties to human-research participants can therefore conflict with interests in non-publication such as gag clauses.

Publication of COI declarations

Some journals place COI declarations at the beginning of an article but most put it in smaller print at the end. Positioning makes a difference; if readers feel they are being manipulated from the beginning of a text, they read more critically than if the same feeling is produced at the end of a text.

According to the ICMJE, "each journal should develop standards with regard to the form the [COI] information should take and where it will be posted". It is often placed after the body of the article, just before the reference section. Some COI statements, like those of anonymous reviewers, may not be published at all. (see § COIs of peer reviewers) COI statements are sometimes paywalled so they are only visible to those who have paid for full text access.This is not considered ethical by the Committee on Publication Ethics.

In 2017 PubMed began including COI statements at the end of the abstract and before the body of the article^[73] after receiving complaints that because COI declarations were only included in full article texts, they often went unseen in paywalled articles. Only COI statements that are appropriately formatted and tagged by the publisher are included.

Science journalism rarely reports COI information from the academic article reported upon; in some studies, fewer than 1% of stories included COI information.

False statements of COIs

Failure to disclose a conflict of interest may, depending on the circumstances, be considered a form of corruption or academic misconduct.

Publication bias

From Wikipedia, the free encyclopedia

In published academic research, publication bias occurs when the outcome of an experiment or research study biases the decision to publish or otherwise distribute it. Publishing only results that show a significant finding disturbs the balance of findings in favor of positive results. The study of publication bias is an important topic in metascience.

Despite similar quality of execution and design, papers with statistically significant results are three times more likely to be published than those with null results. This unduly motivates researchers to manipulate their practices to ensure statistically significant results, such as by data dredging.

Many factors contribute to publication bias. For instance, once a scientific finding is well established, it may become newsworthy to publish reliable papers that fail to reject the null hypothesis. Most commonly, investigators simply decline to submit results, leading to non-response bias. Investigators may also assume they made a mistake, find that the null result fails to support a known finding, lose interest in the topic, or anticipate that others will be uninterested in the null results. The nature of these issues and the resulting problems form the five diseases that threaten science: "significosis, an inordinate focus on statistically significant results; neophilia, an excessive appreciation for novelty; theorrhea, a mania for new theory; arigorium, a deficiency of rigor in theoretical and empirical work; and finally, disjunctivitis, a proclivity to produce many redundant, trivial, and incoherent works."

Attempts to find unpublished studies often prove difficult or are unsatisfactory. In an effort to combat this problem, some journals require studies submitted for publication pre-register (before data collection and analysis) with organizations like the Center for Open Science.

Other proposed strategies to detect and control for publication bias include p-curve analysis and disfavoring small and non-randomized studies due to high susceptibility to error and bias.

Definition

Publication bias occurs when the publication of research results depends not just on the quality of the research but also on the hypothesis tested, and the significance and direction of effects detected. The subject was first discussed in 1959 by statistician Theodore Sterling to refer to fields in which "successful" research is more likely to be published. As a result, "the literature of such a field consists in substantial part of false conclusions resulting from errors of the first kind in statistical tests of significance". In the worst case, false conclusions could canonize as being true if the publication rate of negative results is too low.

One effect of publication bias is sometimes called the file-drawer effect, or file-drawer problem. This term suggests that negative results, those that do not support the initial hypotheses of researchers are often "filed away" and go no further than the researchers' file drawers, leading to a bias in published research. The term "file drawer problem" was coined by psychologist Robert Rosenthal in 1979.

Positive-results bias, a type of publication bias, occurs when authors are more likely to submit, or editors are more likely to accept, positive results than negative or inconclusive results. Outcome reporting bias occurs when multiple outcomes are measured and analyzed, but the reporting of these outcomes is dependent on the strength and direction of its results. A generic term coined to describe these post-hoc choices is HARKing ("Hypothesizing After the Results are Known").

Evidence

There is extensive meta-research on publication bias in the biomedical field. Investigators following clinical trials from the submission of their protocols to ethics committees (or regulatory authorities) until the publication of their results observed that those with positive results are more likely to be published. In addition, studies often fail to report negative results when published, as demonstrated by research comparing study protocols with published articles.

The presence of publication bias was investigated in meta-analyses. The largest such analysis investigated the presence of publication bias in systematic reviews of medical treatments from the Cochrane Library. The study showed that statistically positive significant findings are 27% more likely to be included in meta-analyses of efficacy than other findings. Results showing no evidence of adverse effects have a 78% greater probability of inclusion in safety studies than statistically significant results showing adverse effects. Evidence of publication bias was found in meta-analyses published in prominent medical journals.

Meta-analyses (reviews) have been performed in the field of ecology and environmental biology. In a study of 100 meta-analyses in ecology, only 49% tested for publication bias. While there are multiple tests that have been developed to detect publication bias, most perform poorly in the field of ecology because of high levels of heterogeneity in the data and that often observations are not fully independent.

As of 1998, "No trial published in China or Russia/USSR found a test treatment to be ineffective."

Impact on meta-analysis

Where publication bias is present, published studies are no longer a representative sample of the available evidence. This bias distorts the results of meta-analyses and systematic reviews. For example, evidence-based medicine is increasingly reliant on meta-analysis to assess evidence.

Meta-analyses and systematic reviews can account for publication bias by including evidence from unpublished studies and the grey literature. The presence of publication bias can also be explored by constructing a funnel plot in which the estimate of the reported effect size is plotted against a measure of precision or sample size. The premise is that the scatter of points should reflect a funnel shape, indicating that the reporting of effect sizes is not related to their statistical significance. However, when small studies are predominately in one direction (usually the direction of larger effect sizes), asymmetry will ensue and this may be indicative of publication bias.

Because an inevitable degree of subjectivity exists in the interpretation of funnel plots, several tests have been proposed for detecting funnel plot asymmetry.These are often based on linear regression including the popular Eggers regression test, and may adopt a multiplicative or additive dispersion parameter to adjust for the presence of between-study heterogeneity. Some approaches may even attempt to compensate for the (potential) presence of publication bias, which is particularly useful to explore the potential impact on meta-analysis results.

In ecology and environmental biology, a study found that publication bias impacted the effect size, statistical power, and magnitude. The prevalence of publication bias distorted confidence in meta-analytic results, with 66% of initially statistically significant meta-analytic means becoming non-significant after correcting for publication bias. Ecological and evolutionary studies consistently had low statistical power (15%) with a 4-fold exaggeration of effects on average (Type M error rates = 4.4).

The presence of publication bias can be detected by Time-lag bias tests, where time-lag bias occurs when larger or statistically significant effects are published more quickly than smaller or non-statistically significant effects. It can manifest as a decline in the magnitude of the overall effect over time. The key feature of time-lag bias tests is that, as more studies accumulate, the mean effect size is expected to converge on its true value.

Compensation examples

Two meta-analyses of the efficacy of reboxetine as an antidepressant demonstrated attempts to detect publication bias in clinical trials. Based on positive trial data, reboxetine was originally passed as a treatment for depression in many countries in Europe and the UK in 2001 (though in practice it is rarely used for this indication). A 2010 meta-analysis concluded that reboxetine was ineffective and that the preponderance of positive-outcome trials reflected publication bias, mostly due to trials published by the drug manufacturer Pfizer. A subsequent meta-analysis published in 2011, based on the original data, found flaws in the 2010 analyses and suggested that the data indicated reboxetine was effective in severe depression (see Reboxetine § Efficacy). Examples of publication bias are given by Ben Goldacre and Peter Wilmshurst.

In the social sciences, a study of published papers exploring the relationship between corporate social and financial performance found that "in economics, finance, and accounting journals, the average correlations were only about half the magnitude of the findings published in Social Issues Management, Business Ethics, or Business and Society journals".

One example cited as an instance of publication bias is the refusal to publish attempted replications of Bem's work that claimed evidence for precognition by The Journal of Personality and Social Psychology (the original publisher of Bem's article).

An analysis comparing studies of gene-disease associations originating in China to those originating outside China found that those conducted within the country reported a stronger association and a more statistically significant result.

Risks

John Ioannidis argues that "claimed research findings may often be simply accurate measures of the prevailing bias." He lists the following factors as those that make a paper with a positive result more likely to enter the literature and suppress negative-result papers:

The studies conducted in a field have small sample sizes.
The effect sizes in a field tend to be smaller.
There is both a greater number and lesser preselection of tested relationships.
There is greater flexibility in designs, definitions, outcomes, and analytical modes.
There are prejudices (financial interest, political, or otherwise).
The scientific field is hot and there are more scientific teams pursuing publication.

Other factors include experimenter bias and white hat bias.

Remedies

Publication bias can be contained through better-powered studies, enhanced research standards, and careful consideration of true and non-true relationships. Better-powered studies refer to large studies that deliver definitive results or test major concepts and lead to low-bias meta-analysis. Enhanced research standards such as the pre-registration of protocols, the registration of data collections and adherence to established protocols are other techniques. To avoid false-positive results, the experimenter must consider the chances that they are testing a true or non-true relationship. This can be undertaken by properly assessing the false positive report probability based on the statistical power of the test and reconfirming (whenever ethically acceptable) established findings of prior studies known to have minimal bias.

Study registration

In September 2004, editors of prominent medical journals (including the New England Journal of Medicine, The Lancet, Annals of Internal Medicine, and JAMA) announced that they would no longer publish results of drug research sponsored by pharmaceutical companies, unless that research was registered in a public clinical trials registry database from the start. Furthermore, some journals (e.g. Trials), encourage publication of study protocols in their journals.

The World Health Organization (WHO) agreed that basic information about all clinical trials should be registered at the study's inception, and that this information should be publicly accessible through the WHO International Clinical Trials Registry Platform. Additionally, public availability of complete study protocols, alongside reports of trials, is becoming more common for studies.

Megastudies

In a megastudy, a large number of treatments are tested simultaneously. Given inclusion of different interventions in the study, a megastudy's publication likelihood is less dependent on the statistically significant effect of any specific treatment, so it has been suggested that megastudies may be less prone to publication bias. For example, an intervention found to be ineffective would be easier to publish as part of a megastudy as just one of many studied interventions, whereas it might go unreported due to the file-drawer problem if it were the sole focus of a contemplated paper. For the same reason, the megastudy research design may encourage researchers to study not only the interventions they consider more likely to be effective but also those interventions that researchers are less certain about and that they would not pick as the sole focus of the study due to the perceived high risk of a null effect.

Stereotype threat

From Wikipedia, the free encyclopedia

Stereotype threat is a situational predicament in which people are or feel themselves to be at risk of conforming to stereotypes about their social group. It is theorized to be a contributing factor to long-standing racial and gender gaps in academic performance. Since its introduction into the academic literature, stereotype threat has become one of the most widely studied topics in the field of social psychology.

Situational factors that increase stereotype threat can include the difficulty of the task, the belief that the task measures their abilities, and the relevance of the stereotype to the task. Individuals show higher degrees of stereotype threat on tasks they wish to perform well on and when they identify strongly with the stereotyped group. These effects are also increased when they expect discrimination due to their identification with a negatively stereotyped group. Repeated experiences of stereotype threat can lead to a vicious circle of diminished confidence, poor performance, and loss of interest in the relevant area of achievement. Stereotype threat has been argued to show a reduction in the performance of individuals who belong to negatively stereotyped groups. Its role in affecting public health disparities has also been suggested.

According to the theory, if negative stereotypes are present regarding a specific group, group members are likely to become anxious about their performance, which may hinder their ability to perform to their full potential. Importantly, the individual does not need to subscribe to the stereotype for it to be activated. It is hypothesized that the mechanism through which anxiety (induced by the activation of the stereotype) decreases performance is by depleting working memory (especially the phonological aspects of the working memory system).

The opposite of stereotype threat is stereotype boost, which is when people perform better than they otherwise would have, because of exposure to positive stereotypes about their social group. A variant of stereotype boost is stereotype lift, which is people achieving better performance because of exposure to negative stereotypes about other social groups.

Some researchers have suggested that stereotype threat should not be interpreted as a factor in real-life performance gaps, and have raised the possibility of publication bias. Other critics have focused on correcting what they claim are misconceptions of early studies showing a large effect. Meta-analyses and systematic reviews have shown significant evidence for the effects of stereotype threat, though the phenomenon defies over-simplistic characterization.

Empirical studies

As of 2015, more than 300 studies have been published showing the effects of stereotype threat on performance in a variety of domains. Stereotype threat is considered by some researchers to be a contributing factor to long-standing racial and gender achievement gaps, such as under-performance of black students relative to white ones in various academic subjects, and under-representation of women at higher echelons in the field of mathematics.

The strength of the stereotype threat that occurs depends on how the task is framed. If a task is framed to be neutral, stereotype threat is not likely to occur; however, if tasks are framed in terms of active stereotypes, participants are likely to perform worse on the task. For example, a study on chess players revealed that female players performed more poorly than expected when they were told they would be playing against a male opponent. In contrast, women who were told that their opponent was female performed as would be predicted by past ratings of performance. Female participants who were made aware of the stereotype of females performing worse at chess than males performed worse in their chess games.

A 2007 study extended stereotype threat research to entrepreneurship, a traditionally male-stereotyped profession. The study revealed that stereotype threat can depress women's entrepreneurial intentions while boosting men's intentions. However, when entrepreneurship is presented as a gender-neutral profession, men and women express a similar level of interest in becoming entrepreneurs. Another experiment involved a golf game which was described as a test of "natural athletic ability" or of "sports intelligence". When it was described as a test of athletic ability, European-American students performed worse, but when the description mentioned intelligence, African-American students performed worse.

Other studies have demonstrated how stereotype threat can negatively affect the performance of European Americans in athletic situations as well as the performance of men who are being tested on their social sensitivity. Although the framing of a task can produce stereotype threat in most individuals, certain individuals appear to be more likely to experience stereotype threat than others. Individuals who highly identify with a particular group appear to be more vulnerable to experiencing stereotype threat than individuals who do not identify strongly with the stereotyped group.

The mere presence of other people can evoke stereotype threat. In one experiment, women who took a mathematics exam along with two other women got 70% of the answers right, whereas women who took the same exam in the presence of two men got an average score of 55%.

The goal of a study conducted by Desert, Preaux, and Jund in 2009 was to see if children from lower socioeconomic groups are affected by stereotype threat. The study compared children that were 6–7 years old with children that were 8–9 years old from multiple elementary schools. These children were presented with the Raven's Matrices test, which is an intellectual ability test. Separate groups of children were given directions in an evaluative way and other groups were given directions in a non-evaluative way. The "evaluative" group received instructions that are usually given with the Raven Matrices test, while the "non-evaluative" group was given directions which made it seem as if the children were simply playing a game. The results showed that third graders performed better on the test than the first graders did, which was expected. However, the lower socioeconomic status children did worse on the test when they received directions in an evaluative way than the higher socioeconomic status children did when they received directions in an evaluative way. These results suggested that the framing of the directions given to the children may have a greater effect on performance than socioeconomic status. This was shown by the differences in performance based on which type of instructions they received. This information can be useful in classroom settings to help improve the performance of students of lower socioeconomic status.

There have been studies on the effects of stereotype threat based on age. A study was done on 99 senior citizens ranging in age from 60–75 years. These seniors were given multiple tests on certain factors and categories such as memory and physical abilities, and were also asked to evaluate how physically fit they believe themselves to be. Additionally, they were asked to read articles that contained both positive and negative outlooks about seniors, and they watched someone reading the same articles. The goal of this study was to see if priming the participants before the tests would affect performance. The results showed that the control group performed better than those that were primed with either negative or positive words prior to the tests. The control group seemed to feel more confident in their abilities than the other two groups. Other studies have found that stereotype activation in older adults can improve memory performance, resulting in a distinction between stereotype threat mechanisms in aging compared with other groups.

Many psychological experiments carried out on Stereotype Threat focus on the physiological effects of negative stereotype threat on performance, looking at both high and low status groups. Scheepers and Ellemers tested the following hypothesis: when assessing a performance situation on the basis of current beliefs the low status group members would show a physiological threat response, and high-status members would also show a physiological threat response when examining a possible alteration of the status quo (Scheepers & Ellemers, 2005). The results of this experiment were in line with expectations. As predicted, participants in the low status condition showed higher blood pressure immediately after the status feedback, while participants in the high-status condition showed a spike in blood pressure while anticipating the second round of the task.

In 2012, Scheepers et al. hypothesized that when high social power is stimulated 'an efficient cardiovascular pattern (challenge)' is produced, whereas, 'an inefficient cardiovascular pattern' or threat is caused by the activation of low social power (Scheepers, de Wit, Ellemers & Sassenberg, 2012). Two experiments were carried out in order to test this hypothesis. The first experiment looked at power priming and the second experiment related to role play. Both results from these two experiments provided evidence in support for the hypothesis.

Cleopatra Abdou and Adam Fingerhut were the first to develop experimental methods to study stereotype threat in a health care context, including the first study indicating that health care stereotype threat is linked with adverse health outcomes and disparities.

Some studies have found null results. The single largest experimental test of stereotype threat (N = 2064), conducted on Dutch high school students, found no effect. The authors state, however, that these results are limited to a narrow age-range, experimental procedure and cultural context, and call for further registered reports and replication studies on the topic. Despite these limitations, they state in conclusion that their study shows "that the effects of stereotype threat on math test performance should not be overgeneralized."

Numerous meta-analyses and systematic reviews have shown significant evidence for the effects of stereotype threat.However they also point to ways in which the phenomenon defies over-simplistic characterization. For instance, one meta-analysis found that with female subjects "subtle threat-activating cues produced the largest effect, followed by blatant and moderately explicit cues" while with minorities "moderately explicit stereotype threat-activating cues produced the largest effect, followed by blatant and subtle cues".

Mechanisms

Although numerous studies demonstrate the effects of stereotype threat on performance, questions remain as to the specific cognitive factors that underlie these effects. Steele and Aronson originally speculated that attempts to suppress stereotype-related thoughts lead to anxiety and the narrowing of attention. This could contribute to the observed deficits in performance. In 2008, Toni Schmader, Michael Johns, and Chad Forbes published an integrated model of stereotype threat that focused on three interrelated factors:

stress arousal;
performance monitoring, which narrows attention; and,
efforts to suppress negative thoughts and emotions.

Schmader et al. suggest that these three factors summarize the pattern of evidence that has been accumulated by past experiments on stereotype threat. For example, stereotype threat has been shown to disrupt working memory and executive function, increase arousal, increase self-consciousness about one's performance, and cause individuals to try to suppress negative thoughts as well as negative emotions such as anxiety. People have a limited amount of cognitive resources available. When a large portion of these resources are spent focusing on anxiety and performance pressure, the individual is likely to perform worse on the task at hand.

A number of studies looking at physiological and neurological responses support Schmader and colleagues' integrated model of the processes that produce stereotype threat. Supporting an explanation in terms of stress arousal, one study found that African Americans under stereotype threat exhibit larger increases in arterial blood pressure. One study found increased cardiovascular activation amongst women who watched a video in which men outnumbered women at a math and science conference. Other studies have similarly found that individuals under stereotype threat display increased heart rates. Stereotype threat may also activate a neuroendocrine stress response, as measured by increased levels of cortisol while under threat. The physiological reactions that are induced by stereotype threat can often be subconscious, and can distract and interrupt cognitive focus from the task.

With regard to performance monitoring and vigilance, studies of brain activity have supported the idea that stereotype threat increases both of these processes. Forbes and colleagues recorded electroencephalogram (EEG) signals that measure electrical activity along the scalp, and found that individuals experiencing stereotype threat were more vigilant for performance-related stimuli.

Another study used functional magnetic resonance imaging (fMRI) to investigate brain activity associated with stereotype threat. The researchers found that women experiencing stereotype threat while taking a math test showed heightened activation in the ventral stream of the anterior cingulate cortex (ACC), a neural region thought to be associated with social and emotional processing. Wraga and colleagues found that women under stereotype threat showed increased activation in the ventral ACC and that the amount of this activation predicted performance decrements on the task. When individuals were made aware of performance-related stimuli, they were more likely to experience stereotype threat. However, a study using fMRI to investigate stereotype threat in older adults showed heightened activation in parietal midline regions including the posterior cingulate cortex (PCC) and precuneus during both working memory and episodic memory tasks. The heightened activation in these brain areas also was associated with better memory accuracy, inconsistent with the notion that stereotype threat always leads to impaired performance.

A study conducted by Boucher, Rydell, Loo, and Rydell has shown that stereotype threat not only affects performance, but can also affect the ability to learn new information. In the study, undergraduate men and women had a session of learning followed by an assessment of what they learned. Some participants were given information intended to induce stereotype threat, and some of these participants were later given "gender fair" information, which it was predicted would reduce or remove stereotype threat. As a result, participants were split into four separate conditions: control group, stereotype threat only, stereotype threat removed before learning, and stereotype threat removed after learning. The results of the study showed that the women who were presented with the "gender fair" information performed better on the math related test than the women who were not presented with this information. This study also showed that it was more beneficial to women for the "gender fair" information to be presented prior to learning rather than after learning. These results suggest that eliminating stereotype threat prior to taking mathematical tests can help women perform better, and that eliminating stereotype threat prior to mathematical learning can help women learn better.

Original study

In 1995, Claude Steele and Joshua Aronson performed the first experiments demonstrating that stereotype threat can undermine intellectual performance. Steele and Aronson measured this through a word completion task.

They had African-American and European-American college students take a difficult verbal portion of the Graduate Record Examination test. As would be expected based on national averages, the African-American students did not perform as well on the test. Steele and Aronson split students into three groups: stereotype-threat (in which the test was described as being "diagnostic of intellectual ability"), non-stereotype threat (in which the test was described as "a laboratory problem-solving task that was nondiagnostic of ability"), and a third condition (in which the test was again described as nondiagnostic of ability, but participants were asked to view the difficult test as a challenge). All three groups received the same test.

Steele and Aronson concluded that changing the instructions on the test could reduce African-American students' concern about confirming a negative stereotype about their group. Supporting this conclusion, they found that African-American students who regarded the test as a measure of intelligence had more thoughts related to negative stereotypes of their group. Additionally, they found that African Americans who thought the test measured intelligence were more likely to complete word fragments using words associated with relevant negative stereotypes (e.g., completing "__mb" as "dumb" rather than as "numb").

Adjusted for previous SAT scores, subjects in the non-diagnostic-challenge condition performed significantly better than those in the non-diagnostic-only condition and those in the diagnostic condition. In the first experiment, the race-by-condition interaction was marginally significant. However, the second study reported in the same paper found a significant interaction effect of race and condition. This suggested that placement in the diagnostic condition significantly impacted African Americans compared with European Americans.

Stereotype lift and stereotype boost

Stereotype threat concerns how stereotype cues can harm performance. However, in certain situations, stereotype activation can also lead to performance enhancement through stereotype lift or stereotype boost. Stereotype lift increases performance when people are exposed to negative stereotypes about another group. This enhanced performance has been attributed to increases in self-efficacy and decreases in self-doubt as a result of negative outgroup stereotypes. Stereotype boost suggests that positive stereotypes may enhance performance. Stereotype boost occurs when a positive aspect of an individual's social identity is made salient in an identity-relevant domain. Although stereotype boost is similar to stereotype lift in enhancing performance, stereotype lift is the result of a negative outgroup stereotype, whereas stereotype boost occurs due to activation of a positive ingroup stereotype.

Consistent with the positive racial stereotype concerning their superior quantitative skills, Asian American women performed better on a math test when their Asian identity was primed compared to a control condition where no social identity was primed. Conversely, these participants did worse on the math test when instead their gender identity—which is associated with stereotypes of inferior quantitative skills—was made salient, which is consistent with stereotype threat. Two replications of this result have been attempted. In one case, the effect was only reproduced after excluding participants who were unaware of stereotypes about the mathematical abilities of Asians or women, while the other replication failed to reproduce the original results even considering several moderating variables.

Long-term and other consequences

Decreased performance is the most recognized consequence of stereotype threat. However, research has also shown that stereotype threat can cause individuals to blame themselves for perceived failures, self-handicap, discount the value and validity of performance tasks, distance themselves from negatively stereotyped groups, and disengage from situations that are perceived as threatening.

Studies examining stereotype threat in Black Americans have found that when subjects are aware of the stereotype of Black criminality, anxiety about encountering police increases. This, in turn, can lead to self-regulatory efforts, more anxiety, and other behaviors that are commonly perceived as suspicious to police officers. Because police officers tend to perceive Black people as threatening, their reactions to these anxiety-induced behaviors are commonly more harsh than reactions to White people with the same behavior, and influences whether or not they decide to shoot the person.

In the long run, the chronic experience of stereotype threat may lead individuals to disidentify with the stereotyped group. For example, a woman may stop seeing herself as "a math person" after experiencing a series of situations in which she experienced stereotype threat. This disidentification is thought to be a psychological coping strategy to maintain self-esteem in the face of failure. Repeated exposure to anxiety and nervousness can lead individuals to choose to distance themselves from the stereotyped group.

Although much of the research on stereotype threat has examined the effects of coping with negative stereotype on academic performance, recently there has been an emphasis on how coping with stereotype threat could "spillover" to dampen self-control and thereby affect a much broader category of behaviors, even in non-stereotyped domains. Research by Michael Inzlicht and colleagues suggest that, when women cope with negative stereotypes about their math ability, they perform worse on math tests, and that, well after completing the math test, women may continue to show deficits even in unrelated domains. For example, women might overeat, be more aggressive, make more risky decisions, and show less endurance during physical exercise.

The perceived discrimination associated with stereotype threat can also have negative long-term consequences on individuals' mental health. Perceived discrimination has been extensively investigated in terms of its effects on mental health, with a particular emphasis on depression. Cross-sectional studies involving diverse minority groups, including those relating to internalized racism, have found that individuals who experience more perceived discrimination are more likely to exhibit depressive symptoms. Additionally, perceived discrimination has also been found to predict depressive symptoms in children and adolescents. Other negative mental health outcomes associated with perceived discrimination include a reduced general well-being, post-traumatic stress disorder, anxiety, and rebellious behavior. A meta-analysis conducted by Pascoe and Smart Richman has shown that the strong link between perceived discrimination and negative mental health persists even after controlling for factors such as education, socioeconomic status, and employment.

Mitigation

Additional research seeks ways to boost the test scores and academic achievement of students in negatively stereotyped groups. Such studies suggest various ways in which the effects of stereotype threat may be mitigated. For example, there have been increasing concerns about the negative effects of stereotype threats on MCAT, SAT, LSAT scores, etc. One effort at mitigation of the negative consequences of stereotype threat involves rescaling standardized test scores to adjust for the adverse effects of stereotypes.

Perhaps most prominently, well replicated findings suggest that teaching students to re-evaluate stress and adopt an incremental theory of intelligence can be an effective way to mitigate the effects of stereotype threat. Two studies sought to measure the effects of persuading participants that intelligence is malleable and can be increased through effort. Both suggested that if people believe that they can improve their performance based on effort, they are more likely to believe that they can overcome negative stereotypes, and thus perform well. Another study found that having students reexamine their situation or anxiety can help their executive resources (attentional control, working memory, etc.), rather than allowing stress to deplete them, and thus improve test performance. Subsequent research has found that students who are taught an incremental view of intelligence do not attribute academic setbacks to their innate ability, but rather to a situational attribute such as a poor study strategy. As a result, students are more likely to implement alternative study strategies and seek help from others.

Research on the power of self-affirmation exercises has shown promising results as well. One such study found that a self-affirmation exercise (in the form of a brief in-class writing assignment about a value that is important to them) significantly improved the grades of African-American middle-school students, and reduced the racial achievement gap by 40%. The authors of this study suggest that the racial achievement gap could be at least partially ameliorated by brief and targeted social-psychological interventions. Another such intervention was attempted with UK medical students, who were given a written assignment and a clinical assessment. For the written assignment group, white students performed worse than minority students. For the clinical assessment, both groups improved their performance, though the gap between racial groups was maintained. Allowing participants to think about a positive value or attribute about themselves prior to completing the task seemed to make them less susceptible to stereotype threat. Self-affirmation has also been shown to mitigate the performance gap between female and male participants on mathematical and geometrical reasoning tests. Similarly, it has been shown that encouraging women to think about their multiple roles and identities by creating self-concept map can eliminate the gender gap on a relatively difficult standardized test. Women given such an opportunity for reflection did equally well as men on the math portion of the GRE, while women who did not create a self-concept map did significantly worse on the math section than men did.

Increasing the representation of minority groups in a field has also been shown to mitigate stereotype threat. In one study, women in STEM fields were shown a video of a conference with either a balanced or unbalanced ratio of men to women. The women viewing an unbalanced ratio reported a lower sense of belonging and less desire to participate. Decreasing cues that reflect only a majority group and increasing cues of minority groups can create environments that mitigate against stereotype threat. Further research has focused on constructing environments such that the physical objects in the environment do not reflect one majority group. For instance, in one study, researchers argued that individuals make decisions about group membership based on the group's environment and showed that altering the physical objects in a room boosted minority participation. In this study, removing stereotypical computer science objects and replacing them with non-stereotypical objects increased female participation in computer science to an equal level as male peers.

Directly communicating that diversity is valued may also be effective. One study revealed that a company's pamphlet stating a direct value of diversity, compared to a color blind approach, caused African Americans to report an increase in trust and comfort towards the company. Promoting cross-group relations between people of varying backgrounds has also been shown to be effective at promoting a sense of belonging among minority group members. For instance, a 2008 study indicates that students have a lower sense of belonging at institutions where they are the minority, but developing friendships with members of other racial groups increased their sense of belonging. In 2007, a study by Greg Walton and Geoffrey Cohen showed results in boosting the grades of African-American college students, and eliminating the racial achievement gap between them and their white peers over the first year of college, by emphasizing to participants that concerns about social belonging tend to lessen over time. These findings suggest that allowing individuals to feel as though they are welcomed into a desirable group makes them more likely to ignore stereotypes. The upshot is that if minority college students are welcomed into the world of academia, they are less likely to be influenced by the negative stereotypes of poor minority performance on academic tasks.

One early study suggested that simply informing college women about stereotype threat and its effects on performance was sufficient to eliminate the predicted gender gap on a difficult math test. The authors of this study argued that making people aware of the fact that they will not necessarily perform worse despite the existence of a stereotype can boost their performance. However, other research has found that merely providing information is not enough, and can even have the opposite effect. In one study, women were given a text "summarizing an experiment in which stereotypes, and not biological differences, were shown to be the cause of women's underperformance in math", and then they performed a math exercise. It was found that "women who properly understood the meaning of the information provided, and thus became knowledgeable about stereotype threat, performed significantly worse at a calculus task". In such cases, further research suggests that the manner in which the information is presented –– that is, whether subjects are made to perceive themselves as targets of negative stereotyping –– may be decisive.

Criticism

Some researchers have argued that stereotype threat should not be interpreted as a factor in real-world achievement gaps. Reviews have raised concerns that the effect might have been over-estimated in the performance of schoolgirls and argued that the field likely suffers from publication bias.

According to Paul R. Sackett, Chaitra M. Hardison, and Michael J. Cullen, both the media and scholarly literature have wrongly concluded that eliminating stereotype threat could completely eliminate differences in test performance between European Americans and African Americans. Sackett et al. argued that, in Steele and Aronson's (1995) experiments where stereotype threat was mitigated, an achievement gap of approximately one standard deviation remained between the groups, which is very close in size to that routinely reported between African American and European Americans' average scores on large-scale standardized tests such as the SAT. In subsequent correspondence between Sackett et al. and Steele and Aronson, Sackett et al. wrote that "They [Steele and Aronson] agree that it is a misinterpretation of the Steele and Aronson (1995) results to conclude that eliminating stereotype threat eliminates the African American-White test-score gap." However, in that same correspondence, Steele and Aronson point out that "it is the stereotype threat conditions, and not the no-threat conditions, that produce group differences most like those of real-life testing."

In a 2009 meta-analysis, Gregory M. Walton and Steven J. Spencer argued that studies of stereotype threat may in fact systematically under-represent its effects, since such studies measure "only that portion of psychological threat that research has identified and remedied. To the extent that unidentified or unremedied psychological threats further undermine performance, the results underestimate the bias." Despite these limitations, they found that efforts to mitigate stereotype threat significantly reduced group differences on high-stakes tests.

In 1998, Arthur R. Jensen criticized stereotype threat theory on the basis that it invokes an additional mechanism to explain effects which could be, according to him, explained by other, at the time better known and more established theories, such as test anxiety and especially the Yerkes–Dodson law. In Jensen's view, the effects which are attributed to stereotype threat may simply reflect "the interaction of ability level with test anxiety as a function of test complexity". However, a subsequent study by Johannes Keller specifically controlled for Jensen's hypothesis and still found significant stereotype threat effects.

Gijsbert Stoet and David C. Geary reviewed the evidence for the stereotype threat explanation of the achievement gap in mathematics between men and women. They concluded that the relevant stereotype threat research has many methodological problems, such as failing to adjust for pre-existing mathematics scores and not having a control group, and that some literature on this topic misrepresents stereotype threat as being more well-established than it is. It was only when using the studies that used adjusted mathematics scores, and not when including the studies that did not make such adjustments, that they found evidence for an effect of stereotype threat.

Publication bias

A meta-analysis by Flore and Wicherts (2015) concluded that the average reported effect of stereotype threat is small, and that those reports may be inflated by publication bias. They argued that, correcting for this, the most likely effect size may be near zero.

Ganley et al. (2013) examined stereotype threat in a well-powered (total number approximately 1000) multi-experiment study and concluded that "no evidence that the mathematics performance of school-age girls was impacted by stereotype threat" was found. Positing that large, well-controlled studies have tended to find smaller or non-significant effects, the authors argued that evidence for stereotype threat in children may reflect publication bias. They also suggested that, among the many underpowered studies run, researchers may have selectively published those in which false-positive effects reached significance.

A 2020 meta-analysis by Liu et al. found that, while publication bias may inflate the effectiveness of interventions to mitigate stereotype threat, the level of bias is insufficient to overturn the consensus that such interventions are associated with performance benefits. The authors broke down the studies they analyzed into three types – belief-based, identity-based, and resilience-based – finding greater evidence for publication bias in the last of these and more robust evidence for the effectiveness of intervention in the first two types.

Search This Blog

Wednesday, September 4, 2024

Academic authorship

Definition

Authorship in the natural sciences

Authorship in mathematics

Authorship in medicine

Authorship in the social sciences

Authorship in the humanities

Growing number of authors per paper

Order of authors in a list

Responsibilities of authors

Unconventional types of authorship

Honorary authorship

Gift, guest and rolling authorship

Ghost authorship

Fraudulent paid-for authorship

Anonymous and unclaimed authorship

Non-human authorship

Conflicts of interest in academic publishing

Avoidance, disclosure, and tracking

Codes of conduct

COIs by agent

COIs of journals

Owners and governing bodies

Reprints

Impact factors, reputation, and subscriptions

Advertising

Sponsored supplements

Publishers

COIs of journal staff

COIs of peer reviewers

COIs of article authors

Ghost authors and non-contributing authors

COIs of study sponsors

Effect on conclusions of research

Interests of research participants

Publication of COI declarations

False statements of COIs

Publication bias

Definition

Evidence

Impact on meta-analysis

Compensation examples

Risks

Remedies

Study registration

Megastudies

Stereotype threat

Empirical studies

Mechanisms

Original study

Stereotype lift and stereotype boost

Long-term and other consequences

Mitigation

Criticism

Publication bias

Historical materialism