People are the subjects of social research, and because people have rights and feelings, special considerations apply in social research that do not confront the chemist studying molecules or the physicist investigating gravity. Consider, for example, the area of substance-abuse control and treatment. In recent years, considerable controversy has arisen over the use of drug testing to deal with the problem. Some employers have instituted random drug testing of their employees as a condition of gaining employment or remaining employed. Some college athletes are randomly tested, as are prison inmates. In the context of social research, drug testing of a population would be a useful method for assessing the effectiveness of drug interdiction programs or substance-abuse counseling programs. To be most effective, such testing should be required of all people affected by a program. So if new security procedures to control the entry of drugs into a prison are established, their effectiveness can be measured by random testing of a sample of inmates. Likewise, if a drug education program is established in a college, random testing of a sample of students at that college could assess the program's effectiveness.

Would such involuntary drug testing of inmates or students, however, be proper? This question can be answered only by referring to some cultural, professional, or personal values that help us decide what is right and proper behavior. In the United States, three such values have been articulated in the Belmont Report (National Commission, 1978), a document that is considered to be a pillar of human research ethics. These principles are:

1. Respect for persons: We recognize the personal dignity and autonomy of individuals, and we should provide special protection of those persons with diminished autonomy.

2. Beneficence: We have an obligation to protect persons from harm by maximizing anticipated benefits and minimizing possible risks of harm.

3. Justice: The benefits and burdens of research should be distributed fairly.

Viewing drug testing in this context, some would argue that people's rights to autonomy and privacy preclude taking bodily fluids against their will to find out something about them. Is such an invasion of the privacy of college students proper or acceptable? What about prison inmates? Involuntary testing of a whole population also would seem to violate self-determination. The subjects have no choice about participating in the research. Because a person who tests positive for drugs might be subject to criminal prosecution or stigmatization, it raises the question of whether the research could bring harm to the person. Finally, who bears the cost of, and who benefits from, such research? Some would be quick to employ drug testing to evaluate a treatment program with prison inmates, but would be reluctant to subject participants in a middle-class community program to such scrutiny, even though the program may be implemented there once it is proven to be effective.

These are complicated issues. We do not presume that our statement of values is the final word or that the application of these values to particular cases is easy or straightforward. They do serve as an introduction, however, to the discussion of ethical issues in social research.

Ethics is the study of what is proper and improper behavior, of moral duty and obligation (Drewry, 2004). Moral principles can be grounded in philosophy, theology, or both. For social researchers, ethics involves the responsibilities that researchers bear toward those who participate in research, those who sponsor research, and those who are potential beneficiaries of research. It covers many specific issues. For example, is it ever permissible to harm people during the course of a research project? Should people who participate in a research project ever be deceived? Is it appropriate to suppress research findings that cast a sponsor's program in a negative light? Should researchers report to the police crimes they uncover while conducting a research project? The ethics of a given action depends on the standards used to assess the action, and those standards are grounded in human values. Because of this, ethical questions have few simple or final answers and no scientific tests that can show us whether actions are ethical. In fact, debate continues among scientists about ethical issues, because such issues involve matters of judgment and assessment.

Our purpose in this chapter is to identify the basic ethical issues in social research and suggest some strategies for making sure that ethical considerations are covered in the conduct and use of research. Furthering that goal, each of the remaining chapters in this book will include a short section titled “Eye on Ethics,” in which a specific ethical issue related to that chapter is discussed.

· My Bookshelf

· TOC/Annotation menu

· Downloads

· Print

· Search

· Profile

· Help

3.2 Ethical Issues

Previous section

Next section

3.2 ETHICAL ISSUES

Six basic ethical issues arise in social science research: informed consent, deception, privacy (including confidentiality and anonymity), physical or mental distress, problems in sponsored research, scientific misconduct or fraud, and scientific advocacy. The unique situation confronting the human service researcher raises two additional considerations: protecting vulnerable clients and withholding treatment for research purposes.

Informed Consent

Informed consent refers to telling potential research participants about all aspects of the research, which might reasonably influence the decision to participate. Very often, people are asked to sign a consent form, which describes the elements of the research that might influence a person's decision to participate (see Table 3.2). General agreement exists concerning the desirability of informed consent in behavioral science research, primarily because in the United States, cultural values place great emphasis on freedom and self-determination. Whether the issue is whom to marry, what career to pursue, or whether to participate in a research project, we value the right of individuals to assess information and weigh alternatives before making their own judgments. To deceive potential research participants is to deny them the ability to determine their own destinies.

Although consensus exists about the general principle of informed consent, the debate regarding exactly how far the obligations of researchers extend in this realm is still a hot issue. At one extreme, researchers known as ethical absolutists argue that people should be fully informed of all aspects of the research in which they might play a part (Baumrind, 1985; Elms, 1982; Kimmel, 1988). Even when research is based on the public record, such as agency documents, or on observations of behavior in public, some absolutists argue that people about whom the observations have been made should be informed of the research. Otherwise, they do not have the full right to decide whether to participate.

TABLE 3.2 Basic Elements of an Informed Consent Form

The Following Information Shall Be Provided to Each Subject:
(1)	A statement that the study involves research, an explanation of the purposes of the research and the expected duration of the subject's participation, a description of the procedures to be followed, and identification of any procedures that are experimental;
(2)	A description of any reasonably foreseeable risks or discomforts to the subject;
(3)	A description of any benefits to the subject or others, which may reasonably be expected from the research;
(4)	A disclosure of appropriate alternative procedures or courses of treatment, if any, that might be advantageous to the subject;
(5)	A statement describing the extent, if any, to which confidentiality of records identifying the subject will be maintained;
(6)	For research involving more than minimal risk, an explanation as to whether any compensation or any medical treatments are available if injury occurs, and, if so, what they consist of, or where further information may be obtained;
(7)	An explanation of whom to contact for answers to pertinent questions about the research and research subjects' rights and whom to contact in the event of a research-related injury to the subject; and
(8)	A statement that participation is voluntary and refusal to participate will involve no penalty or loss of benefits to which the subject is otherwise entitled, and that the subject may discontinue participation at any time without penalty or loss of benefits to which the subject is otherwise entitled.

SOURCE: Part 46, Protection of Human Subjects. Code of Federal Regulations, Title 45, Public Welfare. DHHS National Institutes of Health Office for Protection from Research Risks, 2001.

Rigid adherence to the absolutist position, however, makes social research much more difficult to conduct. First, such adherence rules out many practices that some researchers consider important or essential. Many experiments, for example, rely on some degree of deception, at least to the extent of not telling participants the true research hypotheses. The reason for this is that people might respond differently if they knew about these hypotheses. All such studies would be unacceptable to the absolutists, however, and could not be conducted. Absolutists also would disallow disguised observation, during which people in public settings are not aware they are being observed. No longer could researchers engage in such effective strategies as infiltrating organizations—unless, of course, they told all employees what they were doing. The net result would be to make social science research highly conservative and limited. It would become the study of people who volunteer to be studied; and research shows that people who volunteer differ in many ways from people who do not volunteer. This would seriously reduce the generalizability of research findings.

Second, the absolutist approach would call for obtaining informed consent in all research projects, and research has shown that obtaining written consent can reduce people's willingness to participate in research. One study found that formally requesting informed consent prior to conducting an interview reduced the rate of cooperation by 7 percent in comparison to cases in which a formal request was not made (Singer, VonThurn, & Miller, 1995). Because any reduction in response rate will reduce the generalizability of the findings, as we will explain in Chapter 7, obtaining informed consent in survey research can have serious, negative consequences in terms of the validity of the research.

Written informed consent probably reduces people's willingness to participate because it appears to some respondents to contradict the researcher's assurance of confidentiality. One minute people are being told that their answers will remain confidential, and the next they are being asked to put their names on a consent form! Even though signing a consent form need not impede the maintenance of confidentiality, it is not surprising that respondents may not perceive it that way. Anything that undermines the belief of respondents in the confidentiality of their answers will reduce the response rates. Signing a consent form also may affect the quality of the data obtained, because those who do give consent may be less candid in their responses than they otherwise might have been.

Because of these problems with the absolutist approach to informed consent, many researchers take a less extreme position. They argue, first of all, that people should be informed of factors that might reasonably be expected to influence their decision to participate, such as any harm that might occur or how much time and effort will be involved. However, they also use a risk—benefit approach: Questionable research strategies, such as deception, are appropriate if they are essential to conduct the research, if they will bring no harm to participants, and if the outcome is sufficiently important to warrant the research. Lively debate continues over the use of questionable practices and their routine approval by IRBs (Baumrind, 1985).

A final issue regarding informed consent has to do with the possibility that a person might feel pressured to agree or might not understand precisely to what he or she is agreeing. After all, asking a person to participate in a study can involve social pressures not unlike those in other settings. We often feel pressured to help others when they ask for our assistance, and some people find it difficult to say no to a face-to-face request. In addition, scientific researchers represent figures of some authority and status, and people often are disinclined to refuse their requests. In other cases, people may be temporarily confused about what is being asked of them.

To resolve these problems, a study involving institutionalized elderly people used a two-step consent procedure (Ratzan, 1981). In the first step, people were told what their participation would involve, what risks were entailed, and that several days later, they would be asked whether they were willing to participate. This was meant to reduce any immediate pressures to agree and to enable people to talk with others and clarify any confusing issues. The second step, occurring a few days later, involved obtaining the actual consent. In that study, all those who were asked refused to participate. This probably resulted from the fact that a part of the study involved having a needle inserted in a vein continuously for eight days to take blood samples. However, some of the people might have agreed to participate had they been asked to sign a consent form during the first interview. Yet, it is questionable whether any consent obtained at that time would have been as informed and considered as it should have been. The goal of obtaining informed consent is not to pressure people into participating in the research but, rather, to gain participation that is truly informed.

Deception

Despite its status as a controversial strategy, deception is still fairly common in some areas of social science research such as social psychology (Korn, 1997). One reason for this is that some research would be difficult—or even impossible—to conduct without some level of deception. At a minimum, many experiments necessitate not telling the participants the true research hypotheses; this withholding of information is a form of deception. As another example, field research sometimes uses disguised observation, where people in public settings are observed but are unaware that the observation is occurring. Some argue that this involves an implicit deception of those being observed. Other types of field research can involve more explicit deception to gain the cooperation of those being observed.

Richard Leo (1996), in his field study of police interrogators, withheld or concealed information regarding his own views about crime, punishment, homosexuality, and other issues from the police that he observed. In fact, he dubbed what he did the “chameleon strategy”: “I consciously reinvented my persona to fit the attributes, biases, and world-view of my subjects” (Leo,1995, p. 120). Before the detectives, he feigned opposition to abortion, support for the death penalty, and antipathy toward gays. He created this persona to gain their cooperation and encourage them to act openly and naturally in front of him. If the police interrogators had known his true views on these issues, Leo claims, they might have distrusted him and not acted openly in front of him. Because his observations produced ground-breaking sociological research on this topic, he felt that the deception was warranted, given that little harm came to the police officers.

Ethical absolutists would rule out all such deception on the grounds that it is unethical to deceive other human beings deliberately (Erikson. 1967). Leo, however, argues that the situation is more complicated: “Fieldwork is a morally ambiguous enterprise that is fraught with moral hazards, contingencies, and uncertainties” (Leo, 1996, p. 125). There are trade-offs and compromises to be taken into account. Furthermore, Leo makes the distinction between “acts of commission,” in which the researcher intentionally falsifies information, and “acts of omission,” in which information is withheld. Some would argue that Leo did the former when he “reinvented” his persona, but Leo claims that his research strategy involved only omission, and thus was ethically justifiable. In addition, he argued that his use of deception passed the test of the three major criteria that are applied: (1) The research is important; (2) no other way to conduct the research is available; and (3) no harm came to those who were studied.

When deception is used, it is a sound practice to conclude the period of observations with a debriefing, during which people are told the true purposes of the research and informed of any deceptions that were used. This should be done in a positive and supportive way so the participants feel they were joint partners in a worthwhile enterprise rather than dupes of the researchers. Usually, this can be achieved by explaining the reasons for the deception and the impossibility that anyone could have detected the deceptions before being informed of them. In some research, of course, debriefings are not possible, such as in disguised observations in field settings where the researcher never has a chance to contact the participants after the observations are made.

The Right to Privacy: Anonymity and Confidentiality

The right to privacy is one of the key values and ethical obligations mentioned at the beginning of this chapter. Privacy refers to the ability to control when and under what conditions others will have access to your beliefs, values, or behavior. Intrusions on our privacy have become endemic in modern society; with the growth of social research during the past century, the danger of even greater intrusion arises. Virtually any attempt to collect data from people raises the issue of privacy and confronts investigators with the dilemma of whether threats to privacy are warranted by the research. Two well-known research projects illustrate the complexity of this issue.

Two Case Studies. The sociologist Laud Humphreys (1970), in an effort to understand a particular type of sexual behavior, made observations of men having quick and impersonal sexual encounters with other men in public restrooms. To gather his data without arousing suspicion, Humphreys played the role of the “watch queen”—that is, someone who keeps watch and warns participants of approaching police or “straight” males who might disrupt the activities. None of the men who went to the restrooms to engage in sex was aware that a researcher was recording his behavior. Humphreys also noted the license plates on the cars of these men, and was able to find their home addresses through public motor-vehicle records. He then interviewed them in their homes but did not inform them of the real reason for the interviews or that he had observed them earlier in the restrooms.

Humphreys was heavily criticized for using deception, for not obtaining informed consent, and for violating the privacy of these men who were engaging in highly stigmatized actions that, were they to be made public, might disrupt their family lives or threaten their jobs. Many sociologists believe that research on such sensitive topics is not simply a matter of confidentiality—that is, of not letting people's identities become known. Rather, these sociologists believe such data should not even be collected at all, because the men obviously were trying to conceal their actions. Social science researchers, it is argued, should respect that privacy.

Humphreys defended his research on the grounds that the confidentiality of his subjects was maintained and the results of the study were of significant scientific value. In fact, no one else has devised another method to study such sexual behavior. Humphreys discovered that the men who engage in this type of sexual activity were not unusual or deviant in the rest of their lives and, for the most part, were normal, respectable citizens with a rather unusual sexual outlet. Humphreys believed that the greater understanding of what had been considered to be deviant sexual conduct justifies the threat to privacy that these men experienced. In addition, the public setting in which they performed their acts, he argued, reduced their claim of a right to privacy.

Another field study illustrates further difficulties in protecting people's privacy. When research describes an actual group or community of people, a common practice is to report results by giving fictitious names to the people and places to protect privacy. This procedure works fairly well as long as researchers do not become so detailed in their descriptions of places, events, and people that the protection afforded by fictitious names is undermined. A now infamous example of this was Small Town in Mass Society, by Arthur Vidich and Joseph Bensman. First published in 1958, this observational community study described the power relationships and local governmental operations in a small town the authors called Springdale (Whyte, 1958). The authors had assured all the people they interviewed that confidentiality would be maintained, and no identities were directly revealed. However, it was easy for the residents of the small community to recognize the people and events described in the highly detailed report. Because the study was critical of some residents of Springdale, these people understandably became outraged. Local newspapers vociferously attacked the researchers for betraying the community. The townspeople even held a Fourth of July parade that featured a full manure spreader carrying mired effigies of the authors. To avoid such breaches of confidentiality, one must balance enthusiasm for producing a highly detailed account against the ethical obligation to fully protect the identities of those observed.

Single-system research requires particular caution in this regard. As we will explain in Chapter 11, single-system research involves observing changes in the feelings or behavior of an individual over a period of time. Furthermore, the clients in these studies often suffer from some conditions, such as a mental disorder, that might lead to stigmatization should others find out about it. Therefore great care must be taken to ensure that people's identities are not unintentionally revealed in the process of describing the case. Researchers often must tread a fine line between providing sufficient case detail and minimizing the risk of identifying a client. Final reports should always be written with sensitivity to this issue.

As these two case studies show, the right to privacy often is a difficult ethical issue to resolve. Researchers have come up with three major ways to deal with the problem of protecting people's privacy: (1) Let subjects edit their data; (2) keep the data anonymous; and (3) keep the data confidential.

Editing the Data. One very effective way to protect privacy is to offer participants the opportunity, after the data has been collected, to destroy any data they wish to keep private. This has been done in studies of couples and family problems that use audio or video recordings of interactions between family members (Margolin, 2005). Even though the couples and families consent to the recording, investigators, sensitive to the issue of privacy, sometimes offer the opportunity for participants to review the recordings and edit out anything they wish. The assumption is that, despite agreeing to participate, family members might do something on the spur of the moment that they would prefer not be made public. Surprisingly, few family members exercise this option. This and other research suggests that people are generally more tolerant of invasions of privacy than researchers might expect. Nonetheless, it should always be the research subject's decision; the researcher should never assume that people will be tolerant of invasions or breaches of their privacy.

Anonymity. A second means of ensuring privacy is to accord the participants anonymity , which means that no one, including the researcher, can link any data to a particular respondent. This can be accomplished by not including any identifying names or numbers with the data collected. True anonymity means that even the researcher can never link data to a particular respondent. This method of protecting privacy is ideal, because the data are collected in such a way that it is impossible for anyone to determine which data come from which individual. In many research situations, however, it is not possible or feasible to collect data in this fashion, so researchers turn to confidentiality.

Confidentiality. A third way of protecting privacy is through confidentiality , which means ensuring that information about or data collected from those who participate in a study are not made public in a way that can be linked to an individual. Researchers, of course, commonly release their data to the public, but usually only in aggregate forms— that is, reporting on how a whole group responded rather than how specific individuals responded. In many studies, the data are not of a sensitive nature, and thus confidentiality would seem to be less important; but it is impossible to predict what bits of data all participants will want to be kept confidential. Because aggregate and anonymous reporting of results is all that is necessary in most research studies, confidentiality is routinely extended to encourage people's participation and honest responses.

It is worth pointing out that anonymity and confidentiality are quite distinct. With confidentiality, the researcher can link responses to particular respondents but does not release this information publicly; with anonymity, even the researchers cannot link responses to particular respondents. So if a researcher sends out mailed questionnaires that are returned with no names or other identifiers on them, then the respondents have true anonymity. If the same questionnaires are mailed back with a name or any other identifier on them, however, then only confidentiality is possible, because the researcher can link responses to respondents. Even if the identifiers are removed as soon as the questionnaire is received, it would still be ethical to make promises of confidentiality only, because the researcher could link responses to respondents at the point when the questionnaires are received. Obviously, it is unethical to tell respondents that they will have confidentiality when they will not, but it also is wrong to claim that responses will be anonymous when all that will actually be protected is confidentiality.

In some cases, confidentiality can be breached merely by a person's involvement in a research project becoming known. If it becomes known, for example, that an individual was part of a study of drug users, then knowledge that the individual is a drug user has become public even if none of the data collected from that individual has been made public. So one aspect of confidentiality is to ensure that the identities of those who are research subjects are kept hidden, especially when that research is sensitive.

Confidentiality also can be threatened when third parties, such as the people sponsoring the research or the courts, seek to identify research participants. Intrusion by a sponsor is relatively easy to avoid, however. When establishing a research agreement with a sponsoring agency or an organization, one should make clear in the agreement that identities will not be revealed under any circumstances. If the sponsor objects, then researchers should refuse to accept the agreement. We have more to say about sponsors and ethics later in this chapter.

Court and Legislative Challenges to Confidentiality. The courts as well as some laws and statutes pose a more complicated threat to confidentiality. Most communications that physicians, lawyers, and clergy have with their clients are protected from judicial subpoena. Social workers in their clinical capacity also are afforded such protection in many instances, although the degree of protection varies with agency settings and jurisdictions. Social researchers, however, do not have a legal protection of privileged communication with the people from whom they gather data (Kimmel, 1988; Reece & Siegal, 1986). In addition, physicians, most human service professionals, and, in some cases, researchers are required by many state laws to report to public agencies when they observe evidence for mistreatment of children or the elderly. Thus, courts or public agencies may subpoena research data that reveal participants' identities, and failure to comply with such a subpoena renders researchers open to charges of contempt of court. Actually, social science researchers have been treated somewhat inconsistently by the courts in civil cases. During a civil suit in California, for example, the court refused to force a researcher to reveal the identities of respondents in confidential interviews (Smith, 1981).

In criminal cases, however, the courts generally have held that the right of the public to be protected from criminal activity or the right of suspects to a fair trial supersedes any assurance of confidentiality in research. In one case, for example, the researcher had been making field observations in a restaurant when it was heavily damaged by a suspicious fire (Brajuha & Hallowell, 1986). Police wanted the researcher's field notes to determine whether any evidence of arson could be substantiated. One court squashed a subpoena for the field notes, but another upheld it. Eventually, a compromise was reached: The researcher's field notes were considered to be subject to a subpoena, but the researcher was allowed to remove any material that would have violated confidentiality. With rulings like this, the courts seem to recognize that the confidentiality a legitimate social science researcher establishes in a research relationship should be protected if possible while still protecting the rights of the public and criminal suspects.

In another case, a sociologist actually spent time in jail because he refused to give information in court that he believed violated his promise of confidentiality to his research subjects. Rik Scarce (2005) conducted research on activists in the animal liberation movement during the 1990s. A federal grand jury was investigating break-ins by such activists at university laboratories, and some of the activists in whom the grand jury was interested had been interviewed by Scarce as part of his research. Scarce refused to answer certain questions about these activists that were put to him by the grand jury, because he thought it would violate the confidentiality he had extended to the people he interviewed. He was jailed for four months on contempt-of-court charges. In his study of police interrogators described earlier in this chapter, Leo was compelled to testify in court under threat of contempt-of-court charges. As these cases illustrate, the courts generally have held that confidential communication between a researcher and a research participant is not protected in criminal cases; however, some courts also have recognized that the confidential relationship is an important and special one, and efforts should be made not to violate it. Because of cases like these, some researchers have called for federal legislation to give social science researchers protection—even if limited—from being compelled to violate their confidential relationships (Leo, 1995; Scarce, 2005).

Concern over a possible subpoena of their research data has led some researchers to adopt elaborate measures to protect that data. For example, it is common to establish computer files with the data identified only by numbers rather than by names. Often, it is unnecessary to retain name identification for research purposes once the data have been collected. In such cases, the names should be destroyed. If name identification is required—for example, because you want to interview the same people at a later time—the names should be stored in a separate computer file. This procedure reduces the possibility of unauthorized persons linking names with data.

One of the best means of securing confidentiality for sensitive research, such as that dealing with AIDS, substance abuse, or criminal behavior, is to use certificates of confidentiality, which were made available by the Public Health Service Act Amendments of 1974 (Beskow, Dame, & Costello, 2008; Bonnie & Wallace, 2003). Certificates of confidentiality are issued by the National Institutes of Health (NIH) and other Health and Human Service offices to protect identifiable research information from forced disclosure. They allow the investigator and others who have access to research records to refuse to disclose identifying information about research participants in any civil, criminal, administrative, legislative, or other proceeding, whether at the federal, state, or local level. By some interpretations, the certificates of confidentiality supersede even laws that require mandatory reporting of instances of mistreatment involving children and the elderly. Certificates generally are awarded only for research on sensitive topics, such as illegal drug use or sexual behavior, and regardless of whether the research project receives federal funding. While some believe that the certificates provide almost complete protection, some courts have insisted on at least partial disclosure of some information that the certificates were meant to protect. While courts take the certificates seriously, they also tend to view them as only one factor that influences how civil and criminal legal disputes should be resolved. At least some courts have taken the position that the constitutional rights of a plaintiff or defendant may supersede the legal protections offered by the certificates (Beskow, Dame, & Costello, 2008; National Cancer Institute, 1998).

So intrusion by courts is a real danger with research, especially regarding some topics, although it is not terribly common. Nonetheless, researchers have an obligation to inform potential subjects accurately about any possible threats to confidentiality that might arise, including what would likely happen should their data be subpoenaed by the courts. The lengths to which investigators will go to protect privacy indicate the importance of this ethical issue. The bottom line is that no one should be threatened with harm to themselves or their reputation as a result of participating in a scientific study.

Harm, Distress, and Benefit

Researchers should avoid exposing participants to physical or mental distress or danger. If the potential for such distress exists in a research investigation, then the participants should be fully informed, the potential research findings should be of sufficient importance to warrant the risk, and no possibility should exist of achieving the results without this risk. People should never be exposed to situations that might cause serious or lasting harm.

Research in the human services rarely involves physical danger, but there are research settings in which psychological distress may be an element. Some studies, for example, have asked people to view such things as pornographic pictures, victims of automobile accidents, and the emaciated inmates of Nazi concentration camps. Certainly, these stimuli can induce powerful reactions—in some cases, emotions that the participants had not expected to experience. A strong emotional reaction, especially an unexpected one, can be very distressful. In some studies, people have been given false feedback about themselves to observe how they respond. People have been told, for example, that they failed an examination or that tests show they have some negative personality characteristics. Any situation in which people might learn something about themselves of which they were previously unaware also can be distressing. Even the minor deceptions that are part of much research can be distressful to people who thought they could not be deceived so easily.

Another area in which social science research contains the potential for harm to those who are studied is in field research of people engaging in stigmatized, deviant, or illegal behavior. The potential harm comes if the researcher says or does anything that might lead the research subjects to be sanctioned for the deviant or illegal behaviors they displayed in front of the researcher (Klockars, 1979). This is a complex topic that will be explored in detail in Chapter 10, but the problem can arise because field research often involves an implicit or explicit agreement between the researcher and those being observed. In this agreement, the observed consent to letting the researcher join them and make observations on them; in return, they expect the researcher will support their illegal or deviant behavior—or at least not do anything to get them in trouble for it. For example, if police interrogations of suspects were likely to include some harsh or even illegal treatment of the suspects, police officers would be unlikely to let a researcher observe the interrogations unless they believed that researcher would support their actions or not report them to authorities or testify against them in court. In other words, the police would likely expect the researcher, in return for being permitted to observe them, to act like fellow officers and support them in their actions. In developing a relationship with the police, the researcher may give the impression that he or she understands this, or the police may just assume that the researcher understands. This actually creates a moral dilemma for the researcher, because either choice of the researcher is morally compromised. If the researcher supports the police and doesn't report the harsh treatment of the suspect, the suspect is harmed by the researcher's actions, but if the researcher supports the suspect and reports the actions, the police are harmed both by having their implicit agreement with the researcher violated and by being sanctioned for their behavior. This kind of ambiguous dilemma is inherent in some field research on deviant lifestyles and is part of what Leo meant by his comment that field work is “a morally ambiguous enterprise” (Leo, 1996, p. 125).

Assuming that the scientific benefits warrant the risk of distress and the participants are fully informed, it is then the researcher's obligation to alleviate the impact of whatever distress actually does occur. As with problems created by deception, alleviating problems related to harm or distress often is accomplished through a debriefing that assesses people's psychological and emotional reactions to the research.

For some social scientists, an ethical standard that rests only on avoiding harm is far too limited. They would argue that those on whom we conduct research also should gain some positive benefit from their participation. This stance has been put forth most clearly by feminist researchers (as discussed in Chapter 2) and collaborative researchers (Nyden et al., 1997; Reinharz, 1992). Viewing research as a two-way street, they argue that both the researcher and the research subject should gain something positive from the project. Researchers, of course, get data from the research enabling them to publish books and articles, which advances their careers; but what do the research subjects get? This raises the ethical issue of what the research community owes to research subjects. These subjects could be paid for their time and effort, or the research results could be translated into social policies or practices that benefit the community from which the research subjects come. Or, by participating in the research, the subjects might develop knowledge or skills that will enable them to make their own lives better. The point is that, from this perspective, researchers have an ethical obligation not only to not leave research participants worse off, but also to compensate them for their time and effort and, in some way, leave them better off for having participated in the research.

Research in Practice 3.1 explores some of the difficulties surrounding such issues as privacy, confidentiality, harm, and informed consent, especially when the research topic is sensitive.

Sponsored Research

Because much social research is conducted under the auspices of a third-party sponsor, certain ethical considerations arise from that relationship. When research is sponsored, some type of research agreement—essentially, a contract—is developed. Researchers and sponsors alike should exercise great care in drafting this agreement. The potential for ethical problems to arise later is reduced when the research agreement clearly specifies the rights and obligations of the parties involved.

Three areas are of particular concern in sponsored research (King, Henderson, & Stein, 1999). First, it is common for sponsors to want to retain control over the release of the collected data. The precise conditions of release should be specified in the research agreement to avoid conflicts. One limitation that should not be tolerated, however, is the conditional publication of results—that is, agreeing to publish results only if they turn out a certain way (usually so when they support the preconceived notions of the sponsor). Such conditional publication violates the integrity of the research process and the autonomy of the researcher (Wolfgang, 1981). If the researcher agrees to some other type of limitation on release, however, it must be honored. To do otherwise would be a breach of the agreement and, therefore, unethical.

The second major concern in sponsored research is the nature of the research project itself. The precise purpose and procedures of the study should be specified in the agreement. Ethical questions arise if the researcher heavily modifies the study to cover matters not in the agreement. Often sponsors will allow researchers to use the data gathered for scientific purposes beyond the needs of the sponsor, but it is unethical to agree to do a study that a sponsor wants and then change it for personal reasons so it no longer meets the sponsor's expectations.

A third area of ethical concern in sponsored research relates to the issue of informed consent— namely, revealing the sponsor's identity to participants. Although controversy exists in this regard, some researchers take the stance that truly informed consent can be given only if a participant knows who is sponsoring the study and for what purpose the study is being conducted. Some people might object, for example, to providing data that would help a company better market a product, a political party to better present a candidate, or the government to propagandize its citizens. In fact, studies show that people are less likely to participate in research they know to be sponsored by commercial organizations, which suggests that information about sponsorship can influence the decision of whether to participate (Fox, Crask, & Kim, 1988). Each researcher, then, must carefully consider whether to make the sponsorship of a research project explicit. At a minimum, to deceive people regarding the sponsorship of a study to gain their participation is certainly unethical.

[image]

RESEARCH IN PRACTICE 3.1 Practice Effectiveness: Sex Offenders as Participants

Applied social researchers often find themselves doing socially sensitive research, research on people who arouse intense and sometimes negative emotions in others—murderers, rapists, prostitutes, or illegal drug users, to name only a few. Such research often involves evaluation research or program evaluations (discussed in Chapters 1 and 12) and can be very challenging, especially in terms of maintaining ethical standards.

One example of such research is a program evaluation assessing two methods of treating people convicted of child molestation (Jenkins-Hall & Osborn 1994). The goal of the research was to provide observational evidence regarding which treatment approach is more effective. In conducting the research, the researchers had to consider the rights of the child-molester participants as well as the safety of staff. There were potential negative consequences for each group from participating in the research. The researchers identified three general areas of concern:

1. informed consent and voluntary participation;

2. confidentiality of data and privacy; and

3. protection against dangers that a participant may pose to himself or to others.

Informed consent was considered to be so important by the researchers that they incorporated a multiphase consent process requiring participants to give consent at each stage of the process. All clients completed a general consent form that covered the basic requirements of the DHHS. Prior to a comprehensive psychological and social assessment that was a necessary part of the research, the clients completed an evaluation consent form that spelled out the assessment process. Additional consent forms were completed in conjunction with each subsequent treatment component. Given the complexity of the intervention, this multiphase consent process, though unusual, ensured that participants understood the program and were freely consenting to participation throughout the project.

Confidentiality and privacy also were considered to be very important because of the negative consequences that might occur to the molesters if information about their participation in the program were ever released. One protection of privacy was the use of federal certificates of confidentiality. Additional steps also were taken: the project was located in a public office building suite and was not identified as a treatment site for sex offenders. Each participant was given a six-digit ID number that appeared on all documents. Names and other identifying information were removed from all correspondence, consents, and records from referral sources. Staff addressed participants by first name only, and information was never released without signed waivers of confidentiality that specified what information could be released and to whom. Because the program involved the delivery of treatment services, each participant was provided with a client advocate who was a volunteer clinician not affiliated with the project; this advocate's sole responsibility was to protect the rights of the client.

In addressing the issue of harm or distress, the researchers considered not only harm to the participants, but also, given that the participants were convicted sex offenders, potential danger to the staff or members of the local community. Several measures were employed to reduce such risks. One concern for the project was quick detection if a participant were in danger of committing another sexual offense. Researchers argued that the project had an ethical and, perhaps, legal duty to protect potential victims. The project provided each participant with a 24-hour crisis call service as well as therapists to deal with minor crises, such as loss of employment or a breakup with a significant other. Over a two-year period, three cases required involuntary commitment of participants because of deterioration in their mental health status. In several cases where the participant was exhibiting negative signs, he was temporarily removed from the project and offered alternative treatment. If a participant who was at risk of relapse failed to follow the terms of a crisis intervention plan, the director of the project was notified and authorities were alerted. Failure to comply on the part of the participant would result in termination from the project.

To assure the safety of the staff, the program employed a variety of procedures. Staff were directed to have unlisted personal phone numbers, never to reveal home addresses, never to be alone with clients in the suite, and never to enter a room that was lockable from the inside with a client. Each office and work area contained a “panic button” that sounded an alarm in the office and at the university police office. Entry to the project was strictly controlled so that client interaction with other office building occupants was minimal. Participants were routinely debriefed after treatment sessions to ensure that they would not leave the center in a distressed, agitated, or aroused state.

Scientific Misconduct and Fraud

When a research project reaches the final stage—that is, the dissemination of results—the primary consideration regarding ethical conduct shifts from avoiding harm to the participants in research to making sure the consumers of the research are not adversely affected. Results of a study typically are communicated in the form of a report to a sponsoring organization, a publication in a professional journal, or possibly a news release to the media. The preeminent ethical obligation in this regard is not to disclose inaccurate, deceptive, or fraudulent results. To do so risks misleading scientists who depend on previous research to guide their own work. Ethical violations concerning disclosure of results undermine the very nature of the scientific process, which, as we saw in Chapter 2, depends upon building future knowledge on the foundation of existing knowledge. If we cannot depend on the accuracy of existing knowledge, then the scientific endeavor is threatened. In that case, the credibility of all research is damaged by such violations. Furthermore, deceptive or fraudulent disclosures of research results can cause human service practitioners to design useless—or even dangerous—interventions based on previous faulty studies.

Many ethical violations can occur in the process of reporting research (Gibelman & Gelman, 2001). Fraud is the deliberate falsification, misrepresentation, or plagiarizing of the data, findings, or ideas of others. These include such things as falsifying data, embellishing research reports, reporting research that has not been conducted, or manipulating data in a deceptive way. Misconduct is a broader concept that includes not only fraud, but also carelessness or bias in recording and reporting data, mishandling data, and incomplete reporting of results. Other questionable practices are irresponsible claims of authorship (listing coauthors who did not really make contributions to the research) and premature release of results to the public without peer review.

Estimates regarding the actual amount of scientific misconduct that occurs suggest the problem is relatively small. The Office of Research Integrity of the DHHS, which keeps track of such matters, reports that 100 to 200 new cases of alleged misconduct are brought to its attention each year, and many of these allegations prove to be unfounded (Office of Research Integrity, 2010). Of course, underreporting could hide a larger problem, but even if the number of cases is small, any misconduct in research can still damage the credibility of all research and place human service clients—and, by extension, human service providers—at risk.

The issue of who has the responsibility for detecting scientific misconduct is still very controversial (Gibelman & Gelman, 2001). For many years, governmental agencies that fund research pushed this responsibility onto the shoulders of universities and other institutions where research was actually conducted. By the late 1980s, however, both the NIH and the DHHS had established their own offices to watch for misconduct. Yet, most scientists still consider protection against misconduct to be primarily a responsibility of the scientific community, which uses two major mechanisms to detect fraud: peer review and replication. There are, however, limitations to each of these mechanisms, as illustrated by a case of fraud in medical research involving Dr. Robert Slutsky (Engler et al., 1987).

The prolific Dr. Slutsky had authored or coauthored 137 articles on cardiological and radiological research over a seven-year period. During an evaluation of his appointment as a researcher at a university, questions were raised about duplicate data in two of his papers. In the ensuing investigation, 12 articles were deemed fraudulent, and 48 were considered questionable. In some cases, articles described experiments that had never been conducted. How is it that these articles eluded the net of peer review by respected medical journals? There seemed to have been two problems: First, peer review of article submissions cannot detect plausible, internally consistent fabrications. Second, the sheer number of research articles submitted to publications for review requires a large number of qualified reviewers who understand both the methodology and the content of the article.

If peer review does not detect and deter fraud, then will replication solve the problem? It might. The effectiveness of replication, however, has been severely crippled by the modern research system, because research funds often are not appropriated for replication. Agencies prefer to fund research efforts that delve into new areas. Furthermore, when replication does occur, it tends to be reserved for projects that have produced unusual results. Fraudulent studies that enhance the prestige of researchers but do not run counter to accepted findings in a field are not likely to arouse enough attention to warrant replication (Engler et al., 1987).

Beyond peer review and replication, there are several ways of reducing the chances of fraud or, if it occurs, detecting it. One is to supervise novice researchers until they demonstrate good practices, ethical conduct, and technical competency. Second, organizations need to guard against overly prolific researchers. Senior researchers and coauthors should not simply allow their names to be associated with research reports without carefully examining the work. Third, journals can reduce fraud by requiring more complete data to be submitted to reviewers, even if all the data cannot be included in the published article. When fraud is detected, journals have a responsibility to make the fraud public so that others will not unsuspectingly base research or treatments on the fraudulent material.

Though not as unethical as purposeful deception, careless errors in research have the same effect of creating misinformation. Social researchers owe the scientific community carefully conducted research that is as free of error as possible. As with fraud, errors are discovered and corrected either through critical review of research and reanalysis of data or through replication. For example, a study from the 1980s claimed to document substantial negative economic consequences of divorce for women; this study was an important foundation for developing social policy. Partly because these findings were somewhat at variance with the results of other studies, a reanalysis of the data was undertaken (Peterson, 1996). This reanalysis showed that the original conclusions were partly the result of errors in the original analysis. The reanalysis showed that women do suffer economically after divorce, but not nearly as severely as the original analysis had suggested. Had the original research conclusions not been based on an inaccurate analysis, subsequent policy development might have been different.

Beyond the problems of fraud and carelessness, researchers also have an obligation to report their results thoroughly. Researchers must take care to ensure that what they report does not give a distorted picture of the overall results. In addition, researchers should point out any limitations that might qualify their findings. Researchers also must be concerned with how the findings are applied to human service practice and policy decisions. Research projects on the effectiveness of arrest as a deterrent to intimate partner violence (see Chapter 10 for a detailed discussion) are excellent examples. A study conducted in Minneapolis has been singled out as a primary catalyst for widespread adoption of pro-arrest policies. Critics have argued that propaganda generated by some of the participating agencies was a key factor in the study's impact on policy, an impact they say was unjustified by a single study. Subsequent replications in several cities failed to support the findings of the Minneapolis study and have underscored the call for caution in rushing to apply findings to practice (Sherman, 1992). This was not a case of publishing fraudulent results or misrepresenting data, but rather an issue of the researcher's role in applying the research findings.

Scientific Advocacy

Scientific knowledge rarely remains the exclusive domain of the scientific community but, typically, finds its way into public life in the form of inventions, technological developments, or social policy. This raises potential ethical dilemmas in terms of the role of researchers as advocates for particular uses of their research results. What responsibility, if any, do scientists have for overseeing the use to which their results are put? To what extent should scientists become advocates for applying knowledge in a particular way? Quite naturally disagreements arise on how to resolve these issues. This controversy is compounded in the case of human service researchers, because their research normally is initiated with some explicit, clinical application in mind. The classic approach to these issues derives from the exhortations of the sociologist Max Weber (1946, orig. pub. 1922) that science should be “value-free.” Social scientists, according to Weber, should create knowledge, not apply it. Therefore, they have no special responsibility for the ultimate use to which that knowledge is put. Furthermore, according to Weber, they are under no obligation to advocate particular uses of scientific knowledge. Indeed, advocacy is frowned on as threatening objectivity, which is a central concern of science. So although remaining value-free is difficult, many argue that abandoning the effort would be disastrous, for it would prevent us from acquiring an accurate body of knowledge about human social behavior and might threaten the researcher's credibility as a disinterested expert (Gibbs, 1983; Gordon, 1988; Halfpenny, 1982).

Karl Marx (1964, orig. pub. 1848) originally developed the opposite stance in this controversy. Marx championed the cause of the poor and downtrodden; he believed that social researchers should bring strong moral commitments to their work and strive to change unfair or immoral conditions. Following Marx, some modern researchers believe that social research should be guided by personal and political values and directed toward alleviating social ills (Brunswick-Heinemann, 1981; Fay, 1987). Furthermore, scientists should advocate for uses of their research by others that would help accomplish these personal goals.

Sociologist Alvin Gouldner (1976) developed a compromise position on the value-free controversy. Gouldner pointed to the obvious—namely, that scientists have values just as other human beings do. Furthermore, he noted, those values can influence research in so many subtle ways that their effects can never be totally eliminated. So instead of denying or ignoring the existence or impact of these personal values, scientists need to be acutely aware of and up front about them in research reports. Being thus forewarned, consumers of their research are better able to assess whether the findings have been influenced by personal bias. In addition, Gouldner argued that social scientists have not only the right, but also the duty to promote the constructive use of scientific knowledge (Gouldner, 1976). Because someone will make decisions concerning the use of scientific knowledge, scientists themselves are best equipped to make those judgments. As Gouldner states, technical competence would seem to provide a person with some warrant for making value judgments. People who take this position view the value-free stance as a potentially dangerous dereliction of a responsibility that accrues to scientists by virtue of their expertise and role in developing new knowledge.

Clinician—researchers, in particular, may be attracted by this stance. Because their research is conducted, in part, to advance the practice goals of the profession, they would probably view ensuring that any clinical application of the results be faithful to the outcome of the research as one of their duties. Thus, human service researchers are more likely than many other behavioral scientists to take a strong stand in favor of advocacy.

Nothing is wrong with researchers openly pressing for the application of scientific knowledge in ways they deem to be desirable—so long as their advocacy is tempered with respect for objectivity. The danger of advocacy is that scientists can come to feel so strongly about the issues they promote that those feelings hamper the objective collection and analysis of data.

Protecting Vulnerable Clients

Human service clients, because they often are involuntary recipients of services, may find themselves vulnerable against pressures to cooperate in research projects conducted by the organizations that provide them with services. Such clients are likely to be sought out as research subjects, either because they often are viewed as being deviant in some way, and therefore interesting to study, or because they may be easier than other groups to locate and keep track of during a research project.

Welfare recipients, children in day care settings, patients in public mental hospitals, and participants in job-training programs, to name only a few, are likely candidates for participation in social research projects. In fact, it is common for operators of new or special programs to be required by their funding sources to research the effects of their programs as a condition for receiving those funds. Although such safeguards as codes of ethics and governmental regulations may serve as useful guides, special obligations fall on human service practitioners to safeguard their clients from unreasonable pressures to participate in research.

A crucial issue in this regard is the matter of voluntary and informed consent. Can clients actually give consent freely? This is one of the reasons that much research using prison inmates as subjects has been discontinued: It is debatable how free inmates really are to give consent. If participation in the research brings significant rewards in the form of separate living quarters or greater privacy, these rewards may be almost coercive in the Spartan, degrading, and often dangerous world of the prison inmate. Although participation in the research is, on the face of it, voluntary, inmates may not seriously weigh the disadvantages or dangers of the research when their participation is perceived as a means of avoiding assault or rape.

Similar ethical questions also arise with research on people who have significant psychopathologies. If the psychopathology involves defective comprehension or impaired insight, then the subjects may be unable to give truly informed consent. Research on such a population would have to be approached very carefully and could be justified only if it could not be done on a less vulnerable group. If the same research goals could be accomplished with a less vulnerable group, then that would be the route to follow. If it were decided to do research on a group with significant psychopathology, then informed consent should be approached in ways that protect potential participants from even covert coercion. For example, consent could be sought by some party other than the researchers (to avoid the force of authority), or it could be sought in the presence of a relative, friend, or other advocate for the individual (to provide the social support to allow a refusal). If the psychopathology completely impairs the ability to give consent, then most forms of research would be unethical unless it could be shown that either the patient or society would benefit significantly and the research could be done in no other way (Oeye, Bjelland, & Skorpen, 2007).

The problem of voluntary consent may be somewhat more subtle among other recipients of human services. Even if refusal to participate is not linked to termination of benefits, clients might not be sure of this, and thus might be disinclined to take the risk of finding out. In addition, the reading level of clients receiving public assistance often is less than that of the eighth grade; yet the reading level necessary to comprehend many welfare documents is above 13 years of education. Thus it may be that clients who are accustomed to being confused by welfare requirements might not aggressively seek to determine their right to refuse participation in a research project. We are not suggesting that research never be conducted on prisoners or recipients of public welfare. Rather, clinician—researchers need to exercise additional caution—and, in some cases, possibly forgo valuable research projects—in the interests of ensuring truly voluntary informed consent.

Research in Practice 3.2 discusses ethical issues involving harm, misconduct, and protecting vulnerable clients and what can be done to protect clients who might be negatively affected.

Withholding Treatment for Research Purposes

An issue of particular concern to human service providers is the research practice of withholding treatment from a control group to assess whether a given treatment is effective. The control group serves as a comparison group. If a group receiving the treatment shows more improvement than the control group over a certain span of time, we can feel justified in claiming that the treatment brought about the improvement. With no control group, however, we cannot say with confidence that the improvement shown by the treatment group was the result of the treatment. It could be that the improvement would have occurred even in the absence of the treatment. The control group, which is comparable to the treatment group in all ways except that it does not receive the treatment, makes it possible to rule out this possibility. Control groups thus are very important to the ability to state whether an independent variable causes change in a dependent variable (see Chapters 2 and 10). In fact, the evidence-based practice approaches discussed in Chapter 1 give the most credence to evidence resulting from research designs that use random assignment of people to treatment and control groups. These EBP approaches argue that such research designs provide the “best” evidence.

[image]

RESEARCH IN PRACTICE 3.2 Program Evaluation: Vulnerable Clients at Risk

Science must maintain constant vigilance against research that harms people or is a product of fraud and misconduct. This is especially true in applied research where social or medical interventions may affect people's lives. Several recent cases of such ethical violation demonstrate this point. In 1996, the internationally respected researcher Dr. Francis Collins, famous for his work on the government's project to map all human genes, retracted five research papers because a junior colleague had fabricated data. The studies, which were on leukemia, had been published in leading scientific journals (Altman, 1996). Another case involved the cellular biologist Robert P. Liburdy, who published studies purporting to demonstrate a link between exposure to electromagnetic fields and cancer. Eventually, the Office of Research Integrity of the DHHS concluded that much of his data was fictitious, but not before several multimillion-dollar liability suits had been initiated against power companies (Federal Register, 1999).

When such fraud occurs in the context of human service research, the consequences can be especially chilling. The case of Dr. Stephen Breuning came to light in the 1980s. He had published research alleging beneficial effects of certain drugs for the control of behavior in severely developmentally disabled individuals. Social interventions were implemented to help many developmentally disabled children based on the results of his research. Yet, Breuning's results were based, in part, on falsified data. Eventually, the case led to a criminal indictment, and Dr. Breuning was sentenced to 60 days in a halfway house and five years of probation. However, detecting, documenting, and acting on this fraud was a long battle (Garfield & Welljams-Dorof, 1990).

A recent example of ethical violations in the protection of human subjects in applied research comes from the University of Oklahoma, where it was alleged that the principal researcher and an oversight official repeatedly violated federal regulations meant to protect human subjects in cancer research (Pound, 2000). These researchers reportedly were unqualified to conduct the research, the vaccine injected had not been properly tested, and the IRB had failed to monitor the tests. What makes this case especially significant for this illustration is not only the nature of the violations, but also the manner in which the ethical violations came to light. It was not a high-level scientist or an IRB that detected the ethical problems; rather, it was a registered nurse with a limited background in clinical research who had been recently hired at the project and was troubled by what she saw as violations of safety rules, suppression of information, and lying to patients and federal regulators. When her university superiors failed to address her concerns, she didn't give up; instead, she advocated for the patients by writing to federal health regulators. As a consequence, government regulators shut down all government-funded research at the Tulsa campus and directed the university to address the problems. The case emphasizes the importance for all human service professionals, not just scientists, to understand the principles of research and research ethics. Although cases of ethical violation in social and medical research may be rare, detection and prevention of such conduct may well depend on the vigilance of the human service practitioner who encounters the subjects of research.

These cases highlight several disturbing issues in ethics, research, and human service practice. First, some of this research managed to pass through the filters of review panels and institutional procedures without question. How well can the institutionalized research system prevent fraudulent work? Second, when unethical practices are discovered, who bears the responsibility for alerting practitioners and the public to potentially harmful implications? In manufacturing, the government can require product recalls, but how do we protect the public from the implementation of fraudulent research? Third, how do we safeguard those who detect potential fraud? In many cases, the people most likely to discover such fraud are colleagues, assistants, students, and others close to the research. To the extent that these people perceive disclosure as risking unpleasant consequences for themselves, they will be reluctant to come forward. Finally, the slow, plodding nature of the bureaucratic process raises the question of what can be done to process cases efficiently while also protecting the rights of individuals accused of fraud.

A close study of the Breuning case suggests that the procedures of the scientific process do afford some protection against the widespread impact of fraudulent data on both science and practice. Based on the assumption that influential research will be widely cited in other research studies, an ingenious study used the Science Citation Index and the Social Science Citation Index to examine the number of citations made by other researchers to Breuning's20 publications between 1980 and 1988 (Garfield & Welljams-Dorof, 1990). The 200 citations found suggested that his work was influential. However, when the researchers evaluated who had made the citations and the context of the citations to better understand the effects of the Breuning studies, things looked a bit different. For example, of the 200 citations, 40 percent were self-citations by Breuning or his coauthors. In addition, following the publication of a critical review of Breuning's work, citations to his work by others declined sharply. Finally, of 38 studies judged to be materially affected by Breuning's publications, 21 led to negative conclusions, disagreeing with his findings or criticizing his methods. The authors concluded that the pattern of citations suggests the process of scientific publication and critique is capable of purging science, to a degree, of fraudulent studies.

Presumably, research is being conducted on some treatment because that treatment is believed to be effective. Herein lies the ethical dilemma: Some service providers believe it is unethical to withhold a treatment that might help people. They believe that withholding treatment deprives people in the control group of the possibility of improvement. This is a serious problem that is not easy to resolve, but there are a number of issues to consider: First, we might ask whether it is ethical to use untested treatments. We pointed out in Chapter 1 that evidence-based approaches to human service professions explicitly caution against the use of treatments without proven effectiveness. Providing services is expensive and time-consuming, and it also raises clients' expectations for improvement. Is it ethical to do this when no evidence documents that the treatment will be beneficial? For the most part, we would not think of marketing new medicines without a thorough test of their effects, both positive and negative. We should expect no less from the human services we offer.

Alternatives sometimes can be found to withholding treatment. One alternative would be to offer the control group a treatment that is known to be effective and then see whether the new treatment provides more or less improvement. In this way, all clients are receiving a treatment that is either known to be or suspected of being effective. In the study of the effectiveness of suicide intervention programs, for example, one certainly would be reluctant to evaluate a new intervention by using a control group that received no intervention at all. One study of outpatient interventions targeting suicidal young adults used a randomized experiment in which subjects were assigned either to a new experimental treatment or to a control group that received a treatment that had been commonly used for these interventions. Although both groups showed improvement, the experimental treatment was more effective in retaining the highest-risk participants (Rudd, Rajab, & Orman, 1996).

Another alternative to withholding treatment would be to delay giving the treatment to the control group and make the comparisons between those receiving treatment and the control group over this period of time. For example, in a study of two different approaches to controlling people's smoking, all participants in the study were told that, if they were randomly placed in the group that would not receive immediate treatment, not only would they receive treatment when the study was over, they would receive whichever treatment the study showed to be most effective (Coehlo, 1983). This resolves the ethical problem by ensuring that all participants will receive treatment at some point.

[image]

INFORMATION TECHNOLOGY IN RESEARCH Ethical Challenges in Internet-Based Research

The Internet offers a vast new arena in which to conduct social science research. After all, the online interactions of people represent yet another setting in which human beings engage with one another, and these interactions often are readily accessible and convenient to researchers. However, not all people warmly welcome scientific scrutiny in this venue. Although it has great promise, use of the Internet as a vehicle for research raises new and complicated questions about the application of ethical principles, such as informed consent and privacy.

Eysenbach and Till (2001) use the term “Internet communities” to refer to the various ways people now communicate using the web. Included are discussion boards on websites, chat rooms, social networking sites, blogs, and newsgroups. All these venues could be settings in which social researchers gather data. Some of these communities address sensitive issues such as dealing with breast cancer or surviving childhood sexual abuse. In assessing what constitutes ethical research in such communities, most professional research organizations, such as the Association of Internet Researchers (Ess et al., 2002) and the American Sociological Association (1997), would agree that informed consent is required when the behavior is sensitive and occurs in a private setting. Conversely, it is ethical to collect data from people when their behavior is public and nonsensitive without gaining informed consent.

However, with research on Internet communities, the complexities only begin with such a straightforward statement because what is considered “public” on the Internet can be complicated and ambiguous. In fact, one ethicist (Sveningsson Elm, 2009) has argued that the terms “public” and “private,” when referencing the Internet, represent not a dichotomy but a continuum. In other words, we have to ask how public or how private an Internet venue is. So some online environments seem clearly public (such as chat rooms or web pages) because anyone can access them. Social networking sites, on the other hand, might be called semipublic because they involve some registration or membership before one can use, or fully use, the site. Some Internet venues are available (sometimes after registration) only to members of a particular company or organization, such as parts of a company's website, and these might be called semi-private websites. Finally, private online settings, such as private rooms within chat rooms, are open only to people who are invited.

Generally, researchers would agree that public Internet sites do not require informed consent while private sites do. The other two possibilities call for more judgment and assessment. In making these decisions, researchers might call on fellow researchers or ethicists to help them explore the issues, as well as seek guidance from IRBs. Considering the thoughts of ethicists such as Eysenbach and Till and Sveningsson Elm, we can suggest some considerations in deciding whether to conduct research online or whether informed consent is required:

[image] Intrusiveness—Will the research involve passive analysis of Internet postings versus active involvement of the researcher in the community by participating in communications?

[image] Perceived privacy—Do participants perceive their community as private, possibly because it is a closed group requiring registration or because group norms define what transpires there as private?

[image] Vulnerability and harm—Will the intrusion of the researcher or publication of results potentially upset or bring harm to individuals or the community as a whole?

[image] Confidentiality and anonymity—Can confidentiality or anonymity truly be protected given the potential of online search technology (if verbatim quotes are used in a research report, search engines may be able to locate the source of the quote)?

When a researcher does want to obtain informed consent in such online research, how does one go about doing so? One suggestion is to send a message to the mailing list describing the research and giving participants the opportunity to withdraw. A second option is to retrospectively ask persons whose postings will be used to give permission. In terms of privacy, Eysenbach and Till point out a special wrinkle that can thwart privacy. A researcher may judiciously separate identifiers from comments, but if the report includes verbatim quotes, then the original participant comments sometimes can be located on the Internet by anyone using a search engine.

Thus, research via the Internet is a new field where social scientists, IRBs, and funding agencies are confronting new challenges related to ethics and finding ways to address basic ethical issues in this rapidly changing medium.

In cases without such alternatives, however, clinician—researchers must use the risk—benefit approach: Does the benefit to be gained from the research outweigh the risks of withholding treatment from the people in the control group? If we were studying people at high risk of suicide, we would be cautious about withholding treatment. In the treatment of nonassertiveness, on the other hand, we might decide that a delay of a few weeks in treatment would not be terribly detrimental and that whatever damage occurred could be undone once treatment was initiated. As with so many other ethical issues, this is ultimately a judgmental one over which clinician—researchers will continue to agonize.

· Notebook

· My Bookshelf

· TOC/Annotation menu

· Downloads

· Print

· Search

· Profile

· Help

3.3 Codes of Ethics

Previous section

Next section

3.3 CODES OF ETHICS

A point emphasized in this chapter is that ethical judgments are difficult and often controversial because they involve interpretation and assessment. Most professional organizations establish written codes of ethics to serve as guides for their members to follow. These codes by no means settle all debate, but they do stand as a foundation from which professionals can begin to formulate ethical decisions. These codes can be found at the websites and in publications of the major organizations of professionals who conduct research in the social sciences and human services. See, for example, the web pages of these organizations:

[image] American Sociological Association ( http://www.asanet.org )

[image] National Association of Social Workers ( http://www.naswdc.org )

[image] Association for Applied and Clinical Sociology ( http://www.aacsnet.net )

[image] American Association for Public Opinion Research ( http://www.aapor.org )

The codes of ethics of other social science and human service professions would be similar to these.

· Notebook

· My Bookshelf

· TOC/Annotation menu

· Downloads

· Print

· Search

· Profile

· Help

3.4 Review and Critical Thinking

Previous section

Next section

3.4 REVIEW AND CRITICAL THINKING

Main Points

[image] What is ethical in research and practice is based on human values and varies as those values change.

[image] The mistreatment of minorities in research has been a major impetus to the development of ethical standards for research in the United States and abroad.

[image] Social research typically is evaluated in terms of risks versus benefits, with some “questionable practices” allowed if the research promises sufficient benefits.

[image] Informed consent assures participants the ability to determine their own destiny, but may prevent the use of some research methods and force the study of only those persons who volunteer.

[image] Some level of deception is sometimes used in research, but whether or under what conditions deception is acceptable has been highly controversial.

[image] Research subjects have a right to privacy, and it can be protected by letting them edit data about themselves from a data set, keeping the data anonymous, or keeping the data confidential.

[image] Confidentiality and anonymity are not the same thing. Intrusion by third parties, such as courts of law, occasionally can be a threat to the guarantee of confidentiality.

[image] Exposing subjects to physical or mental distress should be kept to a minimum and should never be done without fully informed consent. Subjects should be thoroughly debriefed at the conclusion of the research.

[image] When conducting research for a sponsor, many ethical difficulties can be avoided by a detailed research agreement that covers and reveals topics such as the purpose and nature of the research, the rights of publication, and the sponsor of the research.

[image] Researchers have an obligation to report their results fully and honestly and avoid any kind of scientific misconduct or fraud. Replication is a major tool of science for correcting research errors and fraudulent reports.

[image] It is the researcher's own decision regarding the degree to which he or she will become an advocate; however, caution is required so that objectivity is not undermined.

[image] Because of their disadvantaged status, human service clients often are vulnerable to coercion to participate in research projects and, therefore, require special protection.

[image] Withholding treatment from control groups raises an ethical dilemma, but it often is justified when testing unproved approaches.

[image] Most professional organizations have established codes of ethics that provide useful guidelines for making ethical decisions.

Important Terms for Review

misconduct (scientific)

privacy

Critical Thinking

Along with human service practitioners and policy makers, people in their everyday lives grapple with some of the same issues that are described as research ethics issues in this chapter. Just as researchers operate in a context of moral principles and cultural values, all people have moral, philosophical, or theological beliefs that tell them what constitutes proper behavior. Everyone must apply these beliefs or standards to particular situations in deciding how to behave. You can learn some lessons from this chapter in terms of how to critically analyze ethical issues in practice and policy settings as well as in your everyday life:

1. In your everyday relationships with other people, how do things like informed consent, deception, confidentiality, privacy, and the other ethical issues discussed in this chapter come into play, especially as they relate to assessing information? Give an example of each from your everyday life or from practice or policy realms.

2. How do we regulate ethical behavior in our everyday lives?

3. What “codes of ethics” do we rely on in order to do this? How are these codes of ethics similar to or different from the codes of ethics of the professional organizations discussed in the chapter?

Exploring the Internet

Use the search engines on the Internet to locate the web pages of professional organizations other than those already mentioned in the text—research organizations, such as the American Psychological Association or the American Society of Criminology, or practitioner organizations, such as the American Medical Association or the American Nurses Association. Explore those sites for topics related to ethics. Also use the search engine to look for a variety of key words such as “ethics,” “science,” “research fraud,” “IRB,” or some combination of these.

A site that is especially informative about IRBs is http://www.irbforum.org, which is the IRB Forum, maintained by a university research institute and a hospital in Pennsylvania. It has links to wide-ranging resources on IRBs. An invaluable source for ethics information is the Office of Research Integrity (http://www.ori.hhs.gov), located in the Office of Public Health and Science in the DHHS. The mission of the ORI is to promote integrity in biomedical and behavioral research supported by the PHS. You can access the full text of the Belmont Report, the Nuremburg Code, and other major documents.

The NIH has an extremely useful website where ethics materials can be downloaded: The NIH guidebooks cover practically all the issues discussed in this chapter, including certificates of confidentiality. Universities often present their guidelines for research on their websites, so you can learn how their IRBs operate and how they protect human subjects. If you go to a university's web page, you may find a link such as “Guidelines for Researchers” or “Research Policies and Procedures,” with issues related to the ethical conduct of research.

Another website is called Ethics in Science (http://www.chem.vt.edu/chem-ed/ethics). This site has links to many resources and other sites related to ethical issues in research. Many of the links relate to the physical and natural sciences. Some apply to the social sciences, however, and many ethical issues span disciplines. A highly recommended Canadian site that does a particularly nice job of explaining key ethics terms is the National Council on Ethics in Human Research (http://www.ncehr-cnerh.org). Another Canadian site especially useful for ethics information related to research on children and families is the Centre for Children and Families in the Justice System (http://www.lfcc.on.ca/ethical.html).

For Further Reading

Caplan, A. L., Ed. (1992). When Medicine Went Mad. Totowa, N.J.: Humana Press. This volume contains articles relating to the research conducted by the Nazis during World War II. Although it deals with an extreme situation, it does provide an illustration of what can happen when science “goes mad.”

Corey, G., Corey, M. S., & Callanan, P. (2011). Issues and Ethics in the Helping Professions. 8th ed. Belmont, Calif.: Brooks/Cole Cengage Learning. A comprehensive review of ethical issues in the many realms of human service practice. Given the linkage between research and practice, the analysis of practice ethics is relevant to clinician—researchers in the human services.

Crossen, C. (1994). Tainted Truth: The Manipulation of Fact in America. New York: Simon & Schuster. An enlightening book by a journalist about the many frauds that can be perpetrated by scientists. The author shows how frauds can be artfully crafted, and how they can adversely affect us all.

Forbes Citro, C., Ilgen, D. R., & Bagle, C. (2003). Protecting Participants and Facilitating Social and Behavioral Sciences Research. Washington D.C.: The National Academies Press. This report examines three key issues related to human participation in social research: obtaining informed, voluntary consent from prospective participants; guaranteeing the confidentiality of information collected from participants; and using appropriate review procedures for “minimal-risk” research.

Lee, R. (1993). Doing Research on Sensitive Topics. Newbury Park, Calif.: Sage. Ethical issues become especially important and complicated when doing research on sensitive topics, and this author suggests ethical guidelines for navigating such treacherous waters.

Mertens, D. M., & Ginsberg, P. E., Eds. (2009). The Handbook of Social Research Ethics. Thousand Oaks, Calif.: Sage. The handbook provides a comprehensive understanding of the history, theory, philosophy, and implementation of applied social research ethics through a collection of writings by eminent, international scholars across the social and behavioral sciences and education. It also examines the ethical dilemmas that arise in the relationship between research practice and social justice issues.

Washington, H. A. (2006). Medical Apartheid: The Dark History of Medical Experimentation on Black Americans from Colonial Times to the Present. New York: Doubleday. The author provides a comprehensive history of medical experimentation on African Americans. The book details the way both slaves and freedmen were used in hospitals for experiments conducted without a hint of informed consent—a tradition that the author argues continues today within some black populations.

Zimbardo, P. G. (2007). The Lucifer Effect: Understanding How Good People Turn Evil. New York: Random House. The researcher who conducted the now famous Stanford prisoner experiment addresses the question: What makes good people do bad things? Although research ethics is not the primary subject, the author's discussion of his controversial study provides rich illustrations of research ethics.

Class Exercises for Competency Assessment

3.1. [Ethics] [Critical Thinking] [Research & Practice] A criminal justice major needs to do a research project as a course requirement. The student is working part-time as an undercover store detective in a large discount department store, and he proposes to do observational research on the shoplifting behavior he encounters as a part of his job. In addition, he has access to store records on suspects who have been apprehended for shoplifting in the past. He also plans to interview a sample of shoppers about shoplifting without disclosing that he is a store detective. Discuss how the issues of informed consent, confidentiality, and disclosure of results apply to this case.

3.2. What recommendations could you make to the student in Exercise 3.1 to safeguard the ethical standards for research in this case? Consider each of the ethical issues discussed in this chapter. Do you think the study could be done ethically at all? Support your conclusion.

3.3. [Ethics] [Critical Thinking] [Research & Practice] Suppose you are working in an alcoholism treatment unit, and a proposal is made to initiate a new treatment program that looks promising but is largely untested. Under what conditions do you think it would be acceptable to use a control group as a comparison group that does not receive any treatment? Under what conditions would it be ethically unacceptable? How might you avoid some of the ethical dilemmas that come from withholding treatment but still use some form of control group for comparison?

3.4. [Ethics] [Critical Thinking] [Research & Practice] In addition to her faculty position, Professor Courtney Thompson receives a contract with a local clinic to take over running a therapy group for women survivors of intimate partner violence. The program has been in operation for several years. The agency protocol for the group requires the participants to complete an intake questionnaire and an exit questionnaire, both of which include a depression inventory and an anxiety scale. They also complete a follow-up telephone interview at three months after treatment. The information is used by the agency to compile its service statistics and evaluate the program. Dr. Thompson has access to the clinic data from previous years. What do you see as ethical issues that she must address if she wants to use this data for the purpose of publishing research articles in professional journals?

3.5. [Ethics] [Critical Thinking] [Research & Practice] Although the routinely collected data are useful for her research, Dr. Thompson proposes adding an additional scale for measuring posttraumatic stress. Although valuable for her research, the additional scale is not required for the agency's purpose. Could Dr. Thompson use the additional scale? If so, what ethical issues would need to be addressed?

3.6. [Ethics] [Research & Practice] Invite a representative of the IRB at your college or university to address your class on the kinds of ethical issues that arise in research done by faculty or students. Explore with this person the safeguards that can be built into research designs to protect against such ethical problems.

· Notebook

3.4 REVIEW AND CRITICAL THINKING

Main Points