SEXUAL ABUSE during childhood is a serious problem with many long-term consequences. One of these consequences is the increased risk for adolescent motherhood.1-3 The National Institutes of Health (NIH) funds much of the research conducted in the United States on the topics of adolescent pregnancy and child abuse and neglect. The funding allocation estimates for fiscal year 2007 was $21 million for adolescent pregnancy research and $37 million for child abuse and neglect research. Similar moneys were allocated again for fiscal year 2008.4 In 1996, the House and Senate Appropriations Committee issued a congressional directive to the NIH to include children in research. In 1998, the NIH issued a policy to include individuals younger than 21 years (defined as children) in research unless ethical reasons or scientific evidence indicated otherwise.5,6 Despite the seriousness of childhood sexual abuse (CSA), the potential consequence of adolescent motherhood, the amount of money allocated each year to these problems, a directive from the Congress, and a mandate from the NIH, CSA among adolescent mothers remains a difficult topic to study. The purpose of this article is to discuss the reasons behind these difficulties, the context in which research is proposed, carried out, and reported, and factors to consider when planning studies.

TOPIC SENSITIVITY

CSA among adolescent mothers is difficult to study because of topic sensitivity. Lee7 developed a conceptual model to categorize sensitive topics, assess for conditions of sensitivity, assess for the kinds of threats that studies may pose, and develop strategies to minimize those threats. Discussion of the entire conceptual model is beyond the scope of this article, but Lee's7 method of categorizing sensitive topics and assessing for conditions of sensitivity are discussed. Sensitive topics are defined as subject matter that, when studied, may be perceived by participants as threatening.7 Threatening research involves potential costs to participants. While research often takes time and can be inconvenient, the costs in terms of topic sensitivity go beyond these usual demands. Such extra demands are created by participant concerns about the research itself. Whether this occurs depends on the context of the situation. A topic may be perceived as sensitive in 1 circumstance but not in another, or it may cause an emotional response in 1 person but not in another. Because topic sensitivity is contextual, Lee7 suggested 3 broad conditions under which topic sensitivity occurs. Research may be threatening when it is perceived by participants to be intrusive, stigmatizing, or political. Lee7 labeled these as intrusive threats, threats of sanction, and political threats.

Lee7 described intrusive threats as research that impinges on privacy or creates stress. For this article, privacy is defined as subject matter that may be perceived by participants as confidential, and impingement on privacy is defined as research that may be perceived by participants as interfering, invasive, prying, or meddlesome. Topics regarded as private vary across cultures and situations; therefore, the domain of privacy is not fixed.7 Researchers must be familiar with the culture and the situation of the community in which they are working to know the topics defined as private by potential participants. Queries about personal finances or sexuality are topics that participants typically regard as sensitive and, therefore, may be reluctant to provide responses on. Lee7 defined stressful topics as those involving emotionally charged personal experiences. These topics may create stress when studied because as participants remember their past experiences, they may also relive their emotional reactions to those experiences. Investigation of some topics may create stress because participants may want to maintain emotional control as they remember and describe their experiences and reactions. Bereavement is an example of a sensitive topic that may create an intrusive threat for some participants. It may be emotionally intrusive because feelings associated with bereavement are intense, inescapable, psychologically painful, and may leave emotional damage after initial feelings of grief subside. Elam and Fenton8 suggested that intrusive topics might be inviolable, therefore, seldom thought about, much less talked about, so participants may have limited vocabulary for discussion.

Sanctioned topics are defined as subject matter related to behaviors that are socially controlled, and may be stigmatizing or incriminating when revealed.7 Research about sanctioned topics may be unwelcome because scrutiny might expose behaviors intended to remain covert. In some instances, these behaviors are innocuous and tacitly accepted as long as they remain hidden, but in other instances, they are harmful and explicitly prohibited. Potential participants may be reluctant to volunteer for studies involving topics about which they may feel regret or fear disapproval, exposure, repercussion, exploitation, or disparagement.7,8 For instance, individuals infected with the human immunodeficiency virus (HIV) may be reluctant to volunteer for a study about HIV for fear that people important to them will learn of the infection.

Politically sensitive topics are defined as subject matter that threatens the balance of power among groups of people.7 Research may produce concerns that a power imbalance will be created by the research process, research results, or researcher. Groups in conflict may regard a topic as threatening because they do not want confidential information disclosed to the opposing group. Groups previously in conflict may have concerns that research will disturb tenuous resolutions that may be in place. Elite groups may not want to share information with their disadvantaged counterparts because this might undermine their privileged position. Vulnerable groups may regard research as threatening because they feel that they have no control over the impact of the study on themselves or their communities. There is some basis for this concern because researchers tend to study down, rather than up, the power gradient.7 Powerful groups may have the resources, expertise, and education to distance themselves from research. These groups also may have the resources to manipulate the circumstances under which the research is conducted so that a positive public image is maintained. This may leave the consumers of research with the idea that only certain groups of people have certain problems, when, in fact, the research is permitted and objectively conducted only with certain groups.

Many topics, including CSA, may fall into more than 1 category. CSA may be categorized as intrusive because individuals may feel guilt, shame, stigma, or embarrassment, and therefore, reluctant to participate in research about their experiences. CSA may be categorized as sanctioned, not because victims want to keep it a secret, but because individuals, families, and society may not want to know about it. Van Dam9^(p37) referred to CSA as the "unwanted discovery" and explained that families and communities will commonly blame the victim and defend the perpetrator when a child discloses sexual abuse. When CSA is discovered, there are legal, moral, and ethical mandates to confront it and prevent further abuse. However, confrontation and prevention are costly. The discovery of CSA is disruptive and may destroy what appeared to be a normal, cohesive, and stable family. Ironically, the appearance of family normalcy is often a veneer created by the perpetrator for self-protection.9,10

Finally, CSA may be categorized as a political topic because researchers are encouraged to look at poor and underserved groups of people with the aim of providing them with more and better healthcare. The study of CSA primarily among these groups may create the image that this is a problem created by poverty. In a historical review, Olafson et al11 reported that researchers have ignored CSA among people with money and power who are presumed to be respectable and thus would not want to engage in sex with children. These authors also pointed out that sometimes the people with money and power perpetuate the image of socioeconomic disparity because this facilitates their use of children for sex. This tactic keeps attention off them and their socioeconomic group.

AGE OF CONSENT

CSA is difficult to study because some adolescent mothers are younger than 18 years. Eighteen is the age of majority in most states.12 The exceptions are Nebraska and Mississippi where the age of majority is 19 and 21, respectively. Persons younger than 18 years are minors. Minors tend to be thought of as children, but the terms minor and child are not synonymous. Santelli et al6 argued that for purposes of confidential and anonymous survey research, adolescent minors should not be regarded as children because research indicates that adolescents aged 14 years and older are capable of understanding conflicting viewpoints and making sound independent decisions similar to those of adults. These authors further argued for encouraging and respecting this emerging capacity. The NIH policy mandating the inclusion of children in research defined a child as anyone younger than 21 years, but the NIH deferred to the individual state definitions of child regarding age of consent.5 In the context of research, age of consent is the age at which individuals can grant their own permission for participation. Eighteen is the age of majority in most states, so this is also the age of consent.12 Researchers need to be familiar with the child maltreatment laws in the states or territories where their research will be conducted.

Emancipation means to become an adult and free from the authority, control, and custody of parents or guardians.13 This occurs automatically at the age of 18 years, the age of majority and consent. While pregnant and parenting minor adolescents have rights similar to those of adults with respect to their own children, motherhood does not confer emancipation.14 The status of some adolescent mothers as unemancipated minors causes research difficulties. First, unemancipated minors are supposed to obtain parent or guardian permission before research participation. Emancipation may be obtained by court order,14 but most minor adolescent mothers do not pursue court-ordered emancipation. This places them in a paradoxical situation with all of the responsibilities of motherhood but not all the legal rights of adulthood. For instance, minor adolescent mothers can consent for their children's participation in research, but they cannot consent for themselves.

Second, parental secrecy may masquerade as parental privacy. While most parents and guardians keep their children's interests in mind and should be involved in research decisions,15,16 abusive parents may protect family privacy to prevent abuse disclosure.16 This likelihood increases among adolescent mothers. Almost one half of adolescent mothers report CSA histories2,3,17 in comparison to one third of adult women.1 Sexually abusive parents or guardians are unlikely to consent to research about behaviors intended to remain surreptitious. Coercion is sometimes used to keep sexual abuse secret. A mere interest in study involvement may place adolescents or their children at risk for physical, emotional, or sexual retaliation. Moreover, adolescents have a right to privacy about their sexuality.6 Parental consent threatens that right.

Third, healthcare workers are mandated reporters of child maltreatment.18 This creates a dilemma. If researchers report abuse volunteered by their participants, then this violates the trust, privacy, anonymity, and confidentiality of those participants. If researchers explain that abuse will be reported, then only those willing to disclose or who misunderstand the conditions of participation will volunteer. Adolescents unwilling to disclose will opt out of the study causing an over representation of nonabused participants and loss of valuable data. If researchers do not report, they may be charged with failure to meet statutory obligations. Forty-six states, the District of Columbia, and several territories have child maltreatment statutes that make failure to report a misdemeanor in most of them and a felony in a few of them.19 If researchers report their suspicions of abuse, but those suspicions are wrong, they and the institutional review board (IRB) requiring reporting may be liable for damages, as researchers are unprotected by reporting laws.20

GAZE AVERSION

CSA among adolescent mothers is difficult to study because of gaze aversion. Gazeaversion was the term used by Krugman and Leventhal21 to discuss the tendency of people to distance themselves from sexual abuse. Gaze aversion for this article means to avoid, deny, ignore, or disregard sexual activity between adults and children. Literature regarding gaze aversion is scattered, and other terms have been used to describe the concept. Olafson et al11 referred to gaze aversion as "cultural denial"11^(p8) and indicated that CSA tends to be avoided because knowledge of the commonness and noxiousness of it "darkens" perceptions of human nature.11^(p19) These authors also explained that CSA has come into public and professional awareness many times but was repeatedly suppressed because of the negative reactions to it. For instance, during the 20th century, victims and mothers of victims rather than perpetrators were blamed for CSA, child advocates were labeled as abnormal for focusing on CSA, and most CSA was relegated to occurrences that happen outside of respectable families. Summit referred to the avoidance of CSA as a "mythic insulator"22^(p177) and a "protective denial."22^(p179) He explained that insulation and denial were natural adult responses to children's pain. He also postulated that children and adults react differently to abuse. Rather than report the abuser and tell their story with a coherent beginning, middle, and end as adults might, children are secretive and helpless and learn to accept their situations. When children disclose sexual abuse, they tell disjointed bits and pieces of their experiences, and they repeatedly disclose and recant. While this method of disclosure is a normal response of children to an abnormal situation, it undermines believability and reinforces the adult need for protection from the truth.

Child predators use gaze aversion to their advantage. Lawson23 conducted interviews with 453 incarcerated child predators and found that many of them counted on at least some tolerance regarding their sexual interests in children. Some of the roadblocks that IRBs might create in response to sexual abuse studies may be gaze aversion. Nicholl24 warned that IRBs might expand their boundaries to science and legal issues instead of ethical issues to avoid the topic. Hays et al25 reported that their IRB found no reason to study intellectually challenged convicted sex offenders despite study justification, funding availability, and an independent evaluation of their proposal by psychology experts. During the 3 years required to gain study approval, these researchers were confronted with bureaucratic and study design issues rather than ethical issues.

There are several reasons for gaze aversion. Perpetrators may want the topic of CSA avoided because quiescence allows them to continue to victimize children. Bystanders may want the topic avoided because they may not want to get involved or were victims of maltreatment themselves and feel helpless. Others may want the topic avoided because they find it unbelievable and repugnant. Krugman and Leventhal21 suggested that CSA is avoided because of the magnitude of the problem and the perception that it only affects others. They suggested that direct confrontation requires an effective child protective system, a clear governmental approach to research, services for victims and families, and funding commensurate with the problem.

CONSIDERATIONS FOR FUTURE STUDIES

Difficulties created by topic sensitivity, the legal status of some adolescent mothers as minors, and gaze aversion cannot be eliminated, but there are ways to diminish these difficulties. Hern et al26 suggested that to gain IRB approval and access to adolescents for recruitment, researchers need to write clearly and persuasively about CSA-related problems and justify every aspect of their studies. In addition, the use of valid, reliable, and age-appropriate instruments and involvement of gatekeepers and stakeholders are important for success in navigating the IRB approval process and for completing data collection. Hays et al25 suggested that before beginning a new proposal, researchers consider the people most appropriate to contact regarding data collection sites, study approval, and participant access.

Trust is important in the study of sensitive topics. Researchers need to establish trust with gatekeepers and stakeholders. Gatekeepers are individuals whose approval is necessary to gain access to participants. They may or may not be in positions of authority. A gatekeeper may be the director of an agency, a receptionist, or an influential community member. Stakeholders are individuals who have an interest in adolescents or study findings. Like gatekeepers, stakeholders may or may not hold positions of authority. Moreover, gatekeepers and stakeholders may or may not be one and the same person. The term gatekeeper will be used in this article because the strategies for gaining access to potential participants are similar whether gatekeepers or stakeholders are involved.

Adolescent mothers make up a vulnerable population, so gatekeepers may be protective of them, and, therefore, researchers may encounter a variety of reactions to their studies. One reaction may be support in principle but not in practice. Support in principle means that gatekeepers understand the value of CSA studies and support this type of research but withhold approval for the study to be conducted in their facilities. Support in practice means that gatekeepers not only understand the value of CSA studies and support the research but also grant their approval. There are several reasons that gatekeepers may withhold approval. They may suspect a high prevalence of CSA in the community but have too few resources to handle the problems. They may have concerns about the stigma associated with CSA and the effect of this on the facility or community where data are collected. When gatekeepers are unwilling to grant approval, they may explicitly deny access and explain the reason. Alternatively, they may implicitly deny access by failing to return researcher's phone calls and emails. The researcher needs to determine between missed communication and implicit denial of participant access.

Other gatekeepers may approve CSA studies because the findings can help them work with community members to prevent CSA and solve the problems created by it. Elam and Fenton8 encouraged collaboration between gatekeepers and researchers to develop studies compatible with the host culture, ethnicity, and race so that findings are beneficial to all involved parties. The researcher can facilitate gatekeeper approval by indicating the benefits of knowing the prevalence of CSA in a community and by working with gatekeepers to identify ways that research may be used to solve CSA-related problems.

To gain trust, the researcher and gatekeeper must become acquainted, and the researcher must teach the gatekeeper about the study and its benefits. Gaining trust takes time, patience, and finesse. The initial contact with gatekeepers may be achieved in a variety of ways. Some contacts may be made through casual personal conversations, but more often, contacts are made through phone calls. During the initial call, the researcher needs to describe the study succinctly. If the gatekeeper is interested in the study, the researcher may offer to provide a study abstract or meet with the gatekeeper. This type of abstract is a 1- or 2-paragraph description of the study that usually focuses on recruitment and data collection procedures, but the abstract may be tailored to the needs of the gatekeeper and the agency from which participants will be recruited. The information in the abstract should be pertinent, clear, and concise. Alternatively, the researcher might offer to give a presentation to staff and other key individuals. The length and location of the presentation are negotiable. Presentation attendees may have ideas about approaching potential participants and streamlining data collection procedures to minimize disruptions to normal agency routines.

There are several factors for researchers to consider for future studies regarding IRBs. Board members may be unfamiliar with CSA, adolescent motherhood, and the data relevant to CSA studies. For instance, CSA demographic questions are used to collect data about the frequency, duration, severity, and types of maltreatment; ages of first and last abuse; family reactions; number of perpetrators; number of incidents; and relationship between victim and perpetrator. This information is important because it provides the contextual description in which the CSA occurred27 and may be used during data analysis to determine associations between these and other variables. Some board members may consider CSA demographic information intrusive and disapprove of its collection because of perceived irrelevance, superfluousness, or inappropriateness. For instance, queries about the relationship between victim and perpetrator may be of concern because victims might identify their abusers by name. Victim and perpetrator identification obligates the researcher to report the abuse if the victim is a minor.

Anonymity and confidentiality must be maintained to reduce reporting angst about CSA research. Anonymity means that participants are unidentified and the information provided by individual participants is indistinguishable from the information provided by other participants. Confidentiality means that the identity of individual participants is not divulged. Anonymity and confidentiality are important so that participants feel comfortable about reporting their feelings and experiences without concerns that their identity will be known or that the information they provide will be linked to them. The purpose of anonymity and confidentiality is to encourage truthfulness and openness in the responses that participants provide. Maintaining anonymity and confidentiality protects participants, researchers, and IRBs.

The researcher may need to argue persuasively for the collection of CSA demographic information and should write the procedures section of the proposal so that the identity of the victim and perpetrator are not revealed. The researcher should indicate that the participants will be instructed orally and in writing not to identify themselves or their abusers by name or in any other way on instruments or in interviews. Large Xs may be marked over identification lines on purchased instruments to discourage respondents from completing these items. The assents and consents should be obtained and collected before other portions of the study, and participants should be asked to place completed materials in envelopes that they seal prior to submission. Finally, the signed assents and consents should be stored separate from the collected data and tracking links between the signed assents and consents and collected data eliminated.

Researchers need to anticipate the concerns of IRB members. For example, they may have concerns about the location where the researcher will talk with participants about the study. Some board members may want all aspects of CSA studies discussed in private because of the subject matter. Researchers need to discuss with gatekeepers the locations where privacy may be maintained during recruitment and data collection. This information needs to be written into the IRB proposal. Other board members may have concerns that some questionnaire items may trigger memories of childhood abuse. Researchers should make available to participants the names and phone numbers of counseling centers or referrals to qualified professionals. Participants may be given 24-hour hotline wallet-cards for organizations such as the Rape Abuse and Incest National Network or the National Domestic Violence Hotline. These organizations are accessible via the Internet or by phone to provide free, confidential, anonymous, and secure access to counseling and crisis centers around the clock and across the United States to victims of childhood maltreatment, sexual assault, dating violence, and domestic violence.

The status of some adolescent mothers as unemancipated minors must be addressed with the IRB. A waiver of informed consent may be obtained from the board that asks underage participants to sign assents rather than expect parents or guardians to sign consents. Consent to participate in research may be given only by individuals who have reached the age of majority. Minors can give only their assents. A waiver is important for several reasons.28 First, the parents of the adolescents may be their abusers. Obtaining parental consent might place adolescents or their children at risk for harm when the parents become aware of participation or consideration of participation. Second, parent or guardian consent would reduce the pool of participants with CSA histories and cause an over representation of participants without CSA histories. Third, some studies require collection of information about sexual histories, so parent or guardian consent would threaten adolescent privacy and confidentiality. Finally, adolescent mothers may live autonomously. Waiving consent does not discourage adolescents from seeking counsel but does permit them to choose whom and how much to tell about their sex lives.

If the IRB lacks expertise in a research area, explore with the board chair the possibility of an expert temporarily sitting on the board for review of the proposal. The researcher should be prepared to offer suggestions regarding the person to serve as the temporary board member. Consider attending the IRB proposal review. Some boards do not permit visitors so ask permission prior to the meeting. Carefully weigh the pros and cons of attending the meeting. Meeting attendance may facilitate communication but also may highlight researcher characteristics such as naivete. Consider the concerns that IRB members may introduce during the meeting and rehearse responses.

CONCLUSIONS

The aim of this article was to raise issues that researchers studying CSA among adolescent mothers might encounter and should consider before beginning a study. Although challenges are inevitable, many of them are surmountable with tenacity, creativity, and patience. Keeping CSA and the value of researching CSA in the foreground is important. CSA is easy to ignore as an irrelevant past event because of the potential for discomfort among victims and researchers. While not every CSA survivor will be at ease reporting and discussing their histories, some survivors may welcome the opportunity to participate in research, contribute to the end of CSA, and counteract the effects of their experiences.

REFERENCES