Cover Memo for Cognitive Testing Request

U.S. Department of Justice

Office of Justice Programs

Bureau of Justice Statistics

Washington, D.C. 20531

MEMORANDUM

To: Lynn Murray

Clearance Officer

Policy and Planning Staff

Justice Management Division

Through: James P. Lynch

Director

From: Allen J. Beck, Ph.D., Senior Statistical Advisor

Shannan Catalano, Ph.D., Statistician /Project Manager

Date: December 18, 2012

Re: BJS Request for OMB Clearance for Cognitive Testing under the National Crime Victimization Survey (NCVS) Redesign Generic Clearance, OMB Number 1121-0325.

The Bureau of Justice Statistics (BJS) requests clearance for cognitive interviewing tasks under the OMB generic clearance agreement (OMB Number 1121-0325) for activities related to the National Crime Victimization Survey Redesign Research (NCVS-RR) program. BJS, in consultation with Westat under a cooperative agreement (Award 2011-NV-CX-K074 Methodological Research to Support the National Crime Victimization Survey: Self-Report Data on Rape and Sexual Assault - Pilot Test), has planned to develop and test two different survey designs for collecting self-report data on rape and sexual assault.

In accordance with the generic agreement, BJS is submitting to OMB for clearance of developmental materials associated the Rape and Sexual and Assault (RSA) Pilot Test. BJS will use the approved generic clearance for cognitive interviewing to test instruments, including re-interview instruments, to be used in the pilot test. This initial clearance will be followed by a request to conduct a feasibility test in 2013 and a pilot test in 2014.

1. Purpose of Overall Project

This request is for clearance to conduct cognitive interviews to develop two questionnaires to measure sexual assault and intimate partner violence against women. The questionnaires to be tested will be used in a larger pilot test that compares an in-person survey and a telephone survey. Before describing the proposed cognitive interviews, we describe the goals and research questions for the larger project.

Estimates of rape and sexual assault from the NCVS have been found to be significantly lower than those in several other studies that have focused on these types of crimes. This has raised questions about the quality of the NCVS data. As described by Fisher and Cullen¹, there are significant differences between the NCVS and these other studies. The NCVS begins with an in-person visit, with the respondent being administered a two-stage instrument (crime screener and detailed incident form) that asks about incidents over the previous 6 months. The interviews are either bounded by a previous interview or adjusted to account for the unbounded status as an incoming rotation group.² The screening instrument covers a wide range of predatory victimization, not just sexually related incidents (e.g., burglaries, household and personal theft, aggravated assault, motor vehicle). The context of the NCVS is on “criminal victimization.” The coverage and response rate of the survey is relatively high.

The alternative surveys take a public health approach and focus on interpersonal violence for extended reference periods (lifetime, last 12 months). These surveys cue respondents with explicit reference to behaviors that make up the legal definition of rape and sexual assault. The study is introduced to respondents as one covering ‘health and injuries.’ The survey classifies events according to the initial questions that reference the behaviors, with no follow-up to assess the nature of the event. Most of the studies completed to date have used random digit dialing (RDD) sample designs within a centrally monitored computer-assisted telephone interviewing (CATI) facility. The response and coverage rates are generally lower than the NCVS, but the interviewers are more closely monitored.

The purpose of the main study is to investigate the differences between the NCVS and these other surveys. Specifically, the study has three key objectives:

1. Develop and pilot test an optimal design using Audio Computer Assisted Self Interviewing (ACASI) to collect self-report data on rape and sexual assault.

2. Develop and pilot test a comparison design using Random Digit Dialing (RDD) to collect self-report data on rape and sexual assault.

3. Conduct detailed analytical comparisons of the two designs against each other and the existing NCVS program.

The primary research questions for the project are:

1. What are the differences in data quality of the two approaches?

2. How do the two approaches compare with the current NCVS measures?

3. What are the comparative costs of the two approaches? Given these costs, how would a survey on rape and sexual assault fit within the ongoing NCVS program?

Toward this end, the main study will conduct two separate surveys. One will be an RDD telephone survey that is similar in design to prior telephone surveys (hereafter referred to as the ‘public health surveys’). The other will be an in-person visit that will administer the assault questions with an ACASI Two weeks after the main interviews, a sample of respondents will be re-interviewed using the same protocols as the first interview. This re-interview will be used to estimate the reliability of the two methodologies.

2. Request to Conduct Cognitive Interviews

We are seeking approval to conduct cognitive interviews that will test the telephone and in-person versions of the questionnaires. We will also conduct cognitive interviews for the re-interview. The plan is to recruit women to participate in two interviews, conducted 2 weeks apart.

Both the telephone and in-person interviews have similar structures. This includes:

1. Introductory sections. This includes demographic characteristics and other variables related to risk of victimization. This section is identical across the two modes of interviewing. The biggest difference is the administration of a life-event calendar on the in-person design, which immediately precedes the assault questions.

2. Sexual and IPV screener. This section asks if the respondent has been a victim of different types of assaults. If a report is made, a set of follow-up questions are asked to collect details about the event in order to classify the event (see below). This two stage approach is used by the current NCVS.³ The questions adopt the behaviorally specific approach used by the public health surveys.

3. Detailed incident form. The content of these surveys are identical between the telephone and in-person approaches. As noted above, this will be used to classify the initial reports into a type of crime or, in some cases, as not being in-scope as a criminal victimization. This section will allow analysts to compare the types of events reported by each methodology.

4. Vignettes. This section will administer a series of vignettes that will be used to measure the respondent’s definitions of key concepts related to the occurrence of rape and sexual assault. The intent is to use the vignettes to measure comprehension on the main survey.

5. Re-interview. Respondents will be re-interviewed, approximately 2 weeks after the initial interview.

Recruitment for Cognitive Interviews

We will be conducting cognitive interviews with 60 subjects in English and 20 subjects in Spanish. All participants will be compensated $40 to participate in a 90 minute interview and another $40 to participate in a second, 90 minute, interview. We are using two strategies to recruit women.

Strategy 1:

Ten victims of rape and sexual assault will be recruited. To recruit subjects, DC Rape Crisis Center personnel will approach victims of sexual assault about participating in the survey by providing them with a flyer which provides basic details about the study (Appendix A). To volunteer, women will call a telephone number and will be given information about the interview and asked to set up an appointment.

A time will be set up for Westat personnel to meet the participants at the DC Rape Crisis Center to complete the cognitive interview. At the time of the interview, the Westat interviewer will administer a consent form (Appendix C) and administer the research protocol (Appendices E1 and E2).

Strategy 2:

Seventy women age 18-40 years old will be recruited for the remaining interviews. Younger women are being recruited because they are more likely to have experienced a recent sexual assault. We will be recruiting respondents by posting fliers in locations such as college campuses and advertisements in local newspapers and Craig’s List. (Appendix B). Women volunteering will be asked to call an 800 number. Those calling in will be asked a series of screening questions. Women who have said they have experienced an incident, who are in college and those in the youngest age groups will be given priority when determining eligibility. If they agree, they will be asked to either come to Westat in Rockville or a central location in Washington DC.

Research Design

The overall goal of the interviews is to test the respondents’ comprehension of the survey questions. This will be accomplished through the use of cognitive interview probing techniques. Interviewers will administer the questions as worded and elicit a response from the respondent. All cognitive interviews will be conducted in-person. ACASI questions will be simulated using hardcopy versions of the screens for the respondent to review. The interviewer will read the questions to the respondent, simulating the ACASI voice, and record their responses. For the telephone interviews, an interviewer will read the questions to the respondents simulating a telephone interview. Respondents will be encouraged to think out loud throughout the interview as they answer the questions. Concurrent probes will be administered if the respondent expresses some confusion. Otherwise, at the end of each section, the interviewer will debrief respondents by probing on questions retrospectively. Interviewers will initially probe with open-ended follow-up questions (e.g., “how did you come up with your answer?” or “tell me what you were thinking about when you answered that question”) and focus more specifically on parts of questions as needed.

In order to keep the interviews within the 90 minutes, the cognitive probes will be administered for only a portion of the interview. For half of the respondents, interviewers will probe the event history calendar and the screener (n=40). For the other half of the respondents, no probes will be administered through the screening instrument and detailed probing will occur when the respondent is answering the items on the detailed incident form (n=40).

Screener and Detailed incident questions

The cognitive interviewing will test different variations of some of the screening questions in the ACASI and Telephone instrument. These variations include:

1. Order of questions. Half of the respondents will receive questions ordered from most to least serious, while it will be the opposite for the other half. We are varying the order to investigate the advantages and disadvantages of each approach. Starting with the most serious incidents will reduce duplicate reports across the screening questions. Starting with the least serious assaults has the advantage that it gradually introduces the sensitive topic and explicit language to the respondent.⁴

2. Cues on ‘force’. Different words are used to describe force. One includes ‘made you have sex by using force or threat of harm’, while the other uses the phrase ‘ used any kind of force or threat to make you have sex against your will’. Testing these involves trying to balance keeping the question as simple as possible and the need to elaborate the key concept of force.

3. Detail on direction of sexual act. The direction of oral sex is specified in one version of the question, while the direction is not specified the other version. As with ‘force’ we are balancing keeping questions short with maintaining detail that might be needed to define the event for the respondent.

4. Consent and alcohol use. One version uses the words ‘unable to consent because you were drunk…’, while the other version uses ‘…too drunk to be able to stop what was happening’. The former focuses on not providing consent, while the latter focuses on not being able to deny consent.

5. Defining psychological aggression. The longer version of this question provides several examples to assist defining the meaning of the question. The shorter version summarizes the definition with fewer words. As with the other examples above, we are balancing keeping questions short with maintaining detail that might be needed to define the event for the respondent.

6. Act specific threat questions. The shorter version rolls up the question on attempted acts into one question including all types of assaults. The longer version breaks this into three questions, one for each type of sexual act. As with the other examples above, we are balancing keeping questions short with maintaining detail that might be needed to define the event for the respondent.

7. Placement of lifetime question. The ACASI survey will ask a lifetime question if the respondent does not report a victimization in the last 12 months. The variation to be tested is whether the question comes directly after the specific 12-month question or whether it comes after all of the sexual screeners are asked.

Half of the respondents will be assigned to a group that will order the questions from most to least serious (Questions SV1-8, SV9-13 – Appendix E1) and half will be in a group from least to most serious (Questions SV9-13, SV1-8). The design will cross these two groups with one of two alternative questionnaires that vary the question wording as noted above in points 2 – 7. One of these variations is shown in the Alternative 1 column in Table 1 (‘SV’ questions) and the other variation is shown in the second column (‘Alt’ questions). This 2 x 2 design will result in four different versions of the questionnaire.

Table 1. Variations of ACASI Screening Questions to be Tested During Cognitive Interviews

Design factors to be evaluated in cognitive testing	Alternative 1	Alternative 2
The level of detail used to cue “force”	SV1-4 SV1-4ever	Alt SV1-4 Alt SV1-4ever
The direction of the forced behavior	SV2	Alt SV2
Handling of ‘ability to consent’ in drug and alcohol facilitated sexual assaults	SV5 SV5ever	Alt SV5 Alt SV5ever
Method of cueing for psychological aggression and coercive control	SV6 SV6ever	ALT SV6 ALT SV6ever
The cueing of attempted but not completed behaviors in a combined vs. separate questions	SV8	SV8a – SV8c
Placement of “lifetime” questions (directly after behavior cue, or after full set of behavior cues)	SV1-7ever

* Question numbers refer to the ACASI screener

A very similar design will be used for the telephone screener. The major difference is the telephone screener will not include a variation related to the placement of the lifetime questions (see variation 7 above). Otherwise, all other variations are identical to the ACASI instrument. Table 2 provides the question numbers for each of these variations on the telephone screener. As with the ACASI interviews, respondents will be assigned to one of four different groups, depending on the ordering of the questions by seriousness and the versions of the questions (see columns 1 and 2 of Table 2).

Table 2. Variations of Telephone Screening Questions to be Tested During Cognitive Interviews

Design factors to be evaluated in cognitive testing	Alternative 1	Alternative 2
The level of detail used to cue “force”	SV1 SV2 SV3 SV4	Alt SV1 Alt SV2 Alt SV3 Alt SV4
Handling of ‘ability to consent’ in drug and alcohol facilitated sexual assaults	SV5	Alt SV5
The direction of the forced behavior	SV2	Alt SV2
Method of cueing for psychological aggression and coercive control	SV6	ALT SV6
The cueing of attempted but not completed behaviors in a combined vs. separate questions	SV8	SV8a – SV8c

Detailed incident form

The detailed incident form is identical for both modes. Appendix F provides the instrument, along with the cognitive protocol for those respondents who will be probed on this portion of the instrument. Respondents in high risk groups (young females) are being recruited for the interviews. However, we anticipate there will be a number of individuals who do not report an incident on the screener. These individuals will be asked to go through the detailed incident form using one of several hypothetical scenarios (Appendix G).

Vignettes

The telephone and in-person interviews in the main survey will measure respondents’ interpretation of key definitions through the use of vignettes. The cognitive interviews will be used to develop and pretest these vignettes.

Respondent interpretation of the initial screening questions on a sexual assault survey is critical to what is eventually reported.⁵ The use of behavioral specific descriptions is the preferred method, as this is linked to more objective criteria related to what happened, rather than reference to terms such as ‘rape’ or ‘assault’. However, even behavioral descriptions contain general concepts that allow some judgment on the part of the respondent. Terms such as ‘force’ or ‘intoxicated’ rely on the respondent making a judgment. The use of vignettes will allow the study to explore the boundaries of how respondents are framing key concepts. Vignettes are brief stories that describe hypothetical characters and situations to which a respondent is asked to react.⁶ For the pilot study, respondents will be administered a number of brief vignettes describing different sexual assault scenarios and then answer a survey question from the perspective of one of the actors in the vignette. By systematically manipulating features of the vignette we will better understand the types of events respondents are including when answering the survey questions. This will provide a measure of the content validity of the measures and test whether there is variation in definitions across the two methodologies.

The vignettes will be used to examine variation across two concepts included on the screening questionnaires (Appendix H). One is the role alcohol and drugs play in defining the event as an assault. The link between alcohol/drugs to sexual assault is difficult to measure on a self-report survey. In the context of legal definitions of sexual assault, the role of alcohol and drugs is viewed as a form of incapacitation rendering the victim unable to provide consent. Most states mention involuntary intoxication as leading to incapacitation, while a minority of states include the possibility of voluntary use of alcohol as leading to incapacitation.⁷ The critical element for prosecution of a sexual assault case is showing someone is intoxicated enough to render the person incapable of consenting. This becomes ambiguous in situations where the person voluntarily consumed alcohol, became intoxicated but was not unconscious.

Public health surveys have treated alcohol in different ways when defining unwanted sexual experiences and have produced different estimates of alcohol/drug related incidents. The National Violence Against Women Survey (NVAWS),⁸ for example, does not refer to alcohol when asking about possible incidents. It is collected as a characteristic of the incident after the event is classified as an unwanted sexual experience. They found approximately 20% of the lifetime reports of rape involved the victim using alcohol or drugs. Another national study asked about alcohol/drug facilitated sexual experiences directly as part of the initial screening items and found that approximately 44% of incidents were in this category.⁹ The largest estimates of alcohol and drug related incidents are from the National Intimate Partner and Sexual Violence Surveillance System (NISVS) which found 66% of lifetime incidents involved drugs or alcohol.¹⁰

The vignettes used on the main study will vary the perceived role alcohol has in defining the event as something to report on the survey. In order to understand how the survey questions are interpreted we have crafted vignettes including three dimensions. The first dimension is the type of relationship between the actors in the scenario. Research has shown that people tend to judge sexual situations differently when there is a prior sexual relationship between two people compared to when there is not.¹¹ This is also related to legal issues when judging capacity to provide consent.¹² For example, respondents might find it more acceptable for someone who has been in a relationship longer to continue buying a date drinks even if their partner has already admitted being drunk. The second dimension is whether or not one of the actors is intentionally trying to get the other drunk. Intent to incapacitate is an important legal factor in determining if an assault has occurred.¹³ The last dimension involves the form of a (non)consent. Consent can be either verbal and/or non-verbal, but can be ambiguous.

As shown in Appendix H, there are 3 conditions for each of the relationship, intent and type of consent dimensions. This results in 27 variations (3x3x3 = 27) of the scenario. We plan on presenting 3 to 4 of these scenarios to each of the cognitive interview respondents. After each scenario is read, two survey questions will be asked from the screener: 1) the initial question on forced sexual assault and 2) the question on alcohol/drug related incidents. In addition, the questions related to alcohol/drug related incidents from the detailed incident form will be administered. The cognitive interviewer will probe to get feedback on the difficulty of the task, the wording of the scenario and clarity of the instructions.

The other sets of vignettes relate to situations in which psychological coercion is used to engage in sex with another person. The literature suggests that there are three different forms of psychological coercion.¹⁴ Positive coercion involves the perpetrator making positive statements or promises to the victim. For example the perpetrator might say something like “You are so beautiful” or promise to “take the relationship to the next step” if the woman engages in sex. Negative coercion involves negative statements made to the woman such as threatening to break-up. Neutral coercion generally involves constant nagging for sex. We will present respondents with each of these scenarios. For the reasons discussed above, we will vary these questions by the same relationship conditions. This results in 9 conditions (3 coercion conditions x 3 relationship conditions = 9). One or two of these will be administered to respondents during the cognitive interviews.

Re-interview

As noted above, the pilot survey will include re-interviewing a sample of the original respondents. To develop this part of the pilot survey, the cognitive interviews will re-interview the respondents two weeks after the first interview. The followup survey will use the same mode and instrument as the initial interview. The protocol for the cognitive interviews is shown in Appendix I.

Protection of Human Subjects

There is some risk of emotional distress for the respondents given the sensitive nature of the topic, particularly since the questions are of a personal nature. While these questions are very explicit, they have been used on a number of other general population surveys that have collected self-reports of rape and sexual assault. One source was the National Intimate Partner and Sexual Victimization Surveillance Survey,¹⁵ a random digit dial (RDD) telephone survey that was completed in 2011 by the Centers for Disease Control (CDC). The other source is the National Violence against Women Survey.¹⁶ There a number of other studies that have used variations of these questions in general population surveys.¹⁷ In related research, Campbell and Adams demonstrate that rape survivors are often willing research study participants and do not find the experience to induce distress.¹⁸.

Interviews with women from the DC Rape Crisis Center will be conducted by senior Westat interviewers who have experience working with sexual assault victims and who have been trained by an expert on interviewing victims of assault. The training will be provided by a licensed social worker who manages the training department of the Rape Abuse Incest National Network (RAINN). The social worker is an expert at recognizing emotional distress that rape and sexual assault victims might experience.

The Interviews will be stopped if requested by the respondent or the respondent shows signs of emotional distress. If the subject becomes distressed, interview personnel will have contact information for rape and sexual assault crisis lines available for the respondent if needed (Appendix F). In addition, there will be a therapist on-site if an in-person follow-up is needed.

For the women recruited from the general population, Westat cognitive interviewers will be trained by the senior interviewers who have been trained by the RAINN social worker. Interviewers will be trained to recognize signs where the respondent is becoming upset. If they do recognize someone becoming upset the interview will be terminated and the individual will be provided telephone numbers where they can access services (Appendix F).

The protocol has been reviewed and approved by Westat’s IRB (Appendix J).

Assurance of Confidentiality

BJS’ pledge of confidentiality is based on its governing statutes Title 42 USC, Section 3735 and 3789g, which establish the allowable use of data collected by BJS.  Under these sections, data collected by BJS shall be used only for statistical or research purposes and shall be gathered in a manner that precludes their use for law enforcement or any purpose relating to a particular individual other than statistical or research purposes (Section 3735). BJS staff, other federal employees, and Westat staff (the data collection agent) shall not use or reveal any research or statistical information identifiable to any specific private person for any purpose other than the research and statistical purposes for which it was obtained.  Pursuant to 42 U.S.C. Sec. 3789g, BJS will not publish any data identifiable specific to a private person (including respondents and decedents).

Burden Hours for Cognitive Testing

The burden for this task consists of screening participants into the study and the administration of the two cognitive interviews (per person). The table below provides the burden on the public. We estimate the screening in of individuals calling in to volunteer will take approximately 1.5 minutes per respondent. We are not planning on screening out participants unless they do not fall within the specified age range or decide they cannot participate in the study. Each of the cognitive interviews with women recruited from the general population will take approximately 1.5 hours and interviews with victims recruited from the DC Rape Crisis Center will take 1 hour. We will limit the second interview to women recruited from the general population. The total burden comes to 223.25 hours.

Table 3. Estimated Respondent Burden for Cognitive interviews

	Participants	Time per contact	Total time
Screening	130	1.5 minutes	3.25 hours
First interview (General population)	70	1.5 hours	105 hours
First interview (DC Rape Crisis Center pop.)	10	1.0 hours	10 hours
Second interview	70	1.5 hours	105 hours
Total			223.25 hours

1 Fisher, B., and Cullen, F. (2000). Measuring the sexual victimization of women: Evolution, current controversies and future research. In National Institute of Justice (ed.), Measurement and Analysis of Crime and Justice, Vol. 4, National Institute of Justice, Washington DC

2The NCVS currently adjusts the initial interview conducted in the housing unit. Procedures do not adjust interviews that are unbounded at later time in sample due to someone moving into the unit or because of a prior non-interview.

3 Fisher, B., and Cullen, F. (2000). Ibid

4 Prior studies on rape and sexual assault have used both approaches.

5Fisher, B., and Cullen, F. (2000). Ibid

6 Martin, E. 2002. Vignettes and Respondent Debriefing for Questionnaire Design and Evaluation. In S. Presser, J. Rothgeb, M. Couper, J. Lesser, E. Martin, E. Signer (eds.), Methods for Testing and Evaluating Survey Questionnaires. New York: Wiley.

7 AEQUITAS, 2012. RAPE AND SEXUAL ASSAULT ANALYSES AND LAWS, AEquitas: The Prosecutors’ Resource on Violence Against Women, a project of the Pennsylvania Coalition Against Rape ,Washington, D.C., last accessed on October 28, 2012 at: http://www.aequitasresource.org/Rape_and_Sexual_Assault_Analyses_and_Laws_10.5.12.pdf, last accessed October 28, 2012

8 Tjaden and Thoennes, 2006. Extent, Nature, and Consequences of Rape Victimization: Findings From the National Violence Against Women Survey. National Institute of Justice, Office of Justice Programs, NCJ 210346.

9 Kilpatrick, D.G., Resnick, H.S., Ruggiero, K.J., Conoscenti, L.M. and J. McCauley (2007) Drug-facilitated, Incapacitated, and Forcible Rape: A National Study. Final Report for award 2005-WG-BX-0006, Document 219181.

10 Black, M.C., Basile, K.C., Breiding, M.J., Smith, S.G., Walters, M.L., Merrick, M.T., Chen, J., & Stevens, M.R. (2011). The National Intimate Partner and Sexual Violence Survey (NISVS): 2010 Summary Report. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention.

11 Livingston, J.A., Buddle, A.M., Testa, M., and VanZile-Tamsen, C. 2004. The Role of Sexual Precedence in Verbal Sexual Coercion. Psychology of Women Quarterly, 28, 287-297.; Humphreys, T. (2007) Perceptions of Sexual Consent: The Impact of Relationship History and Gender. Journal of Sex Research, Vol. 44, No. 4, 307-315.

12 Scalzo, T. (2007). Prosecuting Alcohol Facilitated Sexual Assault. National Districts Attorneys Association, American Prosecutors Research Institute, Alexandria, VA. http://www.ndaa.org/pdf/pub_prosecuting_alcohol_facilitated_sexual_assault.pdf; Tracey, C., Fromson, T., Long, J. and C. Whitman (2012) Rape and Sexual Assault in the Legal System. Presentation at the Workshop and Panel Meeting on Measuring Rape and Sexual Assault in BJS Surveys, Washington DC, June 5-7, 2012.

13 Tracey, et al, (2012), Ibid

14 Koss, M.P., Abbey, A., Campbell, R. Cook, S., Norris, J., Testa, M., Ullman, S., White, C., and White, J. 2007. Revising the SES: A Collaborative Process to Improve Assessment of Sexual Aggression and Victimization. Psychology of Women Quarterly, 2007, 31, 357-370.

15 Black, et. al, 2011, Ibid.

16 Tjaeden, P., and Thoennes, N. (1998). Ibid

17 Fisher, B., and Cullen, F. (2000). Ibid

⁵ Campbell, R., & Adams, A.E. (2009). Why do rape survivors volunteer for face-to-face interviews? A meta-study of victims’ reasons for and concerns about research participation. Journal of Interpersonal Violence, 24(3):395-405.

18 Campbell, R., & Adams, A.E. (2009). Why do rape survivors volunteer for face-to-face interviews? A meta-study of victims’ reasons for and concerns about research participation. Journal of Interpersonal Violence, 24(3):395-405.

7