Abstract
Attempts to investigate non-disclosure are hampered by the very aspect being examined, namely an unwillingness to disclose non-disclosure. Although qualitative interviews may be considered to be an appropriate method for in-depth exploration of personal experiences, a lack of anonymity and the desire to conform to what is perceived to be socially acceptable limit its application in sensitive research. The current study, using a qualitative approach, addresses non-disclosure in the context of non-suicidal self-injury. Twenty-five young adults from diverse cultural backgrounds were interviewed in depth about their perceptions of self-injury, without the researchers asking directly whether the participants had ever self-harmed. Two techniques were used to enhance discussion within the qualitative interview: participants were invited to (a) discuss three hypothetical scenarios and (b) explore alternative interpretations of statistical data on patterns of self-harm. Key themes emerged regarding disclosure, gender issues, and culturally shaped concerns about the consequences of disclosure. The contributions of each element of the interview to understanding participants' perceptions are highlighted and alternative methodological approaches for examining disclosure are discussed.
There are many situations where researchers wish to explore sensitive issues in depth but are aware that potential participants may be unwilling to disclose their own personal behaviours or attitudes. Unwillingness to disclose is not only a problem when obtaining estimates of overall prevalence of behaviour but also when trying to examine the nature of that behaviour and associated risk factors. Responses obtained from people willing to disclose may be unrepresentative and people who choose not to disclose certain behaviours are unlikely to wish to discuss openly the reasons for their non-disclosure.
Previous research on disclosure has been framed within a counseling or interpersonal behaviour perspective (e.g., Jourard & Richman, 1963). More recently, the issue of disclosure has been examined from a quantitative, survey-based perspective, focussing on the growth of online data collection and identifying concerns about disclosing private information on the Internet (Joinson, Paine, Buchanan, & Reips, 2008; Paine, Reips, Stieger, Joinson, & Buchanan, 2007). Recent studies have suggested that there are few differences in willingness to complete general surveys across mail, telephone, and Internet modes (Hines, Douglas, & Mahmood, 2010; Reddy et al., 2006). When sensitive issues are discussed, however, surveys administered online elicit higher levels of self-disclosure, including willingness to answer sensitive questions (Tourangeau, 2004), and reductions in socially desirable responses (Frick, Bachfiger, & Reips, 2001).
When more in-depth exploration of personal issues is required, surveys are likely to be less appropriate than methods that provide opportunities for in-depth, open discussion. Yoshioka and Schustack (2001) used in-depth interviews with male participants from minority ethnic groups in the United States who had been diagnosed as HIV positive. Their focus was on the barriers to disclosing their health status to their families, including the desire to protect their family from shame and obligation and their unwillingness to disclose gay relationships. While this study shows that qualitative techniques can be effective in exploring the reasons for non-disclosure and its impact on individuals, these participants had already disclosed their HIV status to the researchers. In contrast, people who have not yet disclosed a personal issue to anyone are likely to be less inclined to disclose it to researchers.
Non-suicidal self-injury (NSSI) is an example of an issue where non-disclosure is important, not only because it contributes to inaccurate prevalence rates but also because under-reporting by particular groups may limit their access to appropriate support services. NSSI refers to behaviours such as cutting, scratching, or biting skin, or hitting or banging one's head against something to inflict pain. These behaviours are used primarily as a form of emotion management, rather than with suicidal intent, and their functions include alleviation of negative affect (Klonsky, 2007), communication of distress, control, distraction, release of anger, and expression of emotional pain through physical pain (Warm, Murray, & Fox, 2003). NSSI is associated with emotional and avoidant coping styles (Hawton & Rodham, 2006) and rumination (Borrill, Fox, Flynn, & Roger, 2009). It tends to be secretive, concealed behaviour, so that establishing prevalence and patterns is problematic.
Although frequently viewed as “female” behaviour (Laye-Gindhu & Schonert-Reichl, 2005), some recent research suggests that male and female rates may be similar (Klonsky, Oltmanns, & Turkheimer, 2003), particularly among adolescents (Muehlenkamp & Guitierrez, 2004). Men, however, appear to be more likely to use methods which are not readily recognised by themselves as self-harm, such as deliberately engaging in risky behaviour, using violence toward themselves, punching walls, and so forth (Taylor, 2003), so sex differences in reported NSSI may be misleading. Patterns of disclosure are also relevant to understanding the relationship between culture and rates of reported self-injury. Studies of NSSI among adolescents and students have recorded that a higher proportion of white females reported self-harm than females from other ethnic groups (Gratz, 2006; Hawton & Rodham, 2006), and that British Asian males reported significantly lower levels of self-harm than British Asian females and other males in a UK student sample (Borrill, Fox, & Roger, 2011). Previous research has also noted cultural differences in readiness to report suicidal thoughts (Eskin, 1999; Morrison & Downey, 2000) and reluctance to discuss self-harm if communities view it as stigmatizing or a sign of mental illness (Marshall & Yazdani, 1999). The extent to which apparent differences in reported prevalence are due to “real” differences in behaviour or to variations in willingness to disclose, therefore requires further investigation using a methodology that will enable open discussion and reduce socially or culturally desirable responses.
In 2004, the UK's National Institute for Clinical Excellence (NICE) highlighted the need for more research on self-harm using qualitative methods. In particular, the NICE (2004) guidelines for management of self-harm state that “Rigorous qualitative research should be conducted about the meaning of self-harm to people from different ethnic and cultural groups” (p. 34). A number of studies have been carried out since then, but almost all of these have recruited participants who have been identified (usually through attendance at hospital services) as having taken an overdose (Redley, 2003, 2009; Sinclair & Green, 2005). Although interviews with people personally engaging in cutting and other forms of non-suicidal self-injury can provide important insights into the motives and feelings of those who contact websites or other self-harm support services (Adler & Adler, 2011; Sutton, 2007), evidence is provided only by those who are willing to self-disclose. Greydanus and Shek's (2009) review of the literature on self-harm in adolescents reiterated the need for qualitative research and also commented on the need to examine self-harm within different communities.
The challenge addressed by the current study was to find a method of qualitative inquiry that would be appropriate for exploring attitudes to disclosure of a sensitive and hidden behaviour among participants who might or might not have engaged in such behaviour themselves. In the context of self-injury the aim was to provide a way for potential participants who had never injured themselves to be able to discuss feelings about disclosure by considering the behaviour of people with whom they might identify. At the same time, any participants who had personal experience of self-injury could express their feelings about disclosure without having to disclose their own behaviour. The methodology adopted combined a semi-structured interview with the use of hypothetical scenarios, matched to the participants' own gender and cultural background, and then invited participants to interpret and discuss prevalence data on self-injury by their peers and people similar to themselves. Scenarios or vignettes have been used in previous research on sensitive topics, particularly in the health related disciplines and often as part of focus group discussions (e.g., Brondani, MacEntee, Bryant, & O'Neill, 2008; Chan, Lam, & Shae, 2011). However, the use of scenarios to address the issue of disclosure by people who are not actually known to have engaged in a particular behaviour is believed to be a novel approach, and there appears to be little or no use of this technique within the area of self-injury. The underlying rationale for the study was, therefore, to explore the extent to which participants' responses to open-ended questions, hypothetical scenarios, and data interpretation could be helpful in developing an understanding of attitudes to disclosure of self-injury amongst participants who were not patients and who were not asked whether they had ever harmed themselves.
This combination of interview, scenarios, and data interpretation addressed issues of both method and content. The key methodological question was “How useful are hypothetical and indirect interview techniques in eliciting information about the sensitive issue of self-injury?” This question was explored within the context of the following enquiry: What do young people from different cultural backgrounds perceive as potential barriers to disclosure with regard to self-injury and how do they describe the consequences of disclosure within their family and community?
Method
Design
The aim of this study was to explore expectations regarding disclosure of self-injurious behaviour using a qualitative approach combined with techniques for enhancing engagement and discussion. In order to overcome the tautological problem of non-disclosers not wanting to talk about disclosure, the study did not require any personal disclosure by participants and did not identify whether participants had direct personal experience of self-injury, but it focussed instead on exploring an opportunity sample of participants' views about how “people like you” would behave in regard to self-injury and its disclosure. In addition to open-ended questions, three hypothetical scenarios were used to facilitate the interview process. Participants were then invited to comment on recent statistical data on the prevalence of self-injury.
Participants
Twenty-five students (13 men and 12 women, aged 18 and over) were recruited as participants by advertising the study as one of a number of opportunities for first year psychology students to gain required credits for research participation. The advertisement for the study informed potential participants that the study included questions about self-harming behaviours, but it was made clear that they would not be asked any questions about whether they had personally hurt or injured themselves. The 13 male participants included most of the males in the year cohort. Due to the much larger number of female students in the course, female participants were selected from the list of volunteers on the basis of their availability for the interview. The University ethics committee approved the study, and all participants gave written consent (see Ethics section below).
The 13 male participants described their cultural/ethnic identity as follows: White British or Irish (3); White European (2 Italian, 1 Bulgarian); British Asian (5) (i.e., parents originating from Pakistan, Afghanistan, Bangladesh, India/Kenya, and Iran/Pakistan); Black British (1); and Korean (1). The 12 female participants described themselves as: White European (4) (Polish & Italian); British Asian (4) (India & Bangladesh); White British (1); Brazilian/Spanish (1); South African (1); and English/Ethiopian (1). Participants also came from a range of religious backgrounds, including Christian (mainly Catholic), Muslim, Hindu, and Sikh, although the majority were non-practising.
Interview Techniques
The first part of the interview consisted of open-ended questions about knowledge and views concerning a range of self-injurious behaviours. A topic guide was drawn up to ensure that the three female interviewers covered the same range of issues and ideas. This included an exploration of the participants' personal definitions of self-harm and self-injury, their attitudes and attributions for self-harming behaviour, and any observations of other people self-harming. Participants' understanding of self-harm and self-injury was elicited by asking them to consider how and why people “like them” might engage in self-injury. Participants were also asked how they and their friends typically managed stress or distress and what coping strategies they thought might be useful as alternatives to self-harm.
Following these open-ended questions, the specific issue of disclosure was explored further within the interview, using two techniques to enhance discussion. First, participants were presented with three short written scenarios describing different incidents that might be considered as self-harm. The three scenarios represented methods of self-injury: cutting, head banging, and scratching wounds. Cutting was chosen because it is one of the most common methods reported in surveys and is relatively clearly defined by the use of an implement, blade, or sharp object. Any under-reporting is therefore less likely to be because of lack of identification as self-harm. Scratching was included as a contrast, because of its more ambiguous nature, in order to explore what criteria participants used in defining these behaviours as self-harm. Head-banging (i.e., hitting one's head against a wall, door, or other hard surface) was chosen because the limited research on gender differences suggests that males tend to self-harm in more violent ways than females, and in ways that are less likely to be assessed in research (Taylor, 2003).
The scenarios were constructed by the first author on the basis of reading personal accounts of self-harm and interviewing people with this experience, but they were kept as short as possible to avoid leading the participants towards a particular interpretation. The assigned names of the people in the scenarios were varied to provide a reasonable fit with the gender and cultural background of each interviewee (i.e., using common names found in White British, European, British Asian, and Black British student samples). Participants were provided with the version of the three scenarios which most closely matched by name their own gender and ethnic background, in order to encourage them to think about how someone like themselves might feel and behave in this situation.
Scenario 1: Cutting – British Asian female example
Ayesha is 20 years old. Last night she went into her bedroom and sat alone listening to music for about an hour. She heard her parents and sister going to bed but she couldn't sleep because of the thoughts rushing around in her head and the feelings she was experiencing. She went quietly into the bathroom and locked the door. She took a razor blade from her washbag and made a small cut in her lower arm. She watched the blood for a while and then washed the wound and put a plaster over it. She went back to bed and slept. She has done this twice before.
Scenario 2: Head-banging – White British male example
Matt is a 19 year old student, living in a hall of residence near the main campus. On Tuesday he came back to his room after lectures but instead of making himself something to eat he sat down by his desk and banged his head hard against the wall, causing pain and a slight bruise. He has done this several times before, starting when he was 17.
Scenario 3: Scratching – European male example
Stefan is 21 and is spending a year on work experience in a large company. Last week his line manager commented on the fact that while working Stefan has been scratching repeatedly at wounds on his skin so that they fail to heal. Stefan is now careful to cover up the wounds with his clothes while he is at work, but he continues to scratch and bite his skin when he is at home.
The order of presentation of the three scenarios was varied across the interviews to eliminate bias due to prior information. For each scenario, participants were asked how they felt about the incident and the reasons they thought the person was self-harming. This was followed by a discussion about whether or not the person in the scenario would perceive the incident as self-harm and whether they would be likely to report such actions to a researcher or other person.
The second technique used to explore views on disclosure involved showing participants a table of prevalence data from a previous survey of self-harm carried out with students from two universities (Borrill et al., 2009). These figures reported rates of self-harm by students and included a breakdown of self-reported self-injury across different methods, tabulated by gender and cultural background. Participants were told that this data was authentic survey data, unlike the scenarios that were hypothetical, and that half of the data came from students at their own university. They were invited to examine these figures and offer their personal interpretations of the patterns of reported self-harm that had been obtained. The aim of this exercise was to provide an opportunity for participants to generate their own explanations of the apparent variations in self-harm across different groups. The first purpose was to gain insight into the extent to which participants would attribute patterns to differences in disclosure, and the second was to explore any attributions they made spontaneously concerning disclosure and/or beliefs about culturally shaped behaviour.
Ethics
All first year students were required to take part in a range of studies in order to gain research experience. Participation in a number of studies provided them with the credit points required for them to complete the first year research methods module. This study was advertised as one of the range of approved studies that students could choose to take part in. The advertisement informed potential participants that the study included questions about self-harming behaviours, but it was made clear that they would not be asked any questions about whether they had personally hurt or injured themselves. Those who volunteered to participate in this study received one hour of research participation credit. All participants gave written consent for the interview and audiotape recording and were reassured that the data would be treated as confidential. On completion of the interview, participants were provided with written confirmation from the lead researcher that they had taken part, which enabled them to receive the course credits. None of the three interviewers, including the lead researcher, had direct personal or tutorial contact with the participants they interviewed. Following the interview, participants were provided with details of the University counseling service and other suggestions about receiving support if the discussion had caused them any distress.
Data Analysis
The interviews were recorded using a standard audio recorder and were transcribed verbatim by the interviewers. The lead researcher carried out an analysis of the data, beginning with preliminary coding of the participants' responses to open-ended questions. The preliminary coding provided basic data description, as advocated by users of grounded theory (see Pidgeon & Henwood, 1997), and enabled the researcher to identify content and ideas relevant to disclosure. Rather than progressing directly to developing categories from this data, the next stage of analysis was an exploration of data elicited by the other elements of the interview, namely participants' subjective interpretations of the three hypothetical scenarios and the prevalence figures. These interpretations were examined in relation to the participants' own gender and cultural background to understand the context of their ideas and attributions. Comparing and contrasting the material elicited from the different elements of the interview enabled the main themes to be identified and described.
Findings
Participants were articulate and responded with interest to all aspects of the interview, engaging positively with the techniques used to elicit information. Although the scenarios were brief, a number of spontaneous comments were made which acknowledged the researchers' attempts to make these scenarios relevant to people like themselves, for example commenting positively on the use of culturally relevant names. Both the scenarios and the open discussion of the statistical data complemented comments made in response to the open-ended questions but also generated much more specific information, particularly with regard to gender and cultural issues.
Three major themes were identified: (a) definition and disclosure, (b) disclosure and gendered attributions of intent, and (c) culture and the consequences of disclosure. Variations in attitudes related to the different methods of self-harm were interwoven across these themes. Examples of content elicited by the different techniques are provided below for each theme.
Definition and Disclosure
This theme referred to the features of harmful behaviour that participants felt would lead to a clear definition of self-injury. It emerged partly from the early stages of the interview but predominantly from the scenarios. During the interview, participants expressed a range of ideas about what constituted self-harm, with the majority focussing on cutting. When presented with the scenario of cutting, most participants grounded this definition in the idea of intention, with cutting (using a blade) seen as deliberate and differentiated from an unconscious or impulsive action. Washing the cut and putting a plaster over it was interpreted as meaning that the behaviour was planned, habitual, or repetitive, even though these self-caring actions were difficult for some participants to understand: “I think it's a bit silly because … then he puts a plaster over it, it's like—what did you get out of that?” (Male Participant 3: British Asian)
Despite defining the cutting incident as self-harm, there was agreement across participants that the person in the scenario would be unlikely to disclose their cutting to anyone else, because participants defined this behaviour as a personal issue: “It's like her personal thing, so I don't think she would report that she's doing it” (Female Participant 7: White British). Another participant commented, “He probably wouldn't say anything. It's all to do with himself innit—it's self-ness” (Male Participant 10: British, Iranian/Pakistani/British parents).
Participants were less clear about defining head banging as self-injury, and this scenario elicited surprise and puzzlement among several participants. For example, one participant proffered several explanations in her attempt to make sense of this behaviour: “Low self-esteem maybe, maybe she has a complex. Maybe she feels really bad. Maybe she's not good enough for her studying” (Female Participant 1: Polish). Others perceived the behaviour as abnormal and linked to a desire to increase pain, or to some kind of mental disorder: “I think there is something wrong with him … because something like that is not normal” (Male Participant 9: British, Nigerian parents).
A number of participants interpreted head banging as stress relief and/or expression of anger; for example, “through banging his head against the wall he is just sort of letting out his frustration, letting out his rage, I don't know” (Male Participant 6: White British) and “I wouldn't say that this is necessarily self-harm. I just think she doesn't like to control her anger all the time” (Female Participant 3: British Asian, Indian parents). Other participants, predominantly male, commented that some head banging was “an everyday thing that he does” (Male Participant 3: British, Pakistani parents), and not sufficiently serious to be reported. One participant suggested it would not be disclosed as self-harm even in a research study:
I'm quite sure they would say “No,” because, if I would ever do something like that and then I would be interviewed I would say “No” because in my mind that wasn't like self-injury because it is not something really strong. (Male participant 12: Italian)
The process of defining self-harm, as a pre-requisite to disclosing it, was therefore revealed partly through open-ended questions, but particularly through the use of scenarios.
Disclosure and Gendered Attributions of Intent
This theme referred to gendered attributions and interpretations of self-harm, particularly with regard to intent. Comments about gender emerged in the initial part of the interview where participants explored their expectations about what kinds of people might self-harm. Several female participants identified girls as being more at risk; for example, “… mostly girls comes to mind … I think they go through a lot of stuff in their life, have a lot of pressure, yeah” (Female Participant 12: British, Bangladeshi parents) and “I'm assuming girls and age 13 to 17 … because of puberty and going through changes at school … I think they're more vulnerable than boys” (Female Participant 5: British, Polish parents).
Given the common stereotype of self-harm as female behaviour, it was interesting to note that most participants did not identify gender as an important risk factor at this stage of the study; instead they focussed more on upbringing, context, and personal circumstances: “Maybe someone who doesn't have self-esteem or someone who wants to be better than another one or someone who is alone without parents or without someone who can support you” (Female Participant 2: Italian).
Exploration of the fictional scenarios suggested that male and female participants might vary in their interpretations of both behaviour and disclosure. For example, male participants tended to define cutting in terms of external signals (e.g., blood): “… to actually slice himself and watch the blood drip from his arm …” (Male Participant 6: British, Irish parents), and to normalise head-banging as a common way of expressing anger or frustration: “He doesn't know how to deal with his aggression so he's hit the wall, it's not intending to hurt himself” (Male Participant 7: British, Indian parents). Similarly, the behaviour described in the wound scratching scenario was viewed by almost all male participants as a bad or annoying habit, not to be defined as self-harm, and unlikely to be reported. Female participants rarely mentioned blood or anger but appeared to be trying to interpret possible emotions and motivations of people who self-harmed, both in the interview and when considering the hypothetical scenarios:
If you started to hate yourself a lot or you were quite depressed as well and you were really upset and you thought that everything was your fault, I guess you would hurt yourself just to make yourself feel better. (Female participant 7: White British)
Another attempt at interpretation was as follows: “… she feels frustrated, she feels different, she wants to change, she wants to do something extreme, she wanted to avoid a situation in life—just find an escape maybe …” (Female Participant 2: Italian). Attributions of emotion also generated beliefs about reasons for disclosure or non-disclosure, including shame, denial, and lack of recognition: “If she doesn't feel comfortable with the researcher you know she might feel ashamed, but if not I think she would open up” (Female Participant 4: Polish). The presentation of the statistical data on self-harm incidents also provoked comments on gender, but these were intrinsically linked to issues of culture, which are discussed within the third emerging theme.
Culture and Consequences of Disclosure
The third theme referred to the interaction of cultural norms with concerns about the perceived consequences of disclosure. Participants commented on cultural issues throughout the interviews, exploring both risk factors for self-harm and disclosure issues within their own cultural context:
I'm not white enough for the white culture but I am not black enough for my African family, and so that can leave you out on a limb, so I guess that might make you more likely to self-harm to try and get some acceptance from somewhere. (Female Participant 6: British, British/Ethiopian parents)
One participant referred to avoiding disclosure in general because of cultural expectations: “… because of Asian cultures you can't be very open and you have to lie about a lot of things … white lies, things that would be normal in other cultures sort of you're not allowed to do” (Female Participant 5: British, Indian parents). References to the consequences of disclosure also emerged in the discussion of scenarios: “… they may think other people will react to it … so I think he probably wouldn't say anything” (Male Participant 7: British, Indian parents).
The three-way relationship between culture, gender, and perceived consequences of disclosure became particularly clear when participants were invited to interpret the statistical data on self-harm prevalence rates across gender and culture. British Asian female participants interpreted the apparently low rates of self-harm by Asian males as reflecting the impact of gender roles within their community; for example, “In Asian cultures a man is a strong dominant figure in family and society and so … they will never self-harm, just to keep the image and they don't want to seek attention in such ways …” (Female Participant 9: British, Bangladeshi parents) and “… they do other things like smoking, taking drugs … they've got other means of dealing with peer pressure” (Female Participant 12: British, Bangladeshi parents). This need to meet role expectations was, therefore, seen as a self-fulfilling prophecy: “… because (Asian males) have grown up around a culture where they've been told … how strong they are and they need to support the women and their family and stuff, they do actually tend to become more emotionally strong (Female Participant 9: British, Bangladeshi parents). Thus, these young women appeared to have formed or accepted the view that having to live up to a “strong” non-emotional role actually made their male counterparts less likely to carry out acts of self-injury.
In contrast, the majority of male British Asian interviewees related the same cultural expectations, not to greater resilience, but to reduced willingness to disclose: “Asians are proud, especially males, they wouldn't ever tell anyone … I think it's 100% the males would not come forward” (Male Participant 3: British, Pakistani/Afghani parents). Unwillingness by men to disclose was not confined to the British Asian group, but they expressed it the most forcefully and their perceptions of the consequences of disclosing self-harm included stigma, shame, and even rejection:
The male is the dominant one and they might feel belittled if they admit it … A lot of things happen but they see it as in the religion so we shouldn't say anything about it … even the females might not disclose, but definitely not the males. (Male Participant 10: British, Iranian/Pakistani/English parents)
Nevertheless, one of the most powerful comments resulted from considering one of the scenarios in which the participant described the influence of his Islamic faith:
In our religion [self-harm] is not allowed, harming the body that was given to you … If I did something like that and told my parents I know that they would do something to me—they might hit me, get angry, or not talk to me. There would be a lot of consequences. (Male Participant 3: British, Pakistani/Afghani parents)
The interview, scenarios, and data presentation also elicited other discussions, including comments about relationships. Data suggesting that white young women reported slightly more self-harm than Asian young women were interpreted by one British, Bangladeshi female participant in terms of larger and stronger female friendship groups in British Asian communities: “We're always hanging around together.” An Italian participant commented on the strength of support from his own family: “When I need them they will be there for sure.” A male Korean participant also emphasized the support he gained from being able to talk openly to his parents about any problems: “To tell them everything – I just tell them.” Although many of these comments suggested a climate of greater openness within family in these communities, they do not refer specifically to disclosing self-harm, and at least one white male implied that there were gender pressures against disclosure in his cultural group as well: “It would definitely be something that would be less talked about with men.” The interviews also stimulated discussion about alternative coping strategies, including religion, socializing, and sport, with a memorable comment from one Black British male that “If there was no football self-harming would be even higher than it is!”
Discussion
From a methodological perspective, this study provides an example of how indirect and hypothetical methods of qualitative enquiry may be helpful in exploring material that is unlikely to be disclosed directly if disclosure is seen as problematic. The techniques of in-depth interviewing, discussion of scenarios, and subjective interpretation of statistical data elicited a wide range of comments, and sometimes provided different perspectives. For example, inviting participants to discuss and provide interpretations of real-life data, rather than asking them whether they had personally self-harmed, was successful in eliciting spontaneous interpretations concerning non-disclosure. It also provided an opportunity to explore participants' perceptions of how people from their own background and community might come to engage in self-injury, and how people like themselves would feel about disclosing this.
Gender issues and stereotypes were not raised as often as expected during the first part of the interview, but culturally assigned gender roles were frequently cited later as explanations of apparently different patterns of prevalence within the statistical data. Non-disclosure was raised frequently in relation to scenarios, but a proportion of participants initially failed to consider nondisclosure as a potential factor in interpreting the data. Thus, the combination and comparison of different techniques enhanced the study, and this approach may be useful in other areas of research where information is regularly requested from those people who are least willing to disclose it.
Although the design of this study succeeded in eliciting key attitudes and perceptions, the use of face-to-face interviews may still be problematic in terms of increasing the risk of participants providing socially desirable responses. All three interviewers were female, which could have had an impact on willingness to disclose, though none of the interviewers were known personally to the participants at the time of interview. Another methodological question to address is whether responses to the scenarios would have been different if the prevalence data had been presented first, or whether the interpretations of the data in cultural terms were in any way shaped or biased by the emphasis on thinking about “people like you.” While using matched names was noticed and commended by some participants, it is not known how much impact this had overall in terms of helping participants to identify more closely with the hypothetical self-harmers. A larger study would be required to randomize this effect.
Another limitation of the study is that the sample size is relatively small and it is tempting to over-interpret the diverse comments made in terms of group differences. For example, males and females appeared to focus on different features and use different attributions of intent, but it would be premature to present this as a definitive example of sex-differences in self-disclosure. A meta-analysis of 205 quantitative studies of sex-differences in disclosure (Dindia & Allen, 1992) concluded that women disclose more than men but that the effect size in quantitative studies is small. The review concluded that “it is time to stop perpetuating the myth that there are large sex differences in men's and women's self-disclosure” (p. 118). However, that meta-analysis included only studies of White North-Americans.
Despite the limitations of the present study, the use of hypothetical scenarios helped to uncover the process through which the definition of the behaviour as self-harm, including perceptions of severity, intent, desire for secrecy, and degree of deviation from “normal” behaviour, may lead to the decision to disclose. It also highlighted that young men in particular viewed potentially brain-damaging behaviour (e.g., repeated head banging) as normal, stress-reducing behaviour. Particularly interesting comments on the data and scenarios were elicited from young British Asians about the consequences of disclosure and the perceived pressures to meet the requirements of their gender role. Several female participants believed that expectations of emotional strength and resilience would actually reduce male self-harming behaviour, whereas male participants only talked about the negative consequences that might occur if they disclosed such behaviour. This suggests that prevalence studies using self-report measures will significantly underestimate self-injurious behaviour in those whose culture or gender strongly prohibits self-injury, and for whom the consequences of disclosure are perceived as negative (disapproval) rather than positive (help and support).
Non-disclosure by males may also contribute to a lack of appropriate interventions. There is some evidence that more men than women in the UK are referred for self-inflicted injuries in Accident and Emergency departments (Hawton & Catalan, 1987), but also that men are more likely to have their deliberately inflicted injuries interpreted as accidental, thus “keeping male self-harm from the public consciousness” (Taylor, 2003). It is likely that men, whose gender and/or cultural roles militate against disclosing self-inflicted harmful behaviours, will not seek traditional forms of help, such as counseling or professional support. Similarly, self-harm websites reporting primarily the experiences of women and girls are unlikely to be attractive or appropriate for men. Young men might be more likely to identify and acknowledge repeated head injury if this was approached using the language of stress management rather than the language of self-harm. Similarly, qualitative explorations in other areas of under-reported behaviour could identify subtle changes in language or different approaches to offering support that could increase willingness to disclose and engage.
In retrospect, it could have been interesting to compare this study with an alternative method, for example, using the scenarios and data discussion within a focus group. While there may be concerns about confidentiality, Kitzinger (1995) argued that focus groups might be particularly effective in eliciting cultural values and also in providing support when discussing sensitive issues. Researchers could also consider combining the methods of hypothetical scenarios and data interpretations with new techniques of online data collection. Computer-aided, self-interviews have been found to increase responses to sensitive personal questions, as well as yield more honest answers (Dayan, Paine, & Johnson, 2009). In-depth qualitative research, however, still requires a degree of human interaction in order to probe responses to open-ended questions and react appropriately and supportively to interviewees. Future studies could usefully introduce discussion of hypothetical scenarios and other relevant material into an anonymous on-line discussion forum or use online blogs to discuss personal attitudes and perceptions. A combination of enhanced qualitative techniques and on-line settings could thus help to resolve the problem of non-disclosing non-disclosure.
In conclusion, this study has explored the contribution of different techniques within a qualitative interview to explore the tautological problem of not disclosing non-disclosure, in the specific context of self-injury. Despite the methodological limitations discussed above, the methods used have generated some unique personal accounts and interpretations that challenge the veracity of quantitative measures of self-injury in a diverse population. Future research on disclosure could assess the impact of embedding these techniques within online and/or group discussions.