Sex Roles: A Journal of Research: Attitudes and beliefs about suicidal behavior when coming out is the precipitant of the suicidal behavior

FindArticles > Sex Roles: A Journal of Research > Nov, 2003 > Article > Print friendly

Attitudes and beliefs about suicidal behavior when coming out is the precipitant of the suicidal behavior

Jennifer Ellen Cato

Young persons who describe themselves as lesbian, gay, or bisexual (LGB) are at least twice as likely than their heterosexual peers to report a history of suicidal behavior (Russell & Joyner, 2001). It has been estimated that one in three LGB youths has engaged in suicidal behavior (D'Augelli, Hershberger, & Pilkington, 2001; Remafedi, Farrow, & Deisher, 1991; Safren & Heimberg, 1999).

Many factors probably contribute to the high risk for nonfatal suicidal behavior among LGB youths. The experiences associated with being a stigmatized sexual minority while young and vulnerable are likely a component of this risk (Cochran, 2001). Coming to terms with one's sexual minority status can be psychologically challenging. At this vulnerable time, LGB youths are often rejected by key persons in their lives, including family, teachers, and friends. For many LGB youths, coming out leads to isolation and even homelessness. These experiences likely tax LGB youths' capacity for coping. Problems with depression and substance abuse may emerge. Given this context, it is not surprising that a frequently reported antecedent of nonfatal suicidal behavior among lesbian and gay youths is the personal and interpersonal turmoil associated with coming to terms with one's sexual identity (see Saulnier, 1998, for a review).

At the same time, studies show that suicidal behavior is not simply a function of difficulties, stress, or victimization. In other words, suicidal behavior is not necessarily more prevalent among those who experience adversities (Canetto & Lester, 1995). For example, in the United States, suicidal behavior is uncommon among African American women despite their social and economical disadvantage (Canetto, 1992). Also, the stress related to being a member of a stigmatized sexual minority may not be directly associated with suicidal behavior. For example, according to one study of LGB youth, victimization increases the risk for mental disorders but not for suicidality (Hershberger & D'Augelli, 1995).

It has been suggested that the likelihood that someone responds to adversity with suicidal behavior depends, among other things, on cultural factors, including the prevailing social meanings and scripts of suicidal behavior (Canetto, 1997b; Canetto & Lester, 1998; Canetto & Sakinofsky, 1998; Rubinstein, 1987). The social meanings of suicidal behavior are the social interpretations of the suicidal act. For example, one culture (e.g., the U.S.A.) may view suicidal behavior, especially suicidal behavior in young persons, as a symptom of a mental disorder. Another culture (e.g., India) may idealize suicide among one group of individuals (e.g., widows of all ages) but may condemn it for all other persons. The cultural scripts represent the specific forms that the suicidal behavior typically takes in different cultures. These scripts include actors, scenarios, and method. For example, in one culture, suicide may typically involve the use of guns, as is the case among European Americans (Canetto & Sakinofsky, 1998), whereas in another culture, guns may rarely be used in suicide despite their accessibility, as is the case among the Inuit of Canada (Kral, 1998). These meanings and scripts are said to influence the social consequences of the suicidal behavior (e.g., whether the suicidal person is hospitalized, incarcerated, or, if dead, worshipped). They also probably affect the choices suicidal people make when a suicidal act is considered (Canetto & Sakinofsky, 1998). As Rubinstein (1987) has argued, based on his research on suicide among adolescents in Micronesia, "individuals draw upon these cultural meanings in choosing their course of action and in giving this course of action some public legitimacy" (p. 145).

In a recent suicide prevention initiative, the Surgeon General called attention to the role of cultural beliefs in the risk for suicidal behavior. He recommended more research in this area because some beliefs (such as "the belief that suicide is a noble solution to a personal dilemma") may encourage suicidal behavior, whereas other beliefs (such as the belief that the decision to kill oneself is cowardly) may prevent suicidal behavior (U.S. Public Health Service, 1999, p. 9). In fact, he specifically urged an increased focus on cultural factors in the suicidal behavior of understudied populations, such as LGB youth.

Past studies have revealed that situational variables (e.g., the precipitant of the suicidal act) affect judgments of suicidal behavior. There is consistent evidence that suicidal behavior (fatal as well as nonfatal) is perceived as more permissible when it is motivated by a physical illness (Dahlen & Canetto, 2002; Deluty, 1988-1989a, 1988-1989b; Droogas, Siiter, & O'Connell, 1982-1983; Ellis & Hirsch, 1995; Hammond & Deluty, 1992; Ingram & Ellis, 1995; Lester, Guerriero, & Wachter, 1991; Lo Presto, Sherman, & DiCarlo, 1994-1995; Range & Martin, 1990; Singh, Williams, & Ryther, 1986). The evidence is mixed on whether reactions to nonfatal suicidal behavior differ depending on the sex of the suicidal person. For example, one study showed that suicidal men received less sympathy than suicidal women for their suicidal behavior, especially from other men (White & Stillion, 1988). In another study, it was older adult suicidal women who received the least sympathy. In this second study, suicidal men, regardless of age, obtained intermediate amounts of sympathy (Stillion, White, Edwards, & McDowell, 1989). These findings make sense in light of the fact that "attempting" suicide and "failing" to kill oneself are considered youthful, weak, feminine behavior (Canetto, 1997a). However, a more recent study showed that nonfatal suicidal behavior was evaluated equally negatively in women and men (Dahlen & Canetto, 2002).

Responses to suicidal behavior also vary across types of respondents. In general, women are less accepting of suicidal behavior than are men, especially if respondents are explicitly asked to judge the suicidal decision rather than the suicidal person. For example, in one study, men tended to agree with and accept the suicidal decision more than did women, independent of precipitant (Dahlen & Canetto, 2002). In other studies, men were more likely than women to view the decision to kill oneself as an individual right (Eskin, 1995; Marks, 1988-1989; Wellman & Wellman, 1986). However, in a study where the focus was less clearly on the suicidal decision than on the suicidal person (for instance, participants were shown "pictures" of the suicidal person), women were more likely than men to consider the suicidal decision as justified (van Winkle, Calhoun, Cann, & Tedeschi, 1998). Finally, evidence suggests that androgynous persons tend to see the decision to engage in suicidal behavior as foolish, independent of the reasons for the suicidal behavior (Dahlen & Canetto, 2002).

So far, only one study has contributed some insights on attitudes about suicidal behavior by LGB youths. This study involved semistructured interviews with more than 60 LGB high school students in Salt Lake City, Utah, following a controversial decision by the Salt Lake City School Board and the state legislature to ban all noncurricular clubs rather than allow the establishment of a Gay/Straight Alliance club. The interviews dealt broadly with the experience of LGB youths in Salt Lake City. One unexpected finding of this study concerned reports of suicidal behavior. Suicidal behavior was described by these LGB youth as inevitable, almost as a rite of passage. Many respondents referred "with casual familiarity" to their own past suicidal behavior, "as well as to late-night conversations with suicidal friends" (Russell, Bohan, & Lilly, 2000, p. 80). Talk about suicide was described as ordinary and "mundane" (p. 81). One of the interviewees was quoted as saying: "We all wear our stripes on our sleeves and on our wrists, so to speak" (p. 80). The authors of the study concluded that "some youth may actually feel inadequate as queers if they have not attempted suicide" (p. 80).

The Utah study offered a much-needed glimpse into the meaning of suicidal behavior for LGB youths. A limitation of the Utah study is that the data were collected at a time of community turmoil. Apparently the Gay/Straight Alliance club's founding and the School Board's decision "ignited a media storm that reached international proportions" (Russell et al., 2000, p. 70). The unique circumstances and cultural location of the study raise questions about the findings' generalizability. Furthermore, the Utah study did not specifically focus on attitudes and beliefs about suicidal behavior in LGB youths. Finally, the Utah study only included LGB respondents. An assessment of cultural attitudes about suicidal behavior in LGB individuals would require an evaluation of the attitudes of heterosexual respondents as well.

There is clearly insufficient information on cultural beliefs and attitudes about suicidal behavior in LGB youths. The present study is aimed at beginning to fill this gap in knowledge. We focused on young adults' reactions to a suicidal decision when coming out and being rejected by one's parents was the precipitant of the suicidal act. We compared attitudes about suicidal behavior in the coming out situation to attitudes about suicidal behavior following other well-researched precipitants (i.e., a physical illness, a relationship loss, and an academic failure). To enhance comparisons with past studies, we also explored how evaluations of the suicidal decision may vary depending on the sex of the suicidal person, the respondent's sex, and the respondent's gender identity. One expectation was that suicidal behavior precipitated by the stress of coming out would be viewed in relatively understanding terms, as is the case for suicidal behavior following a physical illness. On the basis of past studies, we also expected that women and androgynous individuals would be more critical of the suicidal decision than would men and other gender-identity types (i.e., undifferentiated, conventionally masculine, or conventionally feminine individuals).

METHOD

Participants

Two-hundred and thirty-seven women and 219 men participated in the study. The average age of these participants was 20 years (SD = 2.01). Ninety-seven percent described their sexual orientation as heterosexual, and the other 3% described themselves as lesbian, gay, bisexual, or transgendered. The majority of participants were single and unattached (72%). Most respondents (83%) were of European American descent. The remaining group described their ethnicity as Hispanic (5%), American Indian/Native American (3%), other (3%), Asian American (2%), or African American (1%). Most respondents were either Roman Catholic (23%) or Protestant (20%). The rest of the participants reported "other" religious beliefs (26%) or no current religious beliefs (25%).

Stimulus Materials

Attitudes toward suicidal behavior were assessed via the Suicide Attitude Vignette Experience (SAVE; Stillion, McDowell, & Shamblin, 1984), Form A, as modified by Dahlen and Canetto (2002). The vignettes describe a young person engaging in nonfatal suicidal behavior following adversities likely to be challenging for young adults: an incurable physical illness, an academic failure, and the loss of an intimate relationship. The SAVE has been reported to have good reliability and validity (Stillion & Stillion, 1998-1999). One additional vignette about a LGB situation was created based on the LGB literature (D'Augelli, Hershberger, & Pilkington, 1998; Rotheram-Borus & Fernandez, 1995). This vignette described a young person engaging in nonfatal suicidal behavior after coming out to, and then being rejected by her/his parents. The vignettes varied on two dimensions, the precipitant of the suicidal behavior (four levels) and the sex of the suicidal person (two levels), for a total of eight vignettes.

Measures

Evaluations of the Suicidal Decision

Respondents were asked to evaluate the suicidal decision described in the vignette using 7-point Likert scales developed and validated by Deluty (1988-1989a, 1988-1989b), as adapted by Dahlen and Canetto (2002). These scales focused on the following dimensions: "wise-foolish," "right-wrong," "selfish-unselfish," "weak-strong," "brave-cowardly," and "active-passive." To reduce response set influences, the location of the positive adjectives was varied (e.g., "foolish-wise," "unselfish-selfish"). In addition, we used Deluty's 7-point scale to measure agreement with the decision to engage in suicidal behavior, as well as his 7-point scale to assess the acceptability of the suicidal decision. Finally, we included Stillion et al.'s 5-point scale to measure degree of sympathy for the suicidal decision (Stillion et al., 1989), but expanded it to 7 points and formatted it to fit with Deluty's scales.

Gender Identity

Gender identity was measured with the short form of the Personal Attributes Questionnaire (PAQ; Spence & Helmreich, 1978). The PAQ uses 5-point Likert-type items to evaluate agreement with bipolar personal attributes. It is composed of two 8-item subscales labeled Femininity (F) and Masculinity (M). The F subscale contains items perceived as desirable for women and men but more common among women (e.g., kindness and understanding). The M subscale includes items judged as positive for both women and men but more common in men (e.g., confidence and independence; Cook, 1985; Helmreich, Spence, & Wilhelm, 1981; Spence & Helmreich, 1978). The PAQ classifies respondents into one of four categories: feminine, masculine, androgynous, or undifferentiated. Feminine persons score high on the F subscale and low on the M subscale, and masculine persons score high on the M subscale and low on the F subscale. Androgynous persons score above the median on both the F and M subscales, and undifferentiated persons score below the median on both subscales. The psychometric properties of the PAQ have been particularly well-evaluated with young adult college samples (Beere, 1990). Factor analyses have shown that the conventionally feminine and the conventionally masculine domains are distinct from one another. The PAQ's internal consistency has been found to be between .61 and .76 (Helmreich et al., 1981).

Procedure

Data were collected from undergraduate students who needed to earn research course credit as well as from students associated with the university Gay/Lesbian/Bisexual/Transgender (GLBT) Student Services Office. The former completed the questionnaires in classrooms. The latter picked up the research packets from the GLBT Student Services Office and mailed the completed questionnaires to the investigators. The survey packet contained an informed consent form, one of the eight possible vignettes, the rating scales for the vignette, the PAQ, and a demographic information questionnaire.

RESULTS

A factor analysis was conducted on the nine items that were used to evaluate the decision to engage in suicidal behavior in order to reduce their number into a smaller set of conceptually and empirically sound composite variables. The items were submitted to a principal factors extraction and then a varimax rotation with Kaiser normalization. On the basis of the criterion of factor loadings of .50 or greater to retain a variable in the factor, and discarding those variables that cross-loaded on more than one factor, two factors were obtained: Factor 1, which was named "Power of the Decision" (eigenvalue = 4.10), and Factor 2, called "Soundness of the Decision" (eigenvalue = 1.10). These two factors accounted for 57.76 percent of the variance. The factor loadings of the items used to evaluate the decision to engage in suicidal behavior are presented in Table I. Because of their high internal consistency (Cronbach's [alpha] = .75), the scores on the items in the "Strength of the Decision" factor were averaged to form a composite variable. Similarly, the four items on the "Soundness of the Decision" factor were averaged to form a composite score (Cronbach's [alpha] = .79). The "active-passive" item did not load onto either factor. It was treated as a separate variable because of its theoretically important and unique content.

A four-factor analysis of variance (suicide precipitant X sex of the suicidal person X respondent sex X respondent gender identity) was computed for each of the three dependent variables that assessed the suicidal decision (i.e., perceived power, soundness, and activity of the suicidal decision). An alpha level of .05 was set for all statistical tests.

Evaluations of the suicidal decision were found to vary depending on the precipitant of the suicidal act, respondent sex, and respondent gender identity (see Table II). There were no significant main effects for target sex on any of the outcome variables. The only significant interaction occurred between target sex and suicide precipitant on the "passive-active" measure.

The Precipitant of the Suicidal Behavior

The precipitant of the suicidal behavior influenced perceptions of the power of the suicidal decision, F(3, 451) = 14.72, p < .001, [[eta].sup.2] = .09. Duncan-Range post hoc tests (p < .05) indicated that engaging in suicidal behavior because of a physical illness was considered a more powerful decision (M = 4.80, SD = 1.20) than engaging in the same behavior following either "coming out" (M = 5.52, SD = 1.19), a relationship loss (M = 5.68, SD = 1.19), or an achievement failure (M = 5.71, SD = 1.19).

The context of the suicidal behavior was also related to how sound the suicidal decision was perceived to be, F(3, 451) = 17.39, p < .001, [[eta].sup.2] = .10. Duncan-Range post hoc tests revealed that suicidal behavior in response to a physical illness (M = 5.08, SD = 1.22) was seen as a more sound decision than suicidal behavior in response to achievement failure (M = 5.93, SD = 1.21), coming out (M = 6.04, SD = 1.21), or relationship loss (M = 6.07, SD = 1.21).

Finally, the precipitant of the behavior influenced the degree to which the suicidal decision was perceived as an active decision, F(3,443) = 7.03, p < .001, [[eta].sup.2] = .05. Specifically, suicidal behavior in response to a physical illness (M = 3.72, SD = 2.01) and coming out (M = 3.70, SD = 2.00) were considered as significantly more active decisions than the same behavior following a relationship loss (M = 4.25, SD = 2.00) or an academic failure (M = 4.76, SD = 2.01; Duncan-Range, p < .05). The latter main effect should be interpreted with caution, due to the interaction between suicide precipitant and target sex, F(3, 435) = 3.47, p < .05, [[eta].sup.2] = .03. A man's suicidal decision was rated as more active if it was in response to a physical illness (M = 3.40, SD = 1.81), as compared to coming out (M = 4.18, SD = 2.08), a relationship loss (M = 4.13, SD = 2.14), or an academic failure (M = 4.64, SD = 2.00). A woman's suicidal decision following coming out (M = 3.18, SD = 2.12) was rated as the most active behavior, as compared to the same decision following a physical illness (M = 4.00, SD = 1.75), a relationship loss (M = 4.35, SD = 2.07), or an academic failure (M = 4.86, SD = 1.91; Duncan-Range, p < .05).

Target Sex

No significant main effect of target sex was found on perceived power, F(1, 453) = 0.06, p > .05, perceived soundness F(1, 453) = 1.17, p > .05, or perceived activity of the suicidal decision F(1, 445) = 0.01, p > .05.

Respondent Sex

Men perceived the decision to engage in suicidal behavior as less "unsound" (M = 5.61, SD = 1.27) than did women (M = 5.93, SD = 1.26), F(1, 453) = 7.47, p < .01, [[eta].sup.2] = .02. However, both men (M = 5.44, SD = 1.24) and women (M = 5.41, SD = 1.25) saw the suicidal decision as one of powerlessness, F(1, 453) = 0.10, ns. Finally, men (M = 4.00, SD = 2.04) and women (M = 4.18, SD = 2.05) were similar in their judgment of the passivity/activity of the suicidal decision, F(1, 445) = 0.86, ns.

Respondent Gender Identity

Participants were classified into the four groups (i.e., feminine, masculine, androgynous, and undifferentiated) using the median split method described by Spence and Helmreich (1978). The medians obtained were 24 for the Femininity Scale and 21 for the Masculinity Scale. Based on these medians, 161 participants (36%) were classified as androgynous, 119 (27%) as feminine, 91 (21%) as masculine, and 72 (16%) as undifferentiated. Respondent gender identity affected perceptions of the soundness of the suicidal decision, F(3, 438) = 3.65, p < .05, [[eta].sup.2] = .02. Duncan post hoc tests revealed that androgynous persons (M = 6.03, SD = 1.26) saw the suicidal decision as significantly less sound than did persons with an undifferentiated gender identity (M = 5.59, SD = 1.26), those with a conventionally masculine gender identity (M = 5.63, SD = 1.26), or those with a conventionally feminine gender identity (M = 5.65, SD = 1.26). Gender identity did not significantly influence perceptions of the potency of suicidal decision, F(3, 438) = 2.30, ns. Finally, gender identity did not significantly affect perceptions of the suicidal decision as active or passive, F(3, 431) = 0.72, ns.

DISCUSSION

In the present study we examined young adults' evaluations of the suicidal decision of a LGB peer who came out and was rejected by her/his parents, and then compared them to evaluations of suicidal decisions in response to other precipitants (i.e., a physical illness, an academic failure, and a relationship loss). Consistent with past research, we also considered the influence that the sex of the suicidal person, the respondent's sex, and the respondent's gender identity may have on judgments of the suicidal decision.

We found that the precipitant of the suicidal behavior affected evaluations of a suicidal decision. However, it was physical illness, not coming out, that was singled out as a relatively understandable motivation for suicidal behavior. Specifically, persons who engaged in suicidal behavior as a result of a physical illness were seen as having made a more powerful and sounder decision than those who became suicidal as a result of a relationship loss or an academic failure. This finding is consistent with those of many previous studies, all of which showed that a physical illness is considered the most understandable reason for suicidal behavior, both nonfatal and fatal (Dahlen & Canetto, 2002; Deluty, 1988-1989a, 1988-1989b; Droogas et al., 1982-1983; Ellis & Hirsch, 1995; Hammond & Deluty, 1992; Ingram & Ellis, 1995; Lester et al., 1991; Lo Presto et al., 1994-1995; Range & Martin, 1990; Singh et al., 1986). Suicidal decisions following a physical illness and coming out were rated as relatively active, as compared with suicidal decisions following a relationship loss or an academic failure. This effect was qualified by an interaction between target sex and precipitant. A man's suicidal decision was perceived as most active when it was in response to a physical illness; a woman's suicidal decision was perceived as most active when it was a response to parental rejection following "coming out."

As noted above, the decision to engage in suicidal behavior following coming out was perceived as unsound and weak. One interpretation is that the decision to engage in a suicidal act following coming out may not be as condoned as a suicidal decision following a physical illness. This could mean that the high rates of nonfatal suicidal behavior among LGB youth are not be related to a greater social acceptance of such behavior in LGB youth. It is important to note that a majority of respondents (97%) in this study were heterosexual. Because of the small number of LGB respondents in our sample, we did not separately examine the beliefs of LGB participants and compare them to those of heterosexual participants. Therefore, we do not really know if LGB individuals consider a suicidal decision following coming out as weak and unsound, as did the heterosexual respondents. This is a question for future research.

In this study, evaluations of the decision to engage in suicidal behavior did not vary depending on the sex of the suicidal person. The suicidal decision was rated as weak, unsound, and relatively passive in both women and men. Because "attempting" suicide is viewed as youthful, feminine behavior, our findings suggest a double bind for women, who may be both expected to engage and condemned for engaging in what is considered as "unsuccessful" suicidal behavior (Canetto, 1997a). Our target-sex findings are consistent with those of one previous study of attitudes toward nonfatal suicidal behavior (Dahlen & Canetto, 2002), but they are at variance with the findings of other studies (Stillion et al., 1989; White & Stillion, 1988). The reasons for the discrepancies across studies are likely due to differences in methodology. In the studies that reported an effect of target sex (e.g., White & Stillion, 1988), respondents were shown vignettes that varied by target sex as well as by other characteristics, and they were not specifically asked to focus on the suicidal decision. In the study that did not find an effect for target sex (Dahlen & Canetto, 2002), the respondents were given single-sex vignettes, and they completed measures that focused explicitly on the suicidal decision. Clearly, the use of a mixture of vignettes featuring suicidal women and men highlights target sex as a variable and brings out target-sex effects. Also, a focus on the suicidal person rather than on the suicidal decision seems to elicit more gendered evaluations of the suicidal behavior.

As expected, based on past studies, men perceived the decision to engage in suicidal behavior as less unsound than did women. This is consistent with findings from related research, which has shown that men are more likely than women to agree with and accept suicidal decisions (Dahlen & Canetto, 2002) and to think that suicidal behavior can be justified and rational (Miller, 1994, cited in Stillion & Stillion, 1998-1999). Men also tend to view the decision to suicide as an individual right (Marks, 1988-1989; Wellman & Wellman, 1986). In addition, in the present study, both women and men rated the suicidal decision as low in power. Similarly, an early study of college students showed that attempting suicide is considered weak behavior (Linehan, 1973). In Linehan's study, "attempted" suicide was also viewed as more feminine than death by suicide. Furthermore, in the present study, women and men were similar in their judgment of the suicidal decision's activity-passivity (their scores being in the middle range of the continuum).

Finally, gender identity influenced perceptions of the suicidal decision. Specifically, androgynous persons were most likely to rate the suicidal decision as unsound, as compared to all other gender identity types. Similarly, Dahlen and Canetto (2002) found that androgynous persons were the least accepting of the suicidal decision of all gender identity types. They tended to view the suicidal choice as foolish, independent of reasons for the suicidal behavior. Androgynous persons have also been reported to score higher on protective factors against suicidal behavior, including survival beliefs, coping beliefs, and moral objections to suicidal behavior (Ellis & Range, 1988).

This study's findings on gender identity confirm and extend the results of past studies of attitudes about the suicidal decision as well as those of studies of attitudes about the suicidal person (Cato & Canetto, 2003; Dahlen & Canetto, 1996, 2002; Ellis & Range, 1988; Stillion, McDowell, Smith, & McCoy, 1986). In short, androgynous individuals are more forgiving of the suicidal person, but they are less accepting of the suicidal behavior. They seem less critical of the suicidal individual, and they take the suicidal behavior seriously, but they are less likely to idealize or even condone the suicidal decision. These findings have implications for the prevention of suicidal behavior. It may be that androgynous persons are the best candidates for training as peer counselors in suicide prevention programs. They may be more capable of listening and expressing care for the suicidal person while also being the most firm in their judgment that suicidal behavior is an unsound choice.

Our findings, together with those of other studies on beliefs about and attitudes toward suicidal behaviors, suggest new directions for suicide prevention programs. Educational programs may benefit from including a didactic component regarding the epidemiology of gender and suicidal behavior across sexual orientations such that participants can be informed about the forms of suicidal behavior typically found in women and men, and how risk for these different suicidal behaviors varies depending on sexual orientation. This didactic presentation could be followed by a discussion of social beliefs about gender and suicidal behavior under different conditions, and then an analysis of the limitations that these beliefs impose on coping (Canetto, 1997a). This focus on unexamined and dysfunctional social beliefs about suicidal behavior (such as the belief, commonly found in men, that suicide is a personal choice and an individual right; or the belief that surviving a suicidal act is feminine, weak behavior) builds upon the methods and wisdom of cognitive and feminist therapies for depression.

The accumulating evidence on the association of androgyny with constructive beliefs and attitudes about suicidal behavior argues for an androgyny component in primary and secondary prevention programs (Dahlen & Canetto, 2002; Westefeld, Whitchard, & Range, 1990). A focus on androgyny is also indirectly supported by the findings of studies of persons who died of suicide. Psychological autopsy studies tend to show that people who killed themselves have a limited range of interests, are closed to feelings, and avoid new experiences and challenges (Canetto, 1997b). These tendencies, which are part of the Openness to Experience personality dimension, suggest rigidity in the sense of self and, possibly, insufficient androgyny. The evidence on attitudes toward suicidal behavior and androgyny may also argue for androgyny as a broad mental health goal, along with self-esteem and openness to new experience. This goal is consistent with the idea that prevention should focus on strengthening resilience factors, not only on preventing risk factors (McDaniel, Purcell, & D'Augelli, 2001).

Table I. Factor Loadings of Items Evaluating the Suicidal Decision

                         Factor 1                     Factor 2
                       (Power of the               (Soundness of
         Item           Decision)                  the Decision)

Weak-strong                .75
Cowardly-brave             .72
Elicits sympathy           .70
Selfish-unselfish          .69
Permission to suicide                                   .87
Acceptable decision                                     .79
Wrong-right                                             .67
Foolish-wise                                            .61
Table II. Evaluations of the Suicidal Decision: Mean Scores and F
Statistics

                  Precipitant of the suicidal behavior
                          Coming           Relation
                 Illness   out    Failure   loss      F

Power             4.80     5.52    5.71     5.68    14.72***
Soundness         5.08     6.04    5.93     6.07    17.39***
Active/Passive    3.72     3.70    4.76     4.25     7.03***

                        Target sex        Respondent sex
                    Women   Men     F    Women    Men    F

Power                5.44   5.41   0.06   5.41   5.44   0.10
Soundness            5.84   5.71   1.17   5.93   5.61   7.47**
Active/Passive       4.10   4.08   0.01   4.18   4.00   0.86

                        Respondent gender identity
                     Fem     Masc    Andro  Undiff   F

Power                5.18    5.54    5.53    5.33   2.30
Soundness            5.65    5.63    6.03    5.57   3.65**
Active/Passive       3.90    4.14    4.25    4.03   0.72

Note. Illness = physical illness: Coming out = parental rejection
following coming out: Failure = academic failure; Relation loss = loss
of an intimate relationship; Fem = feminine; Masc = masculine; Andro =
androgynous; Undiff = undifferentiated. Lower numbers reflect
perceptions of the suicidal decision as more powerful, more sound, and
more active.
* p < .05. ** p < .01. *** p < .001.

(1) Colorado State University, Fort Collins, Colorado.

(2) To whom correspondence should be addressed at Department of Psychology, Colorado State University, Fort Collins, Colorado 80523-1876; e-mail: scanetto@lamar.colostate.edu.

REFERENCES

Beere, C. A. (1990). Gender roles: A handbook of tests and measures. New York: Greenwood Press.

Canetto, S. S. (1992). Gender and suicide in the elderly. Suicide and Life-Threatening Behavior, 22, 80-97.

Canetto, S. S. (1997a). Meanings of gender and suicidal behavior during adolescence. Suicide and Life-Threatening Behavior, 27, 339-350.

Canetto, S. S. (1997b). Gender and suicidal behavior: Theories and evidence. In R. W. Maris, M. M. Silverman, & S. S. Canetto (Eds.), Review of suicidology (pp. 138-167). New York: Guilford.

Canetto, S. S., & Lester, D. (1995). Gender and the primary prevention of suicide mortality. Suicide and Life-Threatening Behavior, 25, 58-69.

Canetto, S. S., & Lester, D. (1998). Gender, culture, and suicidal behavior. Transcultural Psychiatry, 35, 163-191.

Canetto, S. S., & Sakinofsky, I. (1998). The gender paradox in suicide. Suicide and Life-Threatening Behavior, 28, 1-23.

Cato, J. E., & Canetto, S. S. (2003). Young adults' reactions to gay and lesbian peers who became suicidal following "coming out" to their parents. Suicide and Life-Threatening Behavior, 33, 201-210.

Cochran, S. D. (2001). Emerging issues in research on lesbians' and gay men's mental health: Does sexual orientation matter? American Psychologist, 56, 931-947.

Cook, E. P. (1985). Psychological androgyny. New York: Pergamon Press.

Dahlen, E. R., & Canetto, S. S. (1996, April). The role of gender and context in attitudes toward nonfatal suicidal behavior. Paper presented at the 29th Annual Conference of the American Association of Suicidology, St. Louis, MO.

Dahlen, E. R., & Canetto, S. S. (2002). The role of gender and suicide precipitant in attitudes toward nonfatal suicidal behavior. Death Studies, 26, 99-116.

D'Augelli, A. R., Hershberger, S. L., & Pilkington, N. W. (1998). Lesbian, gay, and bisexual youth and their families: Disclosure of sexual orientation and its consequences. American Journal of Orthopsychiatry, 68, 361-371.

D'Augelli, A. R., Hershberger, S. L., & Pilkington, N. W. (2001). Suicidality patterns and sexual orientation-related factors among lesbian, gay, and bisexual youths. Suicide and Life-Threatening Behavior, 31, 250-264.

Deluty, R. H. (1988-1989a). Physical illness, psychiatric illness, and the acceptability of suicide. Omega: Journal of Death and Dying, 19, 79-91.

Deluty, R. H. (1988-1989b). Factors affecting the acceptability of suicide. Omega: Journal of Death and Dying, 19, 315-326.

Droogas, A., Siiter, R., & O'Connell, A. N. (1982-1983). Effects of personal and situational factors of attitudes toward suicide. Omega: Journal of Death and Dying, 13, 127-144.

Ellis, J. B., & Hirsch, J. K. (1995). Attitudes of traditional and non-traditional college students towards suicide. College Student Journal, 29, 445-447.

Ellis, J. B., & Range, L. M. (1988). Femininity and reasons for living. Educational and Psychological Research, 8, 19-24.

Eskin, M. (1995). Adolescents' attitudes toward suicide, and a suicidal peer: A comparison between Swedish and Turkish high school students. Scandinavian Journal of Psychology, 36, 201-207.

Hammond, L. K., & Deluty, R. H. (1992). Attitudes of clinical psychologists, psychiatrists, and oncologists toward suicide. Social Behavior and Personality, 20, 289-294.

Helmreich, R. L., Spence, J. T., & Wilhelm, J. A. (1981). A psychometric analysis of the Personal Attitudes Questionnaire. Sex Roles, 7, 1097-1108.

Hershberger, S. L., & D'Augelli, A. R. (1995). The impact of victimization on the mental health and suicidality of lesbian, gay, and bisexual youths. Developmental Psychology, 31, 65-74.

Ingram, E., & Ellis, J. B. (1995). Situational analysis of attitudes toward suicide behavior. Death Studies. 19, 269-275.

Kral, M. (1998). Suicide and the internalization of culture: Three questions. Transcultural Psychiatry, 35, 221-233.

Lester, D., Guerriero, J. M., & Wachter, S. M. (1991). The suicide attitude vignette experience (SAVE): A search for sexual stereotypes in the perception of suicidal behavior. Death Studies, 15, 435-441.

Linehan, M. M. (1973). Suicide and attempted suicide: Study of perceived sex differences. Perceptual and Motor Skills. 37, 31-34.

Lo Presto, C. T., Sherman, M. F., & DiCarlo, M. A. (1994-1995). Factors affecting the unacceptability of suicide and the effects of evaluator depression and religiosity. Omega: Journal of Death and Dying, 30, 205-221.

Marks, A. (1988-1989). Structural parameters of sex, race, age, and education and their influence on attitudes toward suicide. Omega: Journal of Death and Dying. 19, 327-336.

McDaniel, J. S., Purcell, D., & D'Augelli, A. R. (2001). The relationship between sexual orientation and risk for suicide: Research findings and future directions for research and prevention. Suicide and Life-Threatening Behavior, 31(Suppl.), 84-105.

Range, L. M., & Martin, S. K. (1990). How knowledge of extenuating circumstances influences community reactions toward suicide victims and their bereaved families. Omega: Journal of Death and Dying, 21, 191-198.

Remafedi, G., Farrow, J. A., & Deisher, R. W. (1991). Risk factors for attempted suicide in gay and bisexual youth. Pediatrics, 87, 869-875.

Rotheram-Borus, M. J., & Fernandez, M. I. (1995). Sexual orientation and developmental challenges experienced by gay and lesbian youths. Suicide and Life-Threatening Behavior, 25(Suppl.), 26-34.

Rubinstein, D. H. (1987). Cultural patterns and contagion: Epidemic suicide among Micronesian youth. In A. B. Robillard & A. Marsella (Eds.), Contemporary issues in mental health: Research in the Pacific Islands (pp. 127-148). Honolulu: Social Science Research Institute.

Russell, G. M., Bohan, J. S., & Lilly, D. (2000). Queer youth: Old stories, new stories. In S. L. Jones (Ed.), A sea of stories: The shaping power of narrative in gay and lesbian cultures (pp. 69-92). New York: Harrington Park Press.

Russell, S. T., & Joyner, K. (2001). Adolescent sexual orientation and suicide risk: Evidence from a national study. American Journal of Public Health, 91, 1276-1281.

Safren, S. A., & Heimberg, R. G. (1999). Depression, hopelessness, suicidality, and related factors in sexual minority and heterosexual adolescents. Journal of Consulting and Clinical Psychology, 67, 859-866.

Saulnier, C. F. (1998). Prevalence of suicide attempts and suicide ideation among lesbian and gay youth. Journal of Gay and Lesbian Social Services, 8, 51-68.

Singh, B. K., Williams, J. S., & Ryther, B. J. (1986). Public approval of suicide: A situational analysis. Suicide and Life-Threatening Behavior, 16, 409-418.

Spence, J. T., & Helmreich, R. L. (1978). Masculinity and femininity: Their psychological dimensions, correlates, and antecedents. Austin, TX: University of Texas Press.

Stillion, J. M., McDowell, E. E., & Shamblin, J. B. (1984). The suicide attitude vignette experience: A method for measuring adolescent attitudes toward suicide. Death Education, 8, 65-79.

Stillion, J. M., McDowell, E. E., Smith, R. T., & McCoy, P. A. (1986). Relationships between suicide attitudes and indicators of mental health among adolescents. Death Studies, 10, 289-296.

Stillion, J. M., & Stillion, B. D. (1998-1999). Attitudes toward suicide: Past, present and future. Omega: Journal of Death and Dying, 38, 77-97.

Stillion, J. M., White, H., Edwards, P. J., & McDowell, E. E. (1989). Ageism and sexism in suicide attitudes. Death Studies, 13, 247-261.

U.S. Public Health Service. (1999). The Surgeon General's call to action to prevent suicide. Washington, DC: Author.

vanWinkle, D., Calhoun, L. G., Cann, A., & Tedeschi, R. (1998). Social reactions to attempted suicide: The effects of gender and physical attractiveness. Omega: Journal of Death and Dying, 37, 89-100.

Wellman, M. M., & Wellman, R. J. (1986). Sex differences in peer responsiveness to suicide ideation. Suicide and Life-Threatening Behavior, 16, 360-378.

Westefeld, J. S., Whitchard, K. A., & Range, L. M. (1990). College and university student suicide: Trends and implications. Counseling Psychologist, 18, 464-476.

White, H., & Stillion, J. M. (1988). Sex differences in attitudes toward suicide: Do males stigmatize males? Psychology of Women Quarterly, 12, 357-366.

Jennifer Ellen Cato (1) and Silvia Sara Canetto (1), (2)