Suicide Stigma as a Predictor of Help-Seeking Intention among Undergraduate Students in Jakarta

Previous research has found that stigma tends to impair help-seeking intention in a suicidal crisis for university students across the globe. Yet very little research has investigated how suicide-specific stigma affects help-seeking intention in a suicidal crisis among university students in the Indonesian context. This research examines the adverse relationship between suicide stigma and help-seeking intention. A total of 284 university students in Jakarta, aged 18–24 years (M = 20.14, SD = 1.18), participated in the study. Linear regression analysis demonstrated that suicide stigma positively contributes to the intention to seek help from both formal and non-formal sources. These results contradict previous findings, possibly due to the unique culture in Indonesia. Theoretical and practical implications are discussed for Indonesian suicide prevention efforts. Stigma Bunuh Diri sebagai Prediktor Niatan Pencarian Bantuan pada Mahasiswa Program Sarjana di Jakarta


Introduction
Globally, suicide is the second highest cause of death among young people aged 15-29 years (World Health Organization, 2014), and previous research has shown that university students are one of the populations most vulnerable to suicide (Li et al., 2014;Zhai et al., 2015;Lageborn, Ljung, Vaez, & Dahlin, 2017;Assari, 2018). In Indonesia alone, 6.9% of university students aged 18-30 in Yogyakarta reported having suicidal ideation at some point in their lives (Peltzer, Yi, & Pengpid, 2017). Previous findings also indicate that, in urban areas, up to 21% of university students have a lifetime history of self-harm and suicide attempts (Tresno, Ito, & Mearns, 2012).
The low rate of help seeking during a suicidal crisis is related to many factors, such as questioning the need for help or the effectiveness of available treatments, lack of time, shortage of information on help resources, assumption that one has the self-capacity to resolve the situation, lack of perceived problem severity, financial barriers, and the perception of an isolating campus environment (Arria et al., 2011;Downs & Eisenberg, 2012;Curtis, 2010). In addition to these factors, there are also barriers related to stigma that inhibit university students from seeking help during suicidal crises, including a stigma of professional help seeking (Arria et al. 2011;Downs & Einsenberg, 2012) and stigma of mental illness (Curtis, 2010;Neilson et al., 2014).
When it comes to help-seeking intention, there is also a stigma of suicide that needs to be taken into account. For example, research in Australia has shown differences in the relation of suicide stigma to the intention to seek help for suicidal ideation. Within the general population, the stronger the stigma is, the more negative a person's attitude is toward help-seeking intention, and the lesser the odds that one will seek help from mental health professionals (Calear, Batterham, & Christensen, 2014). Meanwhile, among medical students, there is no significant association between suicide stigma and helpseeking intention (Chan, Batterham, Christensen, & Galletly, 2014). This may indicate the presence of significant differences between medical students and the general population, which includes people of various educational backgrounds. This study aims to further investigate this issue in a broader university student population by analyzing whether university students generally are more similar to medical students or to the general population.
In addition to weighing the more diverse educational backgrounds of university students, studying the relation between suicide stigma and help-seeking intention in Indonesia might reveal other significant factors. For instance, because collectivist cultures have different stigmas associated with suicide than individualist cultures do, Indonesia, which tends to the former, may have a unique type of suicide stigma (Han et al., 2017;Goldston et al., 2008). Moreover, Indonesian culture tends to be influenced by religion, which may lead to a different suicide stigma as compared to other countries (Goldston et al., 2008), and the choice to seek help to overcome mental health issues could be more directed toward religious leaders (Muluk, Hudiyana, & Shadiqi, 2018). Help sources in Indonesia are also more diverse due to a national acknowledgment of traditional medicine (Indonesian Ministry of Health, 2013). Apart from these, online services can be reliable sources of help as they are more accessible to young Indonesians (Gayatri et al., 2015), and people with suicidal tendencies tend to prefer using online help sources (Seward & Harris, 2016). Therefore, Indonesians may have different patterns of suicide stigma, help sources, and help-seeking intentions compared to participants in previous studies. Therefore, Indonesians may have different patterns of suicide stigma, help sources, and help-seeking intentions compared to participants in previous studies.
As the current literature contains gaps in the explanation of differences between medical students and the general population and lacks consideration of the cultural characteristics specific to Indonesia, this research aims to investigate the effect of suicide stigma on helpseeking intention among university students in Jakarta. Hypothesis 1: Suicide stigma has a negative association with formal and non-formal help-seeking intention for university students in Jakarta.
Hypothesis 2: Suicide stigma decreases formal and nonformal help-seeking intention for university students in Jakarta.

Measurement Adaptation Process. Stigma of Suicide
Scale. To measure suicide stigma, this study used the stigma subscale from the Stigma of Suicide Scale (SOSS) by , which was translated into Bahasa Indonesia. The SOSS measures attitudes toward individuals who die by suicide. It contains 58 items with five answer choices ranging from 1 "Strongly disagree" to 5 "Strongly agree." The SOSS has three subscales that can be scored separately: Stigma, Isolation/Depression, and Normalization/Glorification. The overall reliable Cronbach's alpha is 0.90, while it is 0.95 for the Stigma subscale, 0.88 for the Isolation/Depression subscale, and 0.86 for the Normalization/Glorification subscale. Validity testing was conducted by correlating the scores with another suicide stigma questionnaire, the Suicide Opinion Questionnaire ) through principal component analysis. As a result, several low quality items of the SOSS were dropped. The SOSS has been adapted to Asian cultures using a sample of university students in the People's Republic of China with an average age of 19.6 years (Han et al., 2017). The results of this adaptation show the Cronbach's alpha reliability of the Stigma subscale is 0.72.
In this study, the translate-back translate process was used with the permission and direct feedback from the developer of the SOSS, Philip Batterham. Only one subscale of the SOSS, the Stigma subscale containing 31 items, was employed because this study focuses only on negative attitudes toward individuals who die by suicide. Psychometric adaptation was conducted using 74 university students. Psychometric testing of the 31 items showed an overall Cronbach's alpha of 0.962. Item 1, which states, "In general, people who die by suicide are punishing others," was found to be not valid (corrected total item correlation: −0.77, Cronbach's alpha if item deleted: 0.965). The authors evaluated that "punishing others" is not a common negative stigma attached to people who die by suicide in Indonesia; therefore, the item was dropped. Psychometric testing was reconducted afterward and resulted in a higher ranging from 1 "Extremely unlikely" to 7 "Extremely likely." According to the developers (Deane & Wilson, 2007), the total score on the GHSQ can be analyzed as one overall scale that consists of multiple domains of help sources or as a scale of a single domain (e.g., mental health professionals). Overall, the GHSQ scale has a Cronbach's alpha of 0.82. The test-retest reliability over three weeks of the GHSQ showed a Cronbach's alpha of 0.92, while the Cronbach's alpha for personal emotional problems is 0.7. The test-retest reliability over three weeks for emotional problems showed a Cronbach's alpha of 0.86. The Cronbach's alpha for help-seeking intention related to suicidal ideation is 0.83. Meanwhile, the GHSQ test-retest score for suicidal idation showed a Cronbach's alpha of 0.80. There was a significant positive correlation between help-seeking intention and help-seeking behavior related to suicidal ideation after three weeks of data collection.
The GHSQ was adapted to a Southeast Asian context in the Philippines (Tuliao & Velasquez, 2014) in a study with psychology students aged 16-21. The results of this adaptation showed a Cronbach's alpha of 0.83, and test-retest reliability of the overall GHSQ scale showed a Cronbach's alpha of 0.88. This adaptation utilized an indigenous method named pagtatanung-tanong, which prompted Tuliao and Velasquez (2014) to add more detailed types of mental health professions on the formal help sources scale, drop the GHSQ scale for suicidal ideation, and reduce the answer choice format from a seven-to five-point scale.
To apply the GHSQ in the Indonesian context, the help sources of the GHSQ were modified by adding items related to mental health community services available to provide emotional supports, including peer counselors, and dropping the item about a national hotline because the hotline is no longer available. In addition, an item on traditional medicine was added based on the classification of the Indonesian Ministry of Health (2013). The authors also considered the participants' ease in filling out the survey and reduced the sevenpoint answer scale to a four-point scale ranging from 1 "Extremely unlikely" to 4 "Extremely likely." Different from the adaptation for the Philippines used by Tuliao and Velasquez (2014), the GHSQ used in this study consists of the suicidal ideation subscale of the GHSQ.
After the modifications described above, the GHSQ suicidal ideation subscale utilized in this study consists of 20 items: 7 items about non-formal help sources; 11 items about formal help sources that combine mental health professions from psychiatry, psychology, and counseling with religious leaders and traditional medicine practitioners; one item on "not seeking help at all," and one item about other help sources that are not mentioned in any of the other options. The responses to only 18 items were analyzed; Item 19 about "not seeking help at all" and Item 20 about "other help sources" were excluded from the analysis because they do not belong to any domain of formal or non-formal help sources. The overall GHSQ scores have a Cronbach's alpha of 0.837, which means the scale is sufficiently reliable.
The psychometric adaptation process involved 74 participants for the evaluation of two domains of the GHSQ, the Non-Formal GHSQ subscale and Formal GHSQ subscale. The Non-Formal GHSQ consists of seven items and has an overall Cronbach's alpha of 0.840. Item 1, which listed a "Partner" as a help source, was found to be unreliable and invalid (corrected total item correlation: 0.157, Cronbach's alpha if item deleted: 0.874). The authors decided to drop this item because university students may have differing individual relationship statuses and histories; therefore, the tendency to answer this item poses complications and leads to inconsistencies. In addition, Item 2, which listed a "Friend" as a help source, was found to be psychometrically inadequate (corrected total item correlation: 0.4, Cronbach's alpha if item deleted: 0.844). However, the authors decided not to discard this item because its corrected total item correlation score is still deemed to be good. Moreover, it has been found previously that "friend" is one of the most frequently used non-formal help sources by university students (Arria et al., 2011). Reliability analysis of the Non-Formal GHSQ subscale after discarding of the "Partner" item showed an increase in Cronbach's alpha to 0.874, indicating that the scale is valid and reliable. Therefore, six items were used for the Non-Formal GHSQ subscale.
Psychometric testing of the 10 items of the Formal GHSQ showed an overall Cronbach's alpha of 0.755. Item 8, which listed "Psychologist" as formal help source, was found to be psychometrically inadequate (corrected total item correlation 0.238, Cronbach's alpha if item deleted: 0.755). The authors revised this item by combining it with Item 9 "Psychiatrist" as a "Mental health professional," since Indonesians tend to conflate the two professions. Item 15, which listed "Religious leader" as a help source was also found to be psychometrically inadequate (corrected total item correlation: 0.29, Cronbach's alpha if item deleted: 0.759). The authors did not discard this item as there may be a possibility of different help-seeking patterns in different religious groups. Besides, the authors also Note. The frequency and percentage of endorsement from the number of participants who answered "Agree" or "Strongly agree." considered religious leaders as important help sources in Indonesia (Muluk et al., 2018). After refining of the items, the psychometric analysis included nine items in the Formal GHSQ subscale. It showed an overall Cronbach's alpha in the same range of 0.755, which demonstrated that the scale is sufficiently reliable and valid. Thus, nine items were used for the Formal GHSQ subscale.
Psychometric testing was reconducted for the overall GHSQ. The results showed a Cronbach's alpha of 0.844, which is an increase from the previous testing.
This demonstrates that the GHSQ for suicidal ideation is psychometrically sound.
Psychometric Properties. Within the data collection process, the Scale of Suicide Stigma has a Cronbach's alpha of 0.960. The GHSQ without "Not seeking help at all" and "Other help sources" had a Cronbach's alpha of 0.858, with 0.852 for the Non-Formal GHSQ subscale and 0.801 for the Formal GHSQ subscale.

Research
Participants. An online questionnaire consisting of demographic questions, the SOSS, and the GHSQ was shared to a network of both private and public university lecturers and counselors in Jakarta. An age criterion of 18-24 years was established for a more representative depiction of undergraduate university students. The authors also visited several classes in two universities and asked the students to fill out the form at the end of the class. Through this data collection, 329 completed forms were returned. The authors eliminated the data of participants who did not fit the age criteria. The data of those who were not enrolled at Jakarta universities were also eliminated. This process resulted in 284 university students aged 18-24 (M = 20.14, SD= 1.18) whose data were included in the analysis. The participant demographics are presented in Table 1.
Statistical Analysis. For the statistical analysis, SPSS 20 was utilized. Correlation analysis was conducted to test the relationship between two variables, and linear regression analysis was used to test the predictive contribution of independent to dependent variables after a correlation had been established. In addition, ANCOVA was carried out for statistical control on demographic factors (Coolican, 2014
Hypothesis Testing. Correlation analysis was conducted for suicide stigma (M = 79.41; SD = 23.29) and nonformal help-seeking intention (M = 13.80; SD = 4.20). The correlation score was r = 0.297; p = ≤ 0.000. Therefore, suicide stigma has a positive association with non-formal help-seeking intention. Correlation analysis was also conducted for suicide stigma (M = 79.41; SD = 23.29) and formal help-seeking intention (M = 24.02; SD = 5.04). The correlation was r = 0.181; p ≤ 0,001. Therefore, suicide stigma has a positive association with non-formal help-seeking intention. This finding does not confirm Hypothesis 1.
Regression analysis was conducted for suicide stigma and non-formal help-seeking intention. The regression analysis score R 2 = 0.088; F(1,282) = 27.183, p ≤ 0.01. There is an 8.8% contribution of the suicide stigma score to the non-formal help-seeking intention score. The rest of the percentages can be explained by other factors that were not included in this analysis. This means that an increased suicide stigma score can predict an increased non-formal help-seeking intention score (b = 0.05; SE = 0.01; p < 0.001). Each point of the suicide stigma score leads to an increase of 0.053 points of the non-formal help-seeking intention score.
Regression analysis was conducted for suicide stigma and formal help-seeking intention. The regression analysis score was R 2 = 0.033; F(1,282) = 9.359; p ≤ 0.01. There is a 3.3%, contribution of the suicide stigma score to the formal help-seeking intention score. The rest of the percentages can be explained by other factors that were not included in this analysis. This means that an increased suicide stigma score can predict an increased formal help-seeking intention score (b = 0.03; SE = 0.01; p < 0.01). Each point of the suicide stigma score therefore leads to an increase of 0.039 points of the non-formal help-seeking intention score.
ANCOVA was conducted for statistical control against the demographic factors (gender, history of suicidal ideation, religion, ethnicity, and faculty). It was found that the contribution of suicide stigma to non-formal help-seeking intention remained significant after controlling for demographic factors, (F(92,186) = 1.649, p ≤ 0.002). Further, the contribution of suicide stigma to formal help-seeking intention remained significant (F(92,186) = 1.378; p ≤ 0.034). From this finding, Hypothesis 2 is not confirmed. Note. The frequency and percentage of endorsement are obtained from the number of participants who answered "Likely" or "Extremely Likely." Note. The frequency and percentage of endorsement are obtained from the number of participants who answered "Likely" or "Extremely Likely."

Discussion
This study found that suicide stigma has a positive association with both non-formal and formal helpseeking intention, even after demographical factors such as gender, history of suicidal ideation, religion, ethnicity, and faculty are statistically controlled. Both hypothesises are therefore not confirmed in this study. This is a novel finding that is not line with other research results from studies with similar measurement tools, such as done in Australia on a sample of medical students (Chan, Batterham, Christensen, & Galletly, 2014) and the general population (Calear, Batterham, & Christensen, 2014). The finding is also not in line with previous research on various stigmas that constitute barriers to help-seeking intention (Arria et al., 2011;Downs & Einsenberg, 2012;Curtis, 2010;Neilson et al., 2014).
This result could be the consequence of religion playing a significant role in shaping suicide stigma and help-seeking intention in Indonesia. The authors noted that the most endorsed suicide stigma items, which included "Menentang kodrat" / "Unnatural" and "Tidak dapat dibenarkan" / "Unjustifiable," are rooted in religious teachings. Moreover, in the category of formal help sources, the participants have a higher endorsement rate of religious leaders compared to general practitioners. Paranormal/supernatural practitioners are least endorsed, suggesting that there might be religious beliefs that limit the participants in seeking their help. This evidence shows that there is a probability that religion plays an important role in the lives of university students in Jakarta, especially in shaping suicide stigma and the choice of help sources during a suicidal crisis.
The Indonesian religious context and individual religious beliefs are important factors that might explain the unique findings in this study. When individuals have suicidal ideation, those with high suicide stigma may have stronger religious beliefs shaping the stigma. They therefore may feel more compatible with religious formal and non-formal help sources that are easily found in Indonesia. Thus, individuals may perceive the help sources around them in a positive light and increase their help-seeking intention, instead of reducing it like in the Australian community sample. In the Indonesian context, religion is indeed usually used in providing support and caregiving both by formal and non-formal help givers (Muluk et al., 2018). This confirms previous findings indicating that religious values from formal help sources are a very important factor for religious individuals in developing trust of help providers, a positive attitude toward treatment, willingness to access services, and satisfaction with formal services (Weatherhead & Daiches, 2010;Walker, Worthington, Gartner, Gorsuch, & Hanshew, 2011;Iranmanesh, Moghavvemi, Zailani, & Hyun, 2018).
In addition to shaping formal help-seeking intention, religious values help individuals shape their social support network by emphasizing family function, developing experiences of caring networks, and shaping integrated communal religious practices in daily activities (Nelson, Hanna, Houri, & Klimes-Dougan, 2012). Hence, even though individuals have high and negative suicide stigma and acknowledge that suicide is a very stigmatized form of death in their religion, they still have the intention to seek help from non-formal religious sources. This also indicates that individuals with low suicide stigma might not have a strong attachment to religious beliefs. If these individuals perceive the formal and non-formal help sources around them as religious, it might stigmatize their suicidal ideation, therefore also decreasing their intention to seek help. Further studies need to be conducted to explore how religion shapes suicide stigma, help-seeking intention with formal and non-formal sources, and the mechanism and relationship between the three.
Another explanation of the study findings is that individuals have internalized public stigma as self stigma, which might motivate them to seek help as they perceive the negative judgment of society toward suicide death or having continuous suicidal tendencies in the future. Individuals' drive to seek help when they have self stigma may be due to Indonesia's collectivistic culture. In a collectivistic culture, individuals tend to achieve certain goals because of avoidance of negative consequences within their group (Elliot, Chirkov, Kim, & Sheldon, 2001).
This study is limited because it only analyzes helpseeking intention and is not applicable to actual helpseeking behavior. Even though the presence of intention might significantly predict the occurrence of actual behavior (Ajzen, 2005), the help-seeking intention score of individuals when they filled out the questionnaire is based on an imagined scenario about what they would do if they had suicidal ideation. As only 34.5% of the participants had had suicidal ideation within the last year, most participants did not have any help-seeking intention that was shaped by their past experience. The actual help-seeking behavior of the university students might differ from their intention due to many factors present in real-life settings that might moderate/mediate the relationship between suicide stigma and help-seeking behavior. Further research needs to explore the significant role of suicide stigma and help-seeking intention in shaping actual help-seeking behavior in everyday settings.
Besides the association between suicide stigma and help-seeking intention, this study also explored the participants' endorsement of various items on suicide stigma. University students from various faculties (N = 284) gave differing endorsements of the items, ranging from 3.87% to 59.50%. This constitutes a remarkable difference from the findings in Australia, where agreement ranged from 0.7%-39.5% among 676 university students and staff with STEM backgrounds .
The endorsement of "gegabah"/"reckless" (59.50%) and "berpikiran dangkal" / "shallow" (51.40%) by the participants in this study may be based on the myth that people who die by suicide must have a mental disorder (World Health Organization, 2014). This hypothesis is supported by findings that the other types of stigma endorsed by many of the participants are "tidak bertanggungjawab" / "irresponsible" (45.42%) and "kurang akal"/ "senseless" (42.25%). These negative attributes within suicide stigma are also frequently attached to people with mental disorders (Arboleda-Flörez, 2008). Future research should examine how suicide stigma is associated and overlaps with mental illness stigma in Indonesia.
Another type of suicide stigma endorsed by the participants is "menentang kodrat" / "unnatural." This July 2020 ½Vol. 24 ½ No. 1 stigma may be rooted in the religious beliefs held by most university students that deem suicide as a sinful act for trespassing against God's will in deciding human life and death. Moreover, 47.88% of the participants indicated suicide is "unjustifiable." Both types of stigma show that suicide is deemed an irreligious act because it goes against God's will. Given that most participants maintain religious beliefs and religion is an important factor in shaping Indonesian behavior (Muluk et al., 2018), the authors conclude that religion constitutes a vital role in shaping negative stigma toward individuals who die by suicide in Indonesia. This is in line with the historical findings of suicide stigma in many other civilizations with similar religious beliefs (Evans & Farberow, 2003;Tondo, 2014).
This study has successfully mapped help-seeking intention for suicidal ideation among university students in Jakarta. Most students have a greater intention to seek help from formal sources compared to non-formal sources. The students endorse mental health professionals (psychiatrists/psychologists) as the main formal help source (90.84%) and friends (85.91%) as the main non-formal help source. This finding is unlike previous studies such as Arria et al. (2011) and Yap, Reavley, and Jorm (2013), which indicated that parents were the main non-formal help source for youth and university students.
Counselors / university counseling centers constitute the second most endorsed formal help source (79.22%) in the study sample. This might be due to the proximity, affordability, and ease of scheduling for this service. Online counseling services are the third most endorsed formal help source, chosen by 71.83% of the participants. In the last few years, online counseling services have become increasingly common (personal communication with the founders, 2018), including Pijar Psikologi (pijarpsikologi.org), Riliv (riliv.co), and Ibunda (ibunda.id), which were founded in 2015; Save Yourselves (saveyourselves.id), founded in 2016; and Kalm (get-kalm.com), founded in 2018. These counseling services are aggressively promoted through social media and are now among the most accessible help sources, as they rely on internet access utilized by 30 million youths in Indonesia, in particular those who are concentrated in Java and urban areas (Gayatri et al., 2015). Individuals with suicidal ideation tend to seek help in online settings (Seward & Harris, 2016). Therefore, online forums are becoming one of the most favorable help sources for suicidal crises. Nevertheless, there is still a need to evaluate the effectiveness of these services (Arjadi, Nauta, Chowdhary, & Bockting, 2015).
The participants chose mental health community cadres or volunteers as the fourth most endorsed formal help source (71.83%). Since the rise of voluntary mental health community groups in Indonesia, such as Komunitas Peduli Skizofrenia Indonesia which was founded in 2009, mental health communities often serve as help sources and support families and patients with shared experiences of mental illness (Yasmine, 2018). Most of the participants (57.74%) also chose human resource counselors/staff as a formal help source even though only 15.5% of them worked at the time of the study. This might be because university students consider Employee Assistance Programs as one of the main programs of human resource staff.
A unique finding is that 62.67% of the participants indicated they would seek help from a religious leader if they had suicidal ideation. This percentage of endorsement is even higher than that for general practitioners, which is 35.91%. This is interesting because religious leaders are the only formal help source to be endorsed by most participants, regardless of the fact that they do not have any background in mental health or psychology. As religion is already deeply rooted within Indonesian culture, religious leaders play an important role, including in assisting lay people with health issues (Muluk et al., 2018).
Only a few participants endorsed traditional medicine.
Only 15.14% of the participants chose traditional herbal practitioners, while 13.38% chose body technique practitioners and 8.8% chose paranormal/ supernatural practitioners, who were the least chosen help source. This pattern might be due to the perception among Indonesians that body technique and herbal medicine practitioners are more specialized in physical health problems such as infections (Nugraha & Keller, 2011), malaria (Suswardany, Sibbritt, Supardi, Chang, & Adams, 2015), and cancer (Jauhari, Utami, & Padmawati, 2008). Nevertheless, even though paranormal/supernatural practitioners specialize in spiritual aspects, they are the least popular option because 1) they are considered as heretics in a society dominated by Abrahamic religions, and 2) supernatural practices are deemed unscientific by university students who are exposed to scientific thinking (Sutiono, 2014).
Among all the non-formal help sources, 85.91% of the participants endorsed friends. This may be because friends are important figures who shape and develop adolescent health behaviors (Patton et al., 2016) as well as the identity formation of young people (Sarwono, 2014). After friends, the other most endorsed nonformal help sources are siblings (57.39%) and parents (52.81%). This pattern may appear because many university students in Jakarta have migrated from other regions/cities, which makes it easier for them to seek help from their peers or friends. Siblings are more endorsed than parents, probably due to the inconvenience caused by the generation gap when discussing issues related to suicide.
Besides friends and the nuclear family, other nonformal help sources that are less endorsed by the participants are cousin (36.97%), uncle/aunt (26.05%), and grandparent (21.12%). The non-formal helpseeking pattern of relying on extended family members may appear because Indonesian families tend to live with their extended family (Sarwono, 2014).
This study is limited in that the findings based on data from a small number of participants in Jakarta may not be generalizable for university students as a whole. In addition, there is no analysis of variables related to religion in shaping help-seeking intention. Future research should examine the influence of religiousity, religious commitment, and other religion-related variables on suicide stigma and help-seeking intention. Self suicide stigma can also be analyzed as a mediator between public suicide stigma and help-seeking intention.
This study does not investigate the participants' relationship status, such as their romantic relationship status, or their migration status in Jakarta, which may shape different patterns related to help sources. To attain more rigorous findings on cultural and geographical differences, university students from Jakarta should be compared to university students elsewhere. This study is also limited to examining the patterns of university students' suicide stigma and help-seeking intention; future studies may use a qualitative approach to explore the meaning of suicide death, stigma toward people who die by suicide, and help-seeking process among diverse sub-groups of university students from diverse cultural backgrounds.
There are several practical implications of this study. Psychoeducation is one of many techniques that can be conducted to reduce stigma (Arboleda-Flörez, 2008). The findings on suicide stigma type can be used to develop educational activities with the aim of reducing stigma and debunking myths related to suicide. Based on the findings, psychoeducational material development should focus on addressing the most endorsed stigmas among university students, such as "gegabah," "menentang kodrat," and "berpikiran dangkal." In addressing "gegabah" and "berpikiran dangkal," psychoeducational material may emphasize the understanding of suicide as a consequence of the inability to handle pain and trauma that occur in other people, including those with great talents in the arts, sciences, and humanities and who make positive impacts in society. In addressing "menentang kodrat," the involvement of religious leaders may be beneficial, especially in eliciting a compassionate response to those who die by suicide instead of emphasizing "menentang kodrat" for trespassing against God's will.
The findings on formal help sources can constitute a basis for providing training on suicide prevention by these specific sources: mental health professionals, counselor/university counseling centers, online counseling services, and mental health community cadres/volunteers. The scope of materials of these training initiatives should be adapted to the competence and field of each profession. More socialization needs to be done to inform the availability of these sources. Because religious leaders are an important formal help source, it is also recommended to reinterpret ate religious teachings related to help seeking in distressing events to be used as religious preaching materials to open more constructive and supportive discussions on suicide prevention in religion-based university student organizations. Other traditional practitioners can also be trained to detect suicidal ideation in patients, so that they can be referred to other professionals.
Based on the findings on non-formal help sources, universities can empower students to become peer counselors for their friends by providing suicide crisis intervention training. Outside university settings, nuclear and extended families of university students also need to be informed about suicide crisis intervention, especially by family therapist/psychologists who support suicidal university students.
Even though the study found that high suicide stigma is associated with high help-seeking intention and predicts an increase in help-seeking intention, the authors do not recommend any stigmatization toward those who die by suicide. This is because the effect of stigma affects not only those who die by suicide or display suicidal tendencies, but also the bereaved family and irreligious people with suicidal ideation. The authors recommend creating more direct contact in public spaces with people with lived experiences of suicidal ideation so that lay people will have a more comprehensive understanding about this condition, as previous evidence indicates that interpersonal contact may reduce the stigma (Arboleda-Flörez, 2008). This is also recommended to be conducted in religious settings. Efforts to provide this kind of contact may involve collaboration with local mental health community volunteers, especially those who have lived experiences of suicidal ideation.

Conclusion
This study found that suicide stigma is positively associated with help-seeking intention and predicts the help-seeking intention of university students in Jakarta toward non-formal and formal help sources. The Indonesian cultural context, which includes collectivism and religiousity, may be a factor in this unique finding. This study is the first to reach findings of this sort. Future research should be conducted to add more knowledge on suicide stigma and other factors that may July 2020 ½Vol. 24 ½ No. 1 contribute to help-seeking intention in a suicidal crisis for Indonesian university students. On a practical level, the findings on suicide stigma and help sources for university students in Jakarta can be a basis for various programs that enhance skills and awareness for suicide prevention.