Acknowledgements: to the staff of LYAP, without whose help this research would not have been possible.

Funding: University of Hong Kong Small Project Funding. Project Number  200507176179

Objective of research

Western studies on transpeople reveal experiences of transphobia impacting on their quality of life and upon their health, both mental and physical. Little is known about the lives of transpeople in Asia. No research has been done on the lives of Lao transpeople, except that of Doussantousse (2005). This research, paralleling similar studies already done in Thailand and the Philippines (Winter and Vink, a and b) was aimed at an examination of the quality of life and mental health of a sample of transwomen in Laos.

Background situation of TG in Laos

Lao PDR (People’s Democratic Republic) is a country of 5.3 million inhabitants, landlocked between Thailand, Cambodia, Vietnam, and China. Linguistically, historically and culturally it has ties with Thailand, especially north and north-eastern Thailand. At times under the control of Thailand and then France, a centuries-old monarchy was deposed in 1975 by the Pathet Lao revolutionary forces in 1975. The socialist regime has allowed limited economic liberalisation since 1986. Vientiane is the capital.

Transgenderism is a universal phenomenon. Throughout the world, and throughout history, certain persons designated male at birth have grown up to question their identity as male. Some identify and present as female (or sometimes, perhaps more subtly, as transgendered females). Certain persons designated female at birth have grown up to live lives as men (or transmen).

Doussantousse (2005) has reported a transgendered community in Laos, describing large numbers of transwomen who identify and present as females, or in many cases live openly as transwomen, growing their hair long, dressing and behaving as other women do. Where they have attracted the attention of international NGOs working in the field of HIV prevention, they have often been labelled as MSM (Men who have Sex with Men); this for the obvious reasons that they were designated male at birth, and in most cases, are erotically attracted to men.

As often is the case in Thailand, transwomen are often called ‘kathoey’. As in Thailand (Jackson, 1999) kathoey enjoy a certain degree of tolerance. Neither governmental agencies nor broader Lao society pays much attention to them. They are largely left alone; there are few  reports of discrimination or segregation against them.

The questionnaire

Data was collected by way of a Lao-language questionnaire consisting of closed- and open-ended items falling into several categories, as follows:

a.  Background: age; weight; height; ethnic group; religion; education; employment; siblings.

b. Gender identity and presentation: actual gender identification; preferred gender identification; anticipated gender presentation when older; cross-dressing; hair-style; hormones; surgery/implants; sex reassignment surgery,

c. Transition-related issues:  ability to pass; persons who know of respondent’s trans-status; their responses to the respondent’s trans-status; respondent’s satisfaction with own appearance.

d. Impact of being trans: positive and negative impact on respondent’s life, ( a ) when a child (family, friends and school), and ( b ) nowadays (in family, among peers, in education, employment, health services, and in society viewed broadly).

e. Mental health: self-esteem; fear of negative evaluation; depression; loneliness; subjective well-being; ease with trans- identity; suicidal thoughts and behaviour. 

f. Social support: people with whom respondent spends spare time, and to whom she can call upon for help.

h. Miscellaneous other: respondent’s own explanations for her trans-status; own views about the status of transgender as a difference or as a disorder; sexual preference, autogynephilic experiences. Items in this category fell outside the main focus of the study, which was to examine quality of life and mental health of transwomen, and will be the subject of other reports.

The items generated 213 primary variables, many of which in turn enabled  the calculation of numerous secondary variables, to be described later.

All items paralleled those of recent  studies by Winter and Vink (a and b) in Thailand and the Philippines. Some, especially those in categories (a ) to ( b ) and ( h),  were in turn derived from earlier research studies conducted by Winter (2006a and b) and Winter, Sasot and King (submitted) in those countries.

Items for Category ( d ) were initially developed from the senior author’s research notes, compiled over several years,  regarding impact of trans-status in S.E.Asia. They were refined after feedback from ten trans-people across six  S.E. Asian  countries (see acknowledgements).  The resulting modified list, used already in Thailand and the Philippines, was scrutinised for relevance to Laos, and was overall found appropriate.

Items for Category ( e ) were drawn, wherever appropriate, from existing mental health instruments, Self-esteem was measured by way of the Rosenberg Self Esteem Scale (Rosenberg, 1965) used in its entirety (ten items). Depression was measured by a shortened form  of the Centre for Epidemiologic Studies Depression Scale (CES-D scale; Radloff, 1977). Four of the more somatic (and therefore more culturally specific) items were discarded, leaving 16 out of the original 20 items. Loneliness (emotional and social) was measured by way of the Emotional and Social Loneliness Scale (Vincenzi & Grabosky, 1987), used intact (ten items). A key component of social anxiety was measured by way of a shortened form of the Fear of Negative Evaluation Scale (Watson & Friend, 1969) as modified by Leary (1983). Seven items were retained for this study, those discarded being the more linguistically and conceptually difficult ones, and therefore considered possibly problematic for some participants.

A uniform response format was employed for most items in the mental health category; a five-point scale comprising ‘all of the time’, ‘most of the time’, ‘some of the time’, ‘seldom’, ‘never’. For the Rosenberg Self-esteem Scale and the Fear of Negative Evaluation Scale this entailed a change to their usual response format. Two items provided exceptions to this rule. One was an item measuring overall well-being in one’s life, the response-format here being a five-point scale from ‘very happy’ to ‘very unhappy’. The other was an item asking about suicidal attempts, with a three-point response-format comprising ‘more than once’, ‘once’ and ‘never’.

Throughout the questionnaire items were recoded where necessary so that high scores indicated prima facie positive aspects of a respondent’s life; i.e. openness, transition, passing, acceptance, positive impact of being trans-, good mental health, social support.

From all this data it was possible to calculate several secondary variables, as follows:

OTHKNOW: indicating the degree to which the respondent openly lived as a transperson, and operationalised as the number of key people in the respondent’s life who knew she was trans-. Scores were available for each of the following: father, mother, brothers, sisters, other relatives, friends, partners, current employer, strangers. The possible score range was 0 to 17.

TRANSIT: indicating the degree to which the respondent had transitioned, and operationalised in terms of ( i ) cross-dressing (including the number of key people in front of whom she cross-dressed - father, mother, siblings, other elatives, friends, partners, current employers and strangers), ( ii )  growing hair long, ( iii ) taking hormones, ( iv ) undergoing sex reassignment surgery, and ( v ) other assorted surgery and injections. The possible score range was 0 to 46.

REACT: indicating the degree to which key people supported and encouraged (or alternatively accepted, tolerated or rejected) the respondent’s trans life. Scores were once again available for each of the following: father, mother, brothers, sisters, other relatives, friends, partners, current employer, strangers. The possible score range was 0 to 36.

TREAT: indicating the degree to which key people treated the respondent as female (or alternatively as male or as some other sort of social category). Scores were once again available for each of the following: father, mother, brothers, sisters, other relatives, friends, partners, current employer, strangers. The possible score range was 0 to 18.

PASS: indicating the degree to which the respondent is able to pass as a ‘genetic female’ in different situations: at a distance, close-up with make-up, close-up without make-up, and when they talk. The possible score range was 0 to 12.

SOCSUPP: indicating the degree of social support, in terms of key people with whom the respondent could spend spare time and call on for help. Categories were family, female friends, male friends, trans friends, and others. A further (negative) category was ‘no one’.   The possible score range was 0 to 12.

poPSTim1:  indicating the positive impact of being trans  during the respondent’s childhood. Eight items covered family, friends and school. Any items for which being trans made life ‘better’ were summed. This sum was then divided by the number of valid items for that respondent (i.e. allowing for cases in which the respondent had no father, mother or siblings etc).  The possible score range was 0 to 1.

nePSTim1 (1): a score corresponding to poPSTim1, but  indicating  the negative impact of being trans during the respondent’s childhood. Any items for which being trans made life ‘worse’ were summed, and then divided by the number of valid items for that respondent.  The possible score range was again 0 to 1.

poCURim1  and neCURim1. These scores corresponded to poPSTim1 and nePSTim1, but covered the contemporary impact of being trans. Forty items covered family, peers, education, employment, health services, and broader society. In each case the possible score ranges were 0 to 1.

SESTEEM2: this score was based on the items from the Rosenberg Self-Esteem Scale.  Ten items had been originally been included in the questionnaire, but the data showed that two of them (Items 5 and 8) did not fit well with the other items; either because of issues relating to translation, culture or subculture. Data for these two items was discarded, with the effect that the Cronbach alpha rose from 0.595 to 0.676 for this scale. The possible score range was 0 to 32. , 

FREEDEP: this score was calculated from the CED-Depression Scale items. All items were recoded so that a high score indicated good mental health. As a result this variable represents freedom from depression. Cronbach alpha was 0.826.   The possible score range was 0 to 64.

FREELONE2: this score was based on the items from the Emotional versus Social Loneliness Scale. All items were recoded so that a high score indicated good mental health. . As a result this variable represents freedom from loneliness.  Ten items had originally been included, but the data showed that Items 5, 8, 9 and 10 did not fit well with the other items. Data for these four items was discarded, with the effect that the Cronbach alpha rose from 0.506 to 0.588 for this scale. The possible score range was 0 to 24. 

FREEneEV: this score was based on the seven items of the Fear of Negative Evaluation Scale. All items were recoded so that a high score indicated good mental health.  As a result this variable represents freedom from fear of negative evaluation. Cronbach alpha was 0.823 for this scale. The possible score range was  0 to 28.   

A few other single items within the mental health category were used to generate variables thought to be important in any further analysis of the data. They were as follows:

WELLBEING: a single item asking the respondent to indicate how happy she felt about her life (a five point scale running from ‘very happy’ to ‘very unhappy’). The possible score range was 0 to 5.

SATISAPP: a single item asking the respondent to indicate the extent to which she felt satisfied with her appearance (a five point scale running from all the time’ to ‘none of the time’). The possible score range was 0 to 5.

EASEIDEN: a single item asking the respondent to indicate the extent to which she felt comfortable with people knowing she is trans- (a five point scale running from ‘all of the time’ to ‘none of the time’.. The possible score range was 0 to 5.

FREEsuID: a single item asking the respondent to indicate how often she has thought about committing suicide. The possible score range was 0 to 4 (a five-point scale from ‘all the time’ to ‘none of the time’)  with a high score indicating freedom from suicidal thoughts.

FREEsuBE: a single item asking the respondent to indicate how often she has tried to commit suicide. The possible score range was 0 to 2,  with a high score indicating the respondent had never attempted suicide, and a low score indicating more than one attempt.

The sampling method

Respondent-driven sampling or RDS (Heckathorn, 1997) was used to provide  the sample for this study. RDS requires that researchers recruit a small number of ‘seed’ respondents within the population being researched. They provide data (for example, complete a questionnaire), receiving a small payment for doing so (the primary reward). They are then given a predetermined number of vouchers, each marked in a unique way, which they then pass to others in their personal network who fit the same criterion (i.e. identify as ‘kathoey’). When a ‘second-wave’ respondent approaches the researchers and provides data, she hands in her voucher and the ‘seed respondent’ who recruited her then receives an additional payment (the secondary reward). Each second-wave respondent is then given a set number of vouchers and the process continues, with these respondents receiving their secondary rewards when the third-wave respondents provide data. In this way, with just one ‘seed’, and each respondent recruiting three further respondents, the sample size can reach 364 in just six waves of data.

Heckathorn (op.cit.) notes that RDS provides anonymity for respondents, since no one is ever identified to the investigator, but is rather recruited into the study. He argues that by providing incentives, tapping peer networks and incorporating peer pressure, RDS provides a way of accessing otherwise difficult-to-reach or unwilling respondents and providing a representative sample, independent of the initial ‘seed’ respondents (op.cit.). RDS, it is argued,  can even be used to estimate the size of populations, including those which are ‘hidden’ and whose size is therefore normally most difficult to estimate (Heckathorn, 2002; Salganik & Heckathorn, 2004). RDS has used to study populations as varied as jazz musicians (Heckathorn and Jeffri, 2001), Injection Drug Users (Heckathorn, Semaan, Broadhead and Hughes, 2002) and gays (Ramirez-Valles, Heckathorn, Vazquez, Diaz, & Campbell, 2005). This present study represents the first application of RDS to a transgender population.

The procedure

This research was made possible through the involvement of the Lao Youth Aids Project (LYAP), a Buddhist agency working with the kathoey community for many years now, and with an office base close to central Vientiane. LYAP’s programmes in HIV prevention have led it to set up a group, exclusively of transgendered kathoey, who meet regularly and carry out educational activities for other kathoey. Four of its members supervised development of a Lao version of the questionnaire, as well as data collection.  After reviewing and amending the translation of the questionnaire, they organised and supervised the completion of questionnaires by kathoey respondents. All respondents participated voluntarily in the survey.

Translation proceeded as follows. The English version of the questionnaire (already used for development of questionnaires employed in Thailand and the Philippines (Winter and Vink, 2006a and b) was translated into Lao and then reviewed and amended by the LYAP kathoey group, then back-translated into English by a second independent translator.  The research team then reviewed the back-translation, with a further review by the first author. Eventually, the two translators and kathoey research team met and finalized the Lao Version.  Six questionnaires were tested on a small pilot sample, and then, with further minor changes being made, 220 copies of the final questionnaire were  duplicated for use in the research proper.

Five kathoey were selected as ‘seed respondents’. They completed the questionnaire, receiving US$2 for doing so. Each was then given two vouchers, which they could pass on to further potential ‘referee’ respondents. Any of those referee respondents, upon coming to the LYAP office to complete the questionnaire, received US$2, plus, in their own turn, two vouchers for them to pass on to other potential respondents. The respondents who had referred them earned US$3 for each. And so on.

In 12 days over a 17 day period, 217 Kathoey joined the project and filled the questionnaire.  The data were entered using SPSS in a “double entry” methods.

Findings

A large number of transwomen live in Vientiane, forming a fairly well-defined community; so much so that within a space of 12 days it was possible to collect data on 217.

Demographic information

The sample was aged 12 to 44 years old (average age 20.82 years), exclusively Lao and overwhelmingly Buddhist. Around 44% were students, and 23% were unemployed. The rest had jobs of some kind.

Identification

The most common category by which they identified was ‘kathoey’ (54%), with others identifying as ‘third sex’ (19%) or ‘woman of the second kind’ (17%). Relatively few identified as ‘woman’ (7%). Average age of first identifying in this way was 10.8 years. The majority indicated a preference to have been born a woman in this life (51.2%, and to be born one in the next life (67.7%). While the majority anticipated presenting at age 50 as they do now (58.1%) some believed they might be able to present as natal women (26.7%). Some believed they would be presenting as men (13.4%).

Transition

Between 60 and 70% reported that their parents knew of their transgender, on average learning when their child was around ten years old. Outside the family even more seemed to know about their status (for example around 90% of employers). fewer appeared to know knew. The majority (54%) dressed as female (on average doing so from 12.7 years). Others would if they felt able to (26%). Around 38% had grown their hair long (on average from 15.5 years), with another 42% saying they would if able to wanting to do so ) Around 36% were taking hormones (on average from 16.7 years, and in one case from age 8), with another 25% wanting to do so if able.

Social relationships

Within the sample there was great variation in the degrees ( a ) to which respondents had transitioned,  ( b ) to which other people knew respondents were trans, ( c ) to which they could pass in front of strangers, ( d ) to which others responded to the respondent as female (or trans), ( d ) to which others accepted their transgenderism. There were also great differences in the degree of social support that respondents could call on, both in terms of how they spent their spare time, and in whom they could call on for help. 

Impact of being transgendered

The impact (upon respondents’ lives) of being trans were markedly positive during childhood. Evidence for positive impact outweighed that for negative impact by a factor of 7.22. During adulthood positive impact abated somewhat, while negative impact increased greatly. As a result positive impact, though still outweighing negative impact, did so only by a factor of 2.02.

Mental health

Respondents varied greatly in terms of ( a ) self-esteem, ( b ) freedom from depression, ( c ) freedom from loneliness, ( d ) freedom from fear of negative evaluation, ( e ) well-being, ( f) satisfaction with appearance, ( g) ease with being known as transgendered, and ( h) freedom from suicidal thoughts and behaviour. Suicidal behaviour was worryingly common, with 12% reporting that they had attempted suicide at least once in their lives.

Explanations for being transgendered

The most common explanations for being transgendered were, in order of perceived importance: ( a ) influence of trans friends (63%); ( b ) influence of female friends (55%); ( c ) something biological (46%); ( d ) parents’ influence (36%); ( e ) other relatives’ influence (30%);  ( f ) influence of male friends (30%);  ( g ) sisters’ influence (25%); ( h ) brothers’ influence (20%); ( i ) karma (18%).

Beliefs about the nature of transgenderism

Around 50% saw transgenderism as a psychological problem, 44% as a physical problem, 30% as a moral problem. Around 61% appeared to think it was no problem at all, merely a difference.

Sexuality

The vast majority (94%) reported that they were sexually attracted to males, and around 5% to females. Only a few respondents reported autogynephiliac tendencies, with only 35% (of the valid responses) indicating any history of sexual excitement by wearing female clothes, and 34%  indicating any sexual excitement at the thought of oneself as a woman. 

Tables

Demographic information

Identity

Transition

Social Relationships

Impact of being transgendered

Mental health

Explanations for being transgendered

Beliefs about the nature of transgenderism

Sexuality

Relationships between variables: sample findings

N.B. These are selected preliminary findings only – a more comprehensive report is being compiled for publication

A. Impact of being TG: effects on mental health.

Extreme groups (top scores and bottom scores) were identified for the four ‘impact’ variables (poPSTim1, nePSTim1, poCURim1 and neCURim1). The two extreme groups for each variable (top and bottom 25%) were then examined for differences in scores for mental health variables (SESTEEM2, FREEDEP, FREEneEV, FREELONE2, WELLBEING, SATIS, EASEID, FREEsuID, FREEsuBE). Where possible the top 25% scorers for each impact variable were compared with the bottom 25% scorers. Independent t-tests (unequal variances assumed) produced the following findings (all at significance at p less than 0.05).

Positive current impact of being TG (as measured by poCURim1). The top 25% of the sample (3^rd quartile cut-off value for poCURim1 = 0.437) were compared against the bottom 25% (1^st quartile cut-off = 0.148). Significant differences were found for subjective well-being (WELLBEING) and self-esteem (SESTEEM2). High poCURim1 scores were associated with high WELLBEING and SESTEEM2 scores.

Negative current impact of being TG (as measured by neCURim1). The top 25% of the sample (3^rd quartile cut-off = 0.233) were compared with the bottom 25% (1st quartile 0.052). Significant differences were found for freedom from suicidal behaviour (FREEsuBE), satisfaction with appearance (SATIS), freedom from loneliness (FREELONE2) and ease with TG identity (EASEID). High scores for neCURim1 were associated with  low FREEsuBE, SATIS, and FREELONE2.

For the other two variables it was not possible to simply compare the top and bottom 25% of the sample.

Positive childhood impact of being TG (as measured by poPSTim1). The top 25% of the sample (3^rd quartile cut-off value for poPSTim1 = 0.667) were compared against the 30% of sample for whom poPSTim1 = 0. Significant differences were found for subjective well-being (WELLBEING) and satisfaction with appearance (SATIS). High poPSTim1 scores were associated with WELLBEING and SATIS scores.

Negative childhood impact of being TG (as measured by nePSTim1). The top 18% of the sample who reported any negative impact were compared against the 82% who reported none. Significant differences were found for freedom from depression (FREEDEP), freedom from suicidal thoughts (FREEsuID) and freedom from loneliness (FREELONE2). High nePSTim1 scores were associated with low FREEDEP, FREEsuID and FREELONE2 scores.

B. Overall predictors for mental health

Multiple Regression Analysis (MRA) was used to examine the degree to which any or all of the following variables predicted the various mental health scores. Predictors examined were OTHKNOW, TRANSIT, REACT, TREAT, PASS, poPSTim1, nePSTim1, poCURim1, neCURim1, SOCSUPP, (all of which can be regarded as ‘external’ variables, relating to features external to the person concerned, roughly corresponding to quality of life) and SESTEEM2, FREELONE2, FREEneEV, FREEDEP, WELLBEING, SATISAPP, EASEID (all of which are ‘internal’ variables, relating to the person’s mental health). Because of the size of the sample, and number of potential predictors, a three-step process was followed. First of all, predictors among internal variables were examined  using a stepwise MRA. Second, external variables were examined, again to identify predictors. Third, important external and internal predictors were combined in a final stepwise MRA analysis.

Results of these MRA analyses are as follows:

Self-esteem (as measured by SESTEEM2) was predicted by freedom from loneliness (FREELONE2), freedom form depression (FREEDEP), satisfaction with appearance (SATIS) and ease with TG identify (EASEID). The adjusted R squared was 0.434. All effects (as indicated by beta values) were positive.

Freedom from depression (as measured by FREEDEP) was predicted by freedom from negative evaluation (FREEneEV), self-esteem (SESTEEM2) and freedom from loneliness (FREELONE2). The adjusted R squared was 0.574. All effects (as indicated by beta values) were positive.

Freedom from loneliness (as measured by FREELONE2) was predicted by  self-esteem (SESTEEM2), freedom from depression (FREEDEP) and satisfaction with appearance (SATIS). The adjusted R squared was 0.412. All effects (as indicated by beta values) were positive.

Freedom from negative evaluation (as measured by FREEneEV) was predicted by  freedom from depression (FREEDEP), positive impact of TG in childhood (poPSTim1), and ease with TG identity (EASEID). The adjusted R squared was 0.471. The effects the first two predictors (as indicated by beta values) were positive. The effect for the third one (EASEID) was negative.

Subjective well-being (as measured by WELLBEING) was predicted by  positive current impact of being (poCURim1), freedom from depression (FREEDEP), and negative impact of being TG in childhood (nePSTim1). The adjusted R squared was 0.092. The effects of the first two predictors (as indicated by beta values) were positive. The effect for the third one (nePSTim1) was negative.

Satisfaction with appearance (as measured by SATIS) was predicted by self-esteem (SESTEEM2), the degree to which others knew one is TG (OTHKNOW), positive current impact of being TG (poCURim1), ease with TG identity (EASEID), and freedom from loneliness (FREELONE2). The adjusted R squared was 0.232. All effects (as indicated by beta values) were positive.

Ease with TG identity (as measured by EASEID) was predicted by degree to which others knew one is TG (OTHKNOW), satisfaction with appearance (SATIS) and degree to which one had transitioned (TRANSIT). The adjusted R squared was 0.159. All effects (as indicated by beta values) were positive.

Freedom from suicidal thoughts (as measured by FREEsuID) was predicted by freedom from depression (FREEDEP), subjective well-being (WELLBEING) and self-esteem (SESTEEM2). The adjusted R squared was 0.119. All effects (as indicated by beta values) were positive.

Freedom from suicidal ideas (FREEsuID) was predicted by only one variable, subjective well-being (WELLBEING). The adjusted R squared was 0.116, and the beta value was positive.

C. Mental health and beliefs about the nature of transgender

The belief that transgender was a disorder was associated with poorer mental health. This was true regardless of whether respondents thought it was a mental, physical or moral problem. In each case it was self esteem, depression, loneliness and fear of negative evaluation that were affected. Those who believed transgender is a mental problem scored lower on SESTEEM2 (-0.142*), FREEDEP (-0.265**), FREELONE2 (-).242**) and FREEneEV (-0.203**).  Corresponding correlations for ‘physical problem’ were -0.144*, -0.279**, -0.250** and -0.215**. Those for ‘moral problem’ were -0.252**, -0.315**, -0.270** and -0.236**..

D. Age and age of first hormone use.

Among  the 75 respondents who had taken hormones there was a clear tendency for the younger ones to have begun taking hormones at a younger age. The correlation between the relevant variables (age and age of first hormone use) was 0.601 (significant at p less than 0.01 level). The regression coefficient indicted that for every five years that passes, age of first hormone use falls by around one year.

Conclusions

There is a large and well-interconnected community of transwomen in Vientiane, members of which have transitioned to different levels and enjoy varying levels of social acceptance. Overall, the impact of being transgendered upon respondents’ lives seemed positive, particularly in childhood. Positive and negative experiences seemed more evenly balanced in adulthood.

Respondents displayed varying levels of mental health. There was evidence that one’s experiences of being TG, both in childhood and adulthood, influenced one’s mental health (self-esteem, depression, fear of negative evaluation, loneliness, general well-being, satisfaction with one’s appearance, ease with being known as a transwoman, and suicidal thoughts and behaviour).

In addition there was evidence that certain mental health measures were predicted by one’s degree of gender transition, the degree to which key others knew one was transgendered, the degree to which one was comfortable being known as transgendered and the degree to which one was satisfied with one’s appearance.

Further, there were indications that one’s beliefs about the nature of transgenderism predicted certain mental health measures (self-esteem, depression, loneliness, and fear of negative evaluation). Beliefs that transgenderism was a disorder (mental, physical or moral) predicted poorer mental health.

Some evidence was found to suggest that Lao transwomen are, as time progresses, using hormones earlier than was hitherto the case.

The data is still being analysed.

References

Doussantousse, S. (2005). Male sexual health: kathoeys in the Lao PDR, South East Asia - exploring a gender identity.   Retrieved 28th June, 2006, from http://web.hku.hk/~sjwinter/TransgenderASIA/paper_doussantousse.htm

Heckathorn, D. (1997). Respondent-driven sampling: a new approach to the study of hidden populations Social Problems, 44(2), 174-199.

Heckathorn, D. (2002). Respondent-driven sampling II: deriving valid populations estimates from chain-referral samples of hidden populations Social Problems, 49(1), 11-34.

Heckathorn, D., & Jeffri, J. (2001). Finding the beat: using respondent-driven sampling to study jazz musicians. Poetics, 28, 307-309.

Heckathorn, D., Semaan, S., Broadhead, R., & Hughes, J. (2002). Extensions of respondent-driven sampling: a new approach to the study of injection drug-users aged 18-25. AIDS and Behavior, 6(1), 55-67.

Jackson, P. (1999). Tolerant but unaccepting: the myth of a Thai 'gay' paradise. In P. a. C. Jackson, N. (Ed.), Gender and sexualities in modern Thailand. Chiangmai: Silkworm Books.

Leary, M. R. (1983). A brief version of the Fear of Negative Evaluation Scale. Personality and Social Psychology Bulletin, 9, 371-375.

Radloff, L. S. (1977). The CES-D Scale: a self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385-401.

Ramirez-Valles, J., Heckathorn, D., Vazquez, R., Diaz, R., & Campbell, R. (2005). From networks to populations: the development and application of respondent-drivern sampling among IDUs and Latno gay men. AIDS and Behavior., 9(1), 387-402.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

Salganik, M., & Heckathorn, D. (2004). Sampling and estimation in hidden populations using respondent-driven sampling Sociological Methodology, 34(1), 193-240.

Vincenzi, H., & Grabosky, F. (1987). Measuring the emotional/social aspects of loneliness and isolation. Journal of Social Behavior and Personality, , 2 ((2, Part 2)), 257-270.

Watson, D., & Friend, R. (1969). Measurement of social-evaluative anxiety. Journal of Consulting and Clinical Psychology, 33(448-457).

Winter, S. (2006a). Thai transgenders in focus: demographics, transitions and identities. International Journal of Transgenderism, 9(1), 15-27.

Winter, S. (2006b). Thai transgenders in focus: their beliefs about attitudes towards and origins of transgender. International Journal of Transgenderism, 9(2).

Winter, S., Sasot, A., & King, M. Transgender in the Philippines: a close focus.

 Home