Tristan, a college student, is 21 years old.

James, a graduate student, is 32 years old.

Jennifer, an aerospace engineer, is 38 years old.

Virginia, a writer, is 48 years old.

John, a systems administrator, is 24 years old.

Felicia, a human resources consultant, is 50 years old.

Stephen, a real estate salesman, is 30 years old.

Elizabeth, a machine repair technician, is 51 years old.

Allison, who is currently disabled and unemployed, is 48 years old.

Christina, a scientist, is 34 years old.

Travis, a legal assistant, is 25 years old.

Nathaniel, who works with Automated Data Processing, is 23 years old.

Mark, a college student, is 21 years old.

Scott, who works in retail, is 19 years old.

Conner, an assembly line worker, is 23 years old.

The following is a detailed account of my methods in gathering responses to the Transgender/Transsexual Sexual Experiences Survey, and the likely reasons for my failure to collect a larger number of responses. It is my hope that such a candid and straightforward account may be of use to other inexperienced researchers.

In designing the TTSES I set out, as I have explained above, to avoid a number of what I regarded as serious methodological limitations of previous research into the sexual experiences of trans people -- including very small and nonrepresentative samples. Of course, my research would at least have the advantage of being non-clinical in nature (see my analysis in Part I about the problems of the clinical setting), and what I saw as the further advantage of having a trans researcher behind it; I hoped that these factors, in combination, would ensure that individuals would be both more likely to participate and more likely to be honest and forthcoming in their response. But more than this, I hoped to have a considerably larger sample than many of the studies I had come across in my reading; idealistically, I was aiming for one hundred respondents.

I did not think this to be an unrealistic goal, and perhaps, all other things being equal, it was not. After all, there are millions of trans people in the United States alone, and large numbers of them are connected to some kind of community, be it a support group, political organization, newsletter or Internet listserv. Through such channels, I thought, it ought to be relatively easy to spread word of my research project and generate a fair amount of interest.

From the beginning my project had serious limitations, some of which were unavoidable, and some of which I created myself, perhaps unwisely. First and foremost was funding. While I was able to obtain a $150 research grant -- and this money more than covered a six-month Post Office Box, envelopes and postage, and a few other necessary items -- it was not enough to purchase an advertisement in the nation’s foremost trans community publications. Lack of funds combined with the requirements of my academic calendar to prevent me from visiting conferences at which I could have sought respondents. Nor did my small grant seem to be enough to offer any material incentive to prospective respondents. Indeed, with the number of respondents I anticipated, it was not enough to offer return postage to them all; likewise, the advice from one respondent that I enclose a dollar bill with each posted survey, in order that recipients feel a little more obligated to return the survey, did not seem feasible either. In retrospect, of course, both of these techniques would have been feasible if I had been anticipating only 25 or even 50 responses -- instead of shooting for 100 and ending up with 15. But I am getting ahead of myself.

The second unavoidable limitation of the project was my lack of credentials that would help assure both potential respondents and those whose help I required in promoting the project -- the managers of listservs, for example -- that I was, in fact, nothing other than an honest and trustworthy academic researcher. This problem was exacerbated by the insistence of the research ethics committee overseeing my project protocols that I set up a separate email account for the project, and not use my name in the announcements; anonymous posts were viewed with suspicion by some community members and list administrators.

A third, also unavoidable factor that may have limited participation was the explicitly sexual nature of the data I was seeking. Finding participants for sex research always poses this problem, which results not only in difficulty in finding large numbers of people willing to speak openly and in detail about their lives, but in a serious self-selection problem as respondents tend to be more comfortable with sexual matters, more liberal in their views on sexuality, and even more sexually active. This problem was likely compounded for my project by the particular hesitance of many trans people to discuss intimate details of their bodies and sexual experiences -- understandable considering the morbid curiosity of strangers, acquaintances and the popular media, with which many trans people are so familiar.

The fourth and final factor seemingly beyond my control was the insistence of the research ethics committee of my college, whose approval I was required to obtain, on having signed consent forms from all participants. This seemed most curious to me, as I had read any number of studies by both undergraduate and professional researchers who had collected surveys over the Internet, without any use of the postal system, and without signed consent forms. Particularly given the financial limitations of my project, collecting surveys by mail did not seem to guarantee against the same degree of self-selection and non-representativeness that would result from online data collection. More to the point, I thought this an excessive roadblock to create in the name of ethics in the case of something so innocuous as a paper survey to be collected confidentially and used anonymously. Nevertheless, I deferred to the faculty who insisted that this was necessary, and this protocol created an extra hassle for all potential respondents, especially those who did have access to the Internet -- as all those who actually expressed interest in the project did -- and could otherwise have more quickly and without cost have received and returned the survey.

In retrospect, it is impossible to separate these factors beyond my control from what I can only characterize as the amateurish clumsiness with which I handled the entire undertaking. I began with what seemed to be the simplest means of promoting my project: posting announcements to Internet message boards and listservs. I also posted an announcement on the website of the International Foundation for Gender Education. I contacted a few print publications about advertising, and also wrote (by postal and electronic mail) to a number of support groups across the country, in the hope that whoever answered their mail might pass the information on to their members. Some of these points of contact I knew personally, including a number of online communities I participated in; most were more or less randomly selected by searching the Internet (with an eye to geographic diversity). I never managed to print advertisements in trans or queer community publications, in some cases due to lack of funds, and in others due to lack of any helpful response to my inquiries. I also did not gather any respondents from the support groups contacted, and I can only speculate that a) my haphazard selection had led to contacting many support groups which were no longer active (a few no longer had active email accounts); b) a large number of members of support groups contacted did not fit the study population I had defined, e.g. they were not pursuing hormone therapy, had not begun it or had not been on hormones for the specified six months; and/or c) support group coordinators did not pass on information about the project to members. This was very disappointing, as I had hoped to include by way of support groups individuals who did not have Internet access.

In the event, I did receive a few dozen responses from individuals interested in the project. Some I apologetically turned away because they resided outside North America or were not yet undergoing hormone therapy. Others I furnished with further information as requested about my academic background and the purposes of my project, and most confirmed their interest in participating; to these I mailed my introductory letter, consent form and the survey itself.

And here my troubles began. A number of packets were sent back for insufficient postage; most of these lingered for months in my PO Box because the Postal Service failed to forward them to me over the summer. Others reached their destinations, but weeks went by without any response. I had no orderly system for keeping track of all the emails in my project account: which individuals had been sent surveys, which had returned them. Compounding this problem, I had not given a firm deadline for returning completed surveys. Beginning to worry after three months of sending out surveys and getting back only a couple, I contacted all those who had been sent a packet and not returned the survey, thanking them again for their participation and gently urging them to return the surveys as soon as possible. It was only then that I discovered that some had not yet received their surveys! Though I scrambled to correct this situation, and kept sending out surveys now and again to people who inquired, the numbers were small and my practices remained haphazard: in some packets I included return postage, or cash, and in others not; I still anticipated more respondents than would permit such measures. I continued posting announcements on Internet groups, and late in the summer I created a webpage where interested individuals could read about the project and print out the consent letter and survey, saving the trouble of mailing it to them. I even urged respondents to email the completed survey to me and send only the consent form by postal mail, to save themselves a little money.

By the time I returned to school in the fall, I realized that my sample size was going to be much smaller than anticipated. Even though nearly all of the more than forty individuals who had expressed interest in participating had eventually received packets, only fifteen completed surveys were collected. It was at this time that I decided to include in the study a few surveys I had received from individuals who, it turned out, had not yet begun hormone therapy.

In retrospect, perhaps the most significant factor preventing me from collecting a large number of responses was my initial assumption that I would receive a large number of responses. My study population was defined narrowly in large part because I feared that otherwise my sample would be too large and diverse for a first-time researcher to manage. Even accounting for my sloppiness and lack of resources, defining the study population somewhat more broadly -- for example, as individuals who intended to undergo hormone therapy rather than only those who were no less than six months into therapy -- I might yet have collected the larger but still manageable number of responses I was aiming for, and the challenge of interpreting responses from this broader population would not have been too great. Had I been more organized in my data collection, given a firm deadline for returned surveys from the start, done more research and found a number of active and relevant support groups, I would have done better yet.

Back to Index