Carrotte et al.

BMC Medical Research Methodology (2016) 16:86

DOI 10.1186/s12874-016-0193-4

RESEARCH ARTICLE Open Access

“I am yet to encounter any survey that

actually reflects my life”: a qualitative study
of inclusivity in sexual health research
Elise R. Carrotte1*, Alyce M. Vella1, Anna L. Bowring1,2, Caitlin Douglass1,2, Margaret E. Hellard1,2
and Megan S. C. Lim1,2,3

Abstract
Background: Heteronormativity describes a set of norms and assumptions pertaining to heterosexual identities
and binary gender. In 2015, we conducted our annual Sex, Drugs and Rock’n’Roll study, an online health survey
of over 1000 Victorians aged 15–29 years. Feedback from participants suggested that our survey contained
heteronormative language. In response to this, we aimed to make inclusive changes to our survey via consultation
with young gender and sexually diverse (GSD) people.
Methods: We conducted two semi-structured focus groups in Melbourne with a total of 16 participants (age
range: 21–28 years). Participants were mostly cisgender women, and there were two transgender participants
and one non-binary participant. Participants also had a range of sexual identities including lesbian, queer,
bisexual, pansexual, and asexual. Focus group discussions were transcribed verbatim and analysed thematically.
Results: Most participants indicated heteronormativity affects their lives in multiple ways, noting its impacts on
access to sexual healthcare, invalidating sexual experiences and miscommunication in forms and surveys. Overall,
participants emphasised the need for sexual health research to avoid assumptions about behaviour, to be clear
and eliminate question ambiguity and avoiding treating gender as binary. Participants also discussed how the
Sex, Drugs and Rock’n’Roll survey could address a range of sexual behaviours and experiences, rather than
focusing on penetrative sex, which many participants found invalidating.
Conclusions: Our findings have important implications for future health surveys aimed at general populations.
We present recommendations that encourage research to be more inclusive to ensure data collection from GSD
participants is respectful and rigorous.

Background describes a set of societal assumptions and norms which

Gender and sexual diversity is an umbrella term relating are based on heterosexual, cisgender1 experiences, influ-
to gender expressions not matching male and female enced by social biases, privilege and stereotyping [1–3].
gender norms (including transgender and non-binary Heteronormativity can have negative impacts on the
identities), sexual identities, attractions and behaviours psychological well-being of GSD people, contributing
other than heterosexual, and intersex variations where to feelings of invisibility, invalidation and marginal-
reproductive or sexual anatomy do not fit typical male isation [1].
or female classifications. Although these identities are Examples of heteronormativity can be found in mul-
increasingly being recognised and celebrated, many gen- tiple settings. School-based sexual education is largely
der and sexually diverse (GSD) individuals experience heteronormative, often lacking representation and dis-
the effects of heteronormativity [1]. Heteronormativity cussion of GSD people and their experiences [2, 4].
Qualitative research suggests that heteronormativity in
* Correspondence: [email protected] sexual education legitimises homophobic bullying and
1
Centre for Population Health, Burnet Institute, 85 Commercial Road,
Melbourne, VIC 3004, Australia
contributes to the marginalisation of GSD people [4].
Full list of author information is available at the end of the article GSD people may also experience heteronormativity in
© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Carrotte et al. BMC Medical Research Methodology (2016) 16:86 Page 2 of 10

healthcare settings, making it difficult to access appro- research, and – if necessary to use categories – to in-
priate care. Some physicians make assumptions about clude multiple, itemised gender options and the ability
their patients’ gender, sexual identity and behaviour due to select multiple options [22]. Meanwhile, the Gender
to their lack of awareness or hesitancy to approach Identity in U.S. Surveillance group (GenIUSS; a collabor-
sensitive topics. However, GSD people have unique ation between researchers and GSD groups) and the San
health needs, influenced by both sexual orientation and Francisco Department of Public Health provide recom-
behaviour [5, 6]. For example, physicians may inaccur- mendations for population-based surveys; both recom-
ately assume their GSD patients are not at risk of mend a two-step approach (i.e., asking about sex
sexually transmissible infections (STIs) if they do not assigned at birth and also gender identity). This ap-
report penetrative sexual intercourse [7–12], even proach has high sensitivity and specificity with adult
though lesbians and bisexual women are more likely populations; however, there is no clear evidence that this
than heterosexual women to have ever been diagnosed approach is appropriate for surveys including adoles-
with an STI [13]. Inadequate healthcare access can ad- cents [23, 24]. OII (Organisation Intersex International)
versely impact the sexual health and mental wellbeing of Australia recommends intersex status be asked as a sep-
patients and negative experiences may contribute to- arate question; people with intersex variations do not
wards low continuity of care and satisfaction with health necessarily identify with the term ‘intersex’ in relation to
providers [8, 14–16]. gender and/or sex [25]. Regarding sexual orientation, the
Considering unique sexual health needs and barriers Sexual Minority Assessment Research Team (SMART)
to appropriate healthcare access, there is a need to con- published recommendations for general research, and
duct research into the sexual health of GSD people to recommend asking about the three major dimensions of
inform policy and practice. While studies specifically tar- sexual orientation (identity, behaviour and attraction)
geting GSD individuals exist, (e.g. [17, 18]) general separately if relevant to the research [26]; to contrast,
population studies should also include these communi- the LGBTI Alliance recommends asking about attraction
ties in order to accurately reflect the prevalence and be- or behaviour rather than identity for young people who
haviours of GSD individuals and allow methodologically may still be forming identities [19]. However, no known
sound comparisons between groups [19]. However, it is inclusive language recommendations exist that are spe-
common for researchers to make heteronormative as- cific to youth sexual health research, and no known pa-
sumptions about their participants’ gender identities and pers report details of consultations with young GSD
sexual orientation based on limited responses – for ex- people on these issues. Therefore, these recommenda-
ample, making assumptions about sexual identity based tions were not directly applicable to our research.
on behaviour or vice versa [20]. It is also common for In this study, we aimed to fill this gap in the context
researchers to conflate sex assigned at birth, gender of the Sex, Drugs and Rock’n’Roll (SDRR) survey, an an-
identity and genitals, leading to the miscategorisation or nual survey of Australians aged 15–29 years conducted
exclusion of participants with intersex variations, trans- by the Burnet Institute since 2005 [27]. From 2005 to
gender participants and participants with non-binary 2014, this general and sexual health survey was con-
identities [20]. Further, sexual health research commonly ducted at a music festival, where participants self-
relies on categorisation of sexual behaviours that focus completed a brief, paper-based questionnaire. The sur-
on penetration and assume the presence of a penis [20]. vey defined sex as ‘sexual intercourse’ or ‘penis in vagina
These problems are also common in other forms of writ- or penis in anus’ before questions on sexual behaviour, a
ten communication, such as healthcare intake forms, definition similar to that used in other sexual health re-
where GSD patients often struggle to accurately commu- search aimed at general populations. In 2015, the survey
nicate their gender and sexual information with limited was conducted online. Researchers received several
response categories, causing frustration and miscommu- negative comments about the language of our survey,
nication [21]. Sexual health surveys should be designed with comments about heteronormative language relating
to minimise heteronormativity and allow all participants to sexual health and sexual risk behaviour questions,
to accurately and respectfully answer questions. suggesting we failed to develop rapport with some par-
Although there are no universally accepted best ticipants. It is likely that participants felt more comfort-
practice recommendations, several organisations and able communicating these issues online rather than face-
researchers have made recommendations for inclusive to-face. Further, many female participants in our survey
research methodology. Researchers Ansara and Hegarty reported sex in the past year with female partners des-
[22] provide recommendations regarding gender diver- pite the survey’s definition of sex (involving a penis).
sity in the context of psychological research, recom- Many of these women ticked an option indicating no
mending use of blank text boxes for recording contraception was used the last time they had sex,
participants’ gender and only asking for sex if relevant to resulting in data which incorrectly assumed their STI
Carrotte et al. BMC Medical Research Methodology (2016) 16:86 Page 3 of 10

and pregnancy risks were comparable to unprotected, the language of the survey. Interviews were recorded
vaginal intercourse. Therefore, this study explored the and transcribed verbatim. Identifying information was
perspectives of young GSD people on issues of hetero- removed prior to analysis, and all participants were
normativity, sexual health and experiences with research, assigned pseudonyms.
with the aim of improving and making more relevant
the language used in the SDRR survey and similar Analysis strategy
research. Qualitative thematic analysis was performed on both
transcribed data and written notes from participants.
Methods Coding was undertaken by one researcher (EC). After
Recruitment strategy data immersion, open coding was undertaken to apply
Recruitment notices were posted on Burnet Institute’s headings to concepts represented in the text. [30] Codes
website and social media channels as well as the social were consolidated into higher-level themes concentrat-
media channels of a variety of organisations focusing on ing on the impacts of heteronormativity; themes were
gender and sexual diversity. Recruitment notices were refined and organised through an iterative process based
also emailed to relevant university student organisations. on repetition of topics, relationships between codes and
Inclusion criteria were being aged 18–29 years, with the relevance to the study aims. After the FGDs, the re-
invitation requesting ‘young women who identify as gay, searchers discussed and compared the groups’ suggestions
lesbian or bisexual (or otherwise report sexual interest to help inform decisions relating to survey changes. Deci-
in women), transgender people, and/or other young sions for survey changes were also informed by existing
members of GLBTIQ+ communities.’ Further informa- recommendations from community based organisations
tion was provided and consent forms distributed upon and researchers (e.g. [22, 23, 26]).
expression of interest. Participation in the earlier SDRR
survey was not required. Prior to the focus group, partic- Results
ipants reported their gender, sexual identity and pre- Participants
ferred pronouns to be used when reporting results and Sixteen participants (age range 21–28 years, mean age =
when personally addressed within the focus groups. Par- 24.9 years) attended the FGDs. Seven participants
ticipants were reimbursed for their time. attended the first FGD and nine attended the second.
Participants’ listed gender identities included cisgender
Focus group discussions woman (n = 12), cisgender man (n = 1), transgender
Two ninety-minute focus group discussions (FGDs) were woman (n = 1), transgender man (n = 1) and non-binary
held in Melbourne. FGDs have previously been effective (n = 1). Participants sexual identities included lesbian
for obtaining in-depth information with GSD people [1, (n = 6), queer (n = 2), bisexual (n = 2), pansexual/bisex-
28, 29]. Topics for FGDs were decided a priori; a discus- ual (n = 1), pansexual/queer (n = 1), queer/gay (n = 1),
sion guide was developed by the researchers. This guide gay (n = 1) and asexual (n = 1). One participant speci-
was influenced by literature about heteronormativity and fied she identified ‘with’ (not ‘as’) queer.
sexual health needs of GSD people, and the researchers’
experiences working with young people and minority Thematic analysis
groups on sensitive topics such as sexuality. The first FGD focused on issues relating to transgender
The FGDs were facilitated by one member of the re- people, lesbians and bisexual women. The second FGD
search team (EC) and two other members of the team also focused on lesbians and bisexual women, but ex-
assisted each group (CD, ML and AV). As informed by plored further about asexuality and the influence of
the discussion guide, the FGDs began with open discus- popular culture on assumptions and attitudes on GSD
sion about why the participants were interested in dis- people. Despite these minor differences, an overarching
cussing inclusive language in sexual health research. The theme of both FGDs was heteronormativity and its sig-
first half of the FGDs was dedicated to semi-structured nificant impacts on various aspects of participants’ lives.
discussion on key topics including gender, sex and sexu- The prevalence of heteronormativity (a major theme)
ality, and assumptions others make about these topics and its impact on three areas of interest (sub-themes)
and difficulties in written communication of these topics are discussed briefly below, followed by a summary of
including surveys and healthcare forms. In the second suggestions for improving the SDRR survey.
half of the FGDs, participants were divided into groups
of two to four and given printed copies of the SDRR sur- Heteronormativity is everywhere
vey’s questions related to gender and sexual behaviour Participants identified examples of heteronormativity in
(Table 1). Participants annotated the questions and pro- multiple settings, such as in healthcare, sexual education,
vided both written and verbal suggestions for improving the workplace and even in brief interactions with
 Carrotte et al. BMC Medical Research Methodology (2016) 16:86
Table 1 Changes to the Sex, Drugs and Rock’n’Roll (SDRR) survey
Topic 2015 questions Updated questions Description of changes
Gender What is your gender? What is your gender? (Please select all that apply) Participants now have more flexibility in
Options: Male, female, transgender, other [option to specify] Options: Female, male, transgender, non-binary/genderqueer, other [op- specifying their gender identity and can
tion to specify] choose multiple options that describe
them. Transgender participants can
specify their gender identity rather than
just noting they are transgender.
Intersex status Not asked Intersex is a term for people born with atypical physical sex Inclusion of intersex status.
characteristics. There are many different intersex traits or variations. Do
you have an intersex variation?
Options: Yes/No
Sexual identity How do you identify yourself? How do you currently identify yourself? (Please select all that apply) Question reworded to specify current
Options: Heterosexual (straight), bisexual, gay/homosexual/lesbian, Options: Heterosexual (straight), gay/homosexual/lesbian, bisexual, sexual identity. Participants able to
questioning, queer, other [option to specify] pansexual, asexual, queer, questioning, I don’t know/unsure, I don’t label choose multiple terms. Additional
myself, other [option to specify] options included.
Sexual How old were you when you first experienced the following? (drop Questions are asked in relation to the following behaviours: More options included for each sexual
behaviour down box with age options) Touching a partner’s genitals with your hands, being touched on your behaviour rather than just age of
Deep kissing, touching a partner’s genitals with your hands, being genitals by a partner’s hand, giving oral sex, receiving oral sex, vaginal first experience.
touched on your genitals by a partner’s hand, giving oral sex, receiving intercourse (penetration of vagina by penis), anal intercourse ‘Sex’ not defined before questions;
oral sex, vaginal intercourse (penis in vagina), anal sex (penis in anus) (penetration of anus by penis), rather, questions are worded more
In the following questions, ‘sex’ means vaginal and/or anal sex. The following questions are asked for each behaviour: specifically.
How many people have you had sex with in your lifetime? How old were you when you experienced this for the first time? Question asked in table format.
In the last 12 months, how many males have you had sex with? How many people have you done this with in the your lifetime? Participants only asked relevant questions
In the last 12 months, how many females have you had sex with? In the last 12 months, how many partners of each gender identity (e.g., if they have never experienced
(Categories for response ranging 0-51+) below have you done this with? (# male partners/# female partners/# vaginal intercourse, they will not be
other) asked subsequent questions
about vaginal intercourse).
STI risk In the last 12 months how often did you use a condom with [regular/ In the last 12 months, how often did you use a condom during fellatio ‘Sex’ not defined before set of questions;
behaviour casual/new] sex partner/s? (mouth to penis, ‘blow jobs’)? rather, questions are worded more
Categories for response: N/A no partner/s in past 12 months, always In the last 12 months, how often did you use a glove, dam or other specifically.
used a condom, usually (>50 %), sometimes (≤50 %), never used a barrier during cunnilingus (mouth to vulva or vagina, ‘going down’)? Questions asked for fellatio and
condom with partner/s The following questions are related to penetrative sex (i.e., penetration of a cunnilingus as well as penetrative sex.
vagina or anus with a penis). If participants report not engaging
In the last 12 months how often did you use a condom with [regular/ in the specified sexual behaviour
casual/new] sex partner/s during penetrative sex? in the last 12 months in earlier questions,
Categories for response: N/A No [behaviour/behaviour with relevant they will not be shown the related
partner/s] in last 12 months, always used a condom/barrier, usually questions for the behaviour.
(>50 %), sometimes (≤50 %), never used a condom/barrier [with
relevant partner/s]
Contraception The last time you had sex, which form(s) of contraception did you or The last time you had vaginal intercourse (penis in vagina), which Only asked if reporting vaginal
use at last the person you had sex with use? (tick all that apply) Options: Condom, form(s) of contraception did you or your partner use? (Please select all intercourse ever.
penetrative oral contraception (the pill), injection (Depo Provera), implant that apply) More options added.
sex (implanon), emergency/morning after pill, withdrawal/pulling out, none, Options: N/A (one of us was pregnant or trying to become pregnant),
other [option to specify] condom, oral contraception (the pill), injection (Depo Provera), implant
(implanon), intrauterine device (IUD), diaphragm, hormonal ring,
emergency/morning after pill, withdrawal/pulling out, other [option to

Page 4 of 10
specify], I don’t know, none of these
N.B. ‘I don’t wish to say’ is also an option for all questions in the both original and updated surveys
Carrotte et al. BMC Medical Research Methodology (2016) 16:86 Page 5 of 10

strangers. It was described as more frequent than homo- questioned about their sex lives by other people. Further,
phobia. Most participants noted assumptions by others most participants reported frustration and anger at
(typically heterosexual, cisgender individuals) about their school-based sexual education, which they generally de-
gender and sexual identities, their relationships, their scribed as unhelpfully focusing on heterosexual vaginal
bodies and their sexual experiences. Experiencing and intercourse and STI risk and failing to prepare them for
challenging heteronormativity was described as exhaust- their adult sex lives.
ing, invalidating, frustrating and a ‘battle’. The impacts When asked to define ‘sex’, participants struggled to
of heteronormativity in regards to three key concepts identify a specific definition. Participants indicated that
are described below. sex is diverse, personal, usually intimate, and may not
necessarily be reflective of traditional views, even in het-
Heteronormativity is a barrier to sexual healthcare erosexual encounters. Participants also discussed how
Participants described heteronormativity as a barrier to- experiences of sex and sexuality evolve with age and in-
wards sexual health, specifically sexual health-seeking dividual experiences, and focusing on penetrative sex
behaviour. Most participants reported sexual health- fails to capture a spectrum of behaviours, thoughts, atti-
related experiences and frustrations, including doctors tudes and experiences. While participants in both FGDs
ignoring or struggling to acknowledge sexual identities concluded that it was unhelpful to have a specific definition
and practices. Even when health professionals were of sex or define it on behalf of others, they also acknowl-
aware and accepting of participants’ gender and sexual edged that they wanted their individual understandings of
identities, stigmatising language or assumptions about what constitutes sex to be respected.
patients’ sexual experiences (such as assuming that bi-
sexual and pansexual participants led hypersexualised Heteronormativity facilitates text-based
lifestyles) were common and acted as a barrier to devel- miscommunication
oping rapport and receiving appropriate sexual health- In most participants’ experiences, surveys and forms de-
care. Some participants, particularly lesbian and bisexual signed for general populations rarely represent lifestyles
women, described dismissive or ignorant attitudes of of GSD people. Written communications, particularly
doctors towards sexual experiences and STI risk behav- those that rely on selecting a single category, pose a
iours if there is no risk of pregnancy. Transgender par- challenge for participants who do not necessarily fit into
ticipants acknowledged the difficulties of understanding provided categories.
and communicating their sexual health needs to doctors
who appeared uncomfortable or were unfamiliar with I am yet to encounter any survey on any topic at all
transgender experiences and bodies. that actually reflects my life, and the activities that I
do with my life. Ivan, 28, gay, transgender man
Heteronormativity invalidates sexual experiences
Participants described a heteronormative societal as- Participants reported specific challenges with com-
sumption that ‘sex’ is penis-in-vagina or penis-in-anus pleting forms with regards to gender and sex – in-
intercourse, and anything other than this is not ‘sex’. cluding having no options that describe them, not
This assumption was particularly distressing to partici- knowing how to ‘best’ respond to questions, and try-
pants who did not engage in penetrative intercourse ing to balance providing accurate information with in-
regularly or at all. formation that actually describes their experiences. It
was also noted that the common forced choice be-
I just think having to justify stuff… it’s just very tween ‘male’ or ‘female’ can be upsetting for trans-
frustrating… if it doesn’t tick these particular boxes, gender and non-binary people.
these one or two boxes, then that’s not ‘sex’. Lisa, 22, After being presented with the SDRR survey’s defin-
lesbian, cisgender woman ition of sex, participants were asked whether they would
answer questions regarding sexual behaviour in align-
It can make you feel like your sex doesn’t really ment with their own ideas and definitions or with the
matter. Abby, 23, pansexual/bisexual, cisgender survey’s definition. Most participants agreed they would
woman often, if not always, answer with their own understand-
My experience, my relationships are valid. They’re ing of sex in regards to sexual partners and risk behav-
valid to me. Laura, 23, lesbian, cisgender woman iours. Participants noted that the inclusion of such a
definition suggests that penetrative sex is the clinical
Participants described this assumption in many areas of and the only recognised definition of ‘sex’, indicating a
everyday life. For example, female participants in relation- lack of empathy. Some participants reported filling any
ships with other women described being inappropriately available free-text answer boxes with comments about
Carrotte et al. BMC Medical Research Methodology (2016) 16:86 Page 6 of 10

heteronormative language in order to inform researchers know, I am engaging in a social culture? What part of
of their predicament. my experiences with this do you wanna know about?
Leah, 23, identifies ‘with’ queer, cisgender woman.
I don’t go into a survey expecting that the options will
make me feel comfortable, or will be easy… when they
ask me about sexual behaviour, I will probably have Gender/sex
to, like, sit there for fifteen minutes, thinking, “What Participants questioned whether researchers were in-
answer should I put down?” Taylor, 21, queer, terested in the gender of the participants (a cultural
transgender woman concept), or the participants’ sex assigned at birth
(biological). It was noted that only being able to se-
Participants unanimously reported strong negative lect one option for gender was limiting, particularly
feelings towards the survey’s definition of sex, which re- as ‘transgender’ is an adjective and not a gender iden-
sulted in the experience of ‘othering’ and feeling ex- tity. Participants suggested either replacing categories
cluded. These feelings can influence the participants’ with a free text box so survey participants could de-
relationship with the organisation that has developed the scribe their gender however they desired, or having
survey, making it difficult to develop rapport. more diverse options. It was suggested to also have
‘cisgender’ as an option (accompanied by a definition
When [my idea of sex] isn’t reflected in a survey, or for those unfamiliar with the term), to recognise non-
not legitimised, it also feels like that survey and binary identities and intersex variations, and to spe-
whoever’s funding that survey – and a lot of the time cify male-to-female and female-to-male transgender if
that’s the government – doesn’t care about you, or using categories. Another possibility included allowing
your health, or your wellbeing. Ivan, 28, gay, selection of multiple options. Other questions in the
transgender man survey were criticised for treating gender as a binary,
particularly for reporting sexual partners. Wherever
Regardless of negative experiences with surveys, sev- terms like ‘boyfriend’ and ‘girlfriend’ were mentioned,
eral participants reported a willingness to participate in it was suggested we include ‘partner’ either addition-
sexual health research, ideally if it were made more in- ally or as a replacement.
clusive and empathetic. Participants emphasised that
there is a lack of data about GSD communities, particu-
larly transgender and non-binary communities, and they Sexual orientation
wish to give data accurately and ‘respectfully,’ and have It was suggested that the question on sexual identity
research be translated into positive change. (Table 1) should use the word ‘currently’ as sexuality is
fluid. Similar to the question on gender, participants sug-
I think it’s important just to raise it and say, “We’re gested allowing selection of multiple options. It was
here. We’re here and we’re queer.” [laughter] “Take noted as unrealistic to list every possible sexual identity
notice and research us!” Anna, 26, lesbian, cisgender but the most obvious omissions were asexual and pan-
woman sexual; other suggested options included: ‘Never consid-
ered it’, ‘I don’t know’ and ‘I don’t label myself’. Other
Summary of suggestions for improving our survey suggestions included randomising the order of the list
Participants readily agreed that the language of the (to eliminate a perceived hierarchy of sexual identities),
SDRR survey (Table 1) could be improved, but acknowl- modifying future questions based on the participants’ se-
edged that this was not a straightforward task, particu- lected sexual identity, and updating the list each year
larly as language evolves within GSD communities. with options listed in free-form text. Another suggestion
Participants emphasised the need to identify the true was including questions about attraction in place of, or
aim of each question in order to help eliminate ambigu- in addition to, questions about identity.
ity and clarify the desired response.
Sexual behaviour
“What are you actually asking of me?” I think is the The main criticisms of sexual behaviour questions were
thing. I mean, it’s really tough in research, because heteronormative language and dismissiveness of sexual
you don’t wanna manufacture answers, but… when behaviours other than penetrative sex. It was suggested
I’m filling out a survey like this that I know is that the definition of sex could be removed (with sexual
important… I wanna know what it actually is that you behaviour of interest specified for each question), or
wanna know from me. Do you wanna know if I’m at modified to be more inclusive. Participants suggested
blood transmission risk? Do you wanna know if, you lessening the focus on penetrative sex and including
Carrotte et al. BMC Medical Research Methodology (2016) 16:86 Page 7 of 10

sexual behaviours such as masturbation, pegging,2 and restructure of the sexual behaviour section (including
specifying receptive vs. insertive anal sex. addressing more sexual behaviours and related STI
Regarding condom questions, participants suggested transmission risks), and re-wording of several questions.
we restructure to determine the reason why a participant Inclusion and wording of questions is dependent on a
may select never using a condom, e.g., no chance of particular study’s research goals [33]. We believe our
pregnancy or low perceived STI transmission risk. Other changes are most appropriate at this time for the SDRR
suggestions were including a broader definition of sexual survey’s scope, location and audience – that is, sexual
behaviour but listing more barrier methods of contra- health research aimed at young, general populations.
ception, or including a penetrative definition but en- Our recommendations relating to gender are similar to
suring earlier questions are not dismissive of sexual those suggested by Ansara and Hegarty [22] if using cat-
behaviours. egorical response options. Our findings expand the exist-
ing literature by providing recommendations specific to
Changes to the survey inclusivity in sexual health research. Although previous
Our revised SDRR survey questions are summarised in research has discussed the impact of heteronormativity
Table 1; changes were based on the results of FGDs, dis- among GSD populations, (e.g. [1, 3]) this had not previ-
cussions within the research team, other surveys on ously been discussed in the context of research. Our
topics of gender, sex and sexual diversity (e.g. [17, 18]) findings also provide a detailed rationale on how hetero-
and recommendations from various sources (e.g. [22, 23, normative questionnaires can invalidate and frustrate
26]). These changes were made with consideration of GSD participants in the context of their lives, which are
the key challenges of survey-based sexual health re- already influenced by heteronormativity. These findings
search, including identifying accurate measures of be- also indicate that inclusive, respectful language and sur-
haviour, minimising participation bias (including social vey structure can positively influence rapport.
desirability) and comprehension problems, allowing par- Despite our changes, our survey is still unlikely to fully
ticipants a safe space to accurately report behaviours capture the complexities of identities and experiences of
that may be of a sensitive nature, and treating sexuality participants; we have maintained a largely category-
with nuance and respect [20, 31]. Although inclusivity based multiple-choice structure to ease data collection
regarding intersex status was briefly discussed in the for our expected large samples (n > 1000). It was within
FGDs, we did not have any participants who identified the scope of the research to focus on gender rather than
an intersex variation, and our decision to include sex assigned at birth; for other studies, particularly those
intersex status as a separate question was informed by focusing on adult populations, the two-step process may
recommendations from OII (Organisation Intersex be most appropriate [23, 24]. We have focused on iden-
International) Australia [25]. tities, which may not capture experiences and attraction,
likely leading to underreporting of some experiences
Discussion (e.g., some people with transgender life experiences do
This study reports on results of FGDs with young GSD not identify as ‘transgender’) [20, 23]. Of note, we have
people regarding inclusive sexual health research meth- added a significant number of questions to our survey,
odology. Similar to research into healthcare needs of increasing the risk of participant fatigue. Our survey is
young GSD people [32], participants reported negative aimed at a general population; it is likely that some ter-
experiences with heteronormativity, and an overwhelm- minology used in the survey may be confusing for some
ing desire to be treated with respect, competence and to participants. We aim to address this limitation by word-
have their gender and sexual identities recognised in a ing questions clearly [23]. We will review the content of
variety of formats. FGD participants confirmed prior the survey over time from a range of perspectives, and
feedback that some of the language of the SDRR survey modify survey language if deemed necessary.
was heteronormative, invalidating and could result in in- Revisions to the survey introduce some additional lim-
accurate data. Participants were passionate and eager to itations regarding coding and analysis. Importantly, as
have their concerns heard with regards to inclusive lan- our questions have been modified, we will no longer be
guage, and expressions of interest outnumbered the allo- able to directly compare data for some variables across
cated places in the FGDs. The main points raised in the earlier surveys. Although we have included more cat-
FGDs included rethinking the survey structure to ensure egories for gender and sexual identities, it is likely that
a range of sexual risk behaviours are captured accurately, prevalence estimates of some communities in our sam-
to avoid treating gender as binary, and to ensure ple will be small. Sample size may limit data analysis
questions asked in the survey are clear and inclusive. As strategies; for example, although we will ideally conduct
described in Table 1, our updated survey includes recog- analyses among a diverse range of gender identities
nition of more gender and sexual identities, a significant rather than excluding ‘outliers’ [22], any analyses
Carrotte et al. BMC Medical Research Methodology (2016) 16:86 Page 8 of 10

performed on these groups are likely to have relatively recruiting from GSD events or organisations, or use of
large margins of error and we may be unable to make strategies to maximise participation of minority groups,
meaningful conclusions about behaviour [31]. Initially, e.g., time-location sampling or respondent-driven sam-
we may be required to collapse categories during data pling [33]. When deciding on the best approach to inclu-
analyses (although prevalence can still be reported in sivity, future studies will need to consider balancing
text); this has occurred in prior research [17]. Depending issues such as budget, space limitations, data accuracy,
on response rates, we will need to consider using target population demographics, and coding strategies.
appropriate coding schemes which do not code gender
identities as mutually exclusive or treat sexual minorities Conclusions
as a homogenous group [20, 22]. It is possible that over Based on our results, and with consideration of other
time, we can combine data to increase the sample size of methodology recommendations for general health re-
minority groups and perform meaningful analyses [23, search (e.g. [22, 23, 26]) and healthcare of GSD people
33]. Although not within the scope of our research, in general [15, 34], we put forward recommendations for
other sexual health surveys should consider over- inclusive language in sexual health surveys involving
sampling minority populations, such as specifically general populations of young people (Table 2). Some

Table 2 Recommendations for inclusive sexual health research involving general populations
Topic Recommendations
General Eliminate question ambiguity by identifying the true aim of each question and the type of data you wish to collect. To achieve this,
ensure questions are specific, clear and use defined time periods
Avoid language that assumes participants are heterosexual or cisgender, or that certain behaviours are ‘normal’
Use language that is culturally appropriate
Avoid forced-choice, single option items and binary categorisations where possible
Avoid skip patterns based on participants’ recorded gender or sexual identity; these require assumptions about participants’
attraction, identity, behaviour and genitals
Routinely update questions based on cultural changes, feedback and relevant research while still considering data comparability across time
Consult with GSD people and organisations if unsure of best wording of questions – avoid making assumptions
Check relevancy and appropriateness of questions, particularly if space or your budget is limited, and aim to balance these
considerations with making language as inclusive as possible
Be aware of your own biases and knowledge limitations as a researcher
Carefully consider all recommendations within the context of the research (including location and audience)
Gender/sex Determine whether your research is interested in sex assigned at birth, gender identity, neither, or both
If interested in sex and gender, consider a two-step process, i.e., asking for sex and gender separately
When asking for participants’ gender identities, use open-ended or free text options if possible. If free-text is not an option, ensure
options are more diverse than ‘female’ and ‘male’. Include recognition of transgender participants, non-binary/genderqueer partici-
pants, and other gender identities. Allow selection of multiple options or be specific with wording
Ensure gender/sex questions are optional
Avoid treating gender as binary (male/female) throughout the survey, including avoiding language implying a binary such as ‘both
genders’ or ‘opposite sex’
Avoid conflating gender and sex and be consistent with terminology
Ask about intersex variations separately from sex and gender
Sexual Consider sexual orientation in terms of identity, attraction and behaviour, if relevant
orientation
When asking about sexual identity, use open-ended or free text options where possible. If free-text is not an option, avoid ‘othering’
participants by including a range of sexual identities and options such as ‘I don’t know’ or ‘I don’t label myself’.
Consider selection of multiple identities
Sexual Avoid defining ‘sex’; instead, be specific to the sexual behaviour of interest
behaviour
Clearly identify when questions are referring to STI transmission risk, risk of pregnancy, or other objectives
Consider appropriate skip patterns for questions that may be irrelevant to participants based on previous response patterns (this
will also help reduce participant fatigue)
Consider respectful and appropriate inclusion of sexual behaviours beyond penetrative sex
Carrotte et al. BMC Medical Research Methodology (2016) 16:86 Page 9 of 10

recommendations, such as those related to gender and Ethics approval and consent to participate
sexual orientation questions, are also relevant for more Approval was granted by the Alfred Hospital Human Research Ethics
Committee (Project 163/15) and all participants signed informed consent
general research. forms before interviews commenced.
We believe that with careful consideration of the
points raised in this project, investigators can make re- Author details
1
Centre for Population Health, Burnet Institute, 85 Commercial Road,
search significantly more inclusive, whilst maintaining Melbourne, VIC 3004, Australia. 2School of Public Health and Preventive
rigour. It is unlikely that researchers will ever be able to Medicine, Monash University, Alfred Hospital, Commercial Road, Melbourne,
be inclusive of every individual participating in their re- VIC 3004, Australia. 3Melbourne School of Population and Global Health, The
University of Melbourne, Level 4, 207 Bouverie Street, Melbourne, VIC 3010,
search; however, routinely recognising diversity in future Australia.
sexual health research has several benefits. These
changes have the potential to improve rapport with GSD Received: 19 September 2015 Accepted: 22 July 2016

participants. Further, these changes will allow for the

collection of more accurate data which may inform
References
evidence-based healthcare and health interventions. 1. Platt LF, Lenzen AL. Sexual orientation microaggressions and the experience
of sexual minorities. J Homosex. 2013;60(7):1011–34.
2. Riggs DW. Heteronormativity in online information about Sex: a South
Endnotes Australian case study. Contemporary Issues Early Childhood. 2013;14(1):72–80.
1
Gender identity matches sex assigned at birth doi:10.2304/ciec.2014.14.1.72.
2 3. Sue DW. Microaggressions in Everyday Life: Race, Gender, and Sexual
A woman penetrating a man anally with a sex toy Orientation. USA: Wiley; 2010.
4. McNeill T. Sex education and the promotion of heteronormativity.
Sexualities. 2013;16(7):826–46. doi:10.1177/1363460713497216.
Additional file 5. Everett BG. Sexual orientation disparities in sexually transmitted infections:
examining the intersection between sexual identity and sexual behavior.
Additional file 1: 2015 Sex, Drugs and Rock'n'Roll survey. (DOCX 44 kb) Arch Sex Behav. 2013;42(2):225–36. doi:10.1007/s10508-012-9902-1.
6. Kerker BD, Mostashari F, Thorpe L. Health care access and utilization among
women who have sex with women: sexual behavior and identity. J Urban
Abbreviations Health. 2006;83(5):970–9.
FGD, focus group discussion; GLBTIQ+, gay, lesbian, bisexual, transgender, 7. Bjorkman M, Malterud K. Lesbian women’s experiences with health care: a
intersex, queer or other identities; GSD, gender and sexually diverse; STI, qualitative study. Scand J Prim Health Care. 2009;27(4):238–43.
sexually transmissible infections 8. McCann E, Sharek D. Challenges to and opportunities for improving mental
health services for lesbian, gay, bisexual, and transgender people in Ireland:
a narrative account. Int J Ment Health Nurs. 2014;23(6):525–33. doi:10.1111/
Acknowledgements inm.12081.
The authors wish to thank the focus group participants for their valuable 9. Mravcak SA. Primary care for lesbians and bisexual women. Americal Family
contribution to this study. Physician. 2006;74(2):279–86.
10. Hayman B, Wilkes L, Halcomb EJ, Jackson D. Marginalised mothers: lesbian
Funding women negotiating heteronormative healthcare services. Contemp Nurse.
The authors gratefully acknowledge the contribution to this work of the 2013;44(1):120–7.
Victorian Operational Infrastructure Support Program received by Burnet 11. Rondahl G, Innala S, Carlsson M. Heterosexual assumptions in verbal and
Institute. ML is supported by an Australian Government Department of non-verbal communication in nursing. J Adv Nurs. 2006;56(4):373–81.
Health Preventive Research Fellowship. 12. Douglas C, Deacon R, Mooney-Somers J. Pap smear rates among Australian
community-attached lesbian and bisexual women: some good news but
disparities persist. Sex Health. 2015. doi:10.1071/sh14210.
Availability of data and materials
13. Grulich AE, de Visser RO, Smith AM, Rissel CE, Richters J. Sex in Australia:
The authors can provide de-identified transcripts of the two focus groups
sexually transmissible infection and blood-borne virus history in a
from this study. However, as the audio recordings were not de-identified, the
representative sample of adults. Aust N Z J Public Health. 2003;27(2):234–41.
authors cannot provide these recordings due to a need to protect partici-
14. McNair R, Szalacha LA, Hughes TL. Health status, health service use, and
pant identity. The 2015 Sex, Drugs and Rock’n’Roll questionnaire is available as
satisfaction according to sexual identity of young Australian women.
Additional file 1.
Women’s Health Issues. 2011;21(1):40–7. doi:10.1016/j.whi.2010.08.002.
15. Pitts MK, Couch MA, Smith AM. Men who have sex with men (MSM): how
Authors’ contributions much to assume and what to ask? Med J Aust. 2006;185(8):450–2.
The study was conceptualised and designed by EC and ML. EC facilitated 16. Snelgrove J, Jasudavisius A, Rowe B, Head E, Bauer G. “Completely out-at-
the focus groups, recruited participants, completed the literature review, sea” with “two-gender medicine”: a qualitative analysis of physician-side
transcribed interviews, analysed themes and led the writing of the barriers to providing healthcare for transgender patients. BMC Health Serv
manuscript. AV, CD and ML assisted with focus groups. AB assisted with Res. 2012;12(1):1–13. doi:10.1186/1472-6963-12-110.
thematic analysis. All authors have been involved with the design and 17. Hillier L, Jones T, Monagle M, Overton N, Ghan L, Blackman J, et al. Writing
implementation of the Sex, Drugs and Rock’n’Roll survey and discussed and themselves in 3 - The third national study on the sexual health and
agreed upon changes to the survey. All authors read, contributed to and wellbeing of same sex attracted and gender questioning young people.
approved the final manuscript. Melbourne: Australian Research Centre in Sex, Health & Society, Latrobe
University; 2010.
18. Lucke J, Loxton D, Lee C, Dobson A, Fraser I. Contraceptive use, Pregnancy
Competing interests
Intention & Decisions of Australian Women. New South Wales and Bayer
The authors declare that they have no competing interests.
Australia: The University of Queensland and the University of Newcastle in
partnership with Family Planning; 2013.
Consent for publication 19. Irlam CB. LGBTI Data: Developing an Evidence-Informed Environment for
Not applicable. LGBTI Health Policy. Melbourne: National LGBTI Health Alliance; 2015.
Carrotte et al. BMC Medical Research Methodology (2016) 16:86 Page 10 of 10

20. Ansara YG. Improving research methodology in adolescent sexual health

research. J Adolesc Health. 2015;56(4):367–9. doi:10.1016/j.jadohealth.2015.01.013.
21. Goins ES, Pye D. Check the box that best describes you: reflexively
managing theory and praxis in LGBTQ health communication research.
Health Commun. 2013;28(4):397–407.
22. Ansara YG, Hegarty P. Methodologies of misgendering: recommendations
for reducing cisgenderism in psychological research. Feminism Psychol.
2014;24(2):259–70. doi:10.1177/0959353514526217.
23. Herman JL. Best practices for asking questions to identify transgender and
other gender minority respondents on population-based surveys. Los
Angeles: The GeniUSS Group; 2014.
24. Department of Public Health. Principles for Collecting, Coding, and
Reporting Identity Data - Sex and Gender Guidelines. San Francisco: City
and County of San Franscisco; 2013.
25. OII Australia - Intersex Australia. Including intersex in research studies and
surveys. 2012. https://oii.org.au/20042/on-requests-for-research/. Accessed
31 Aug 2015.
26. Badgett MVL. Best Practices for Asking Questions About Sexual Orientation
on Surveys. Los Angeles: The Williams Institute; 2009.
27. Lim MS, Bowring AL, Gold J, Aitken CK, Hellard ME. Trends in sexual
behavior, testing, and knowledge in young people; 2006–2011. Sex Transm
Dis. 2012;39(11):831–4. doi:10.1097/OLQ.0b013e3182663f27.
28. Logie CH, James L, Tharao W, Loutfy MR. “We don’t exist”: a qualitative study of
marginalization experienced by HIV-positive lesbian, bisexual, queer and
transgender women in Toronto, Canada. J Int AIDS Soc. 2012;15(2):17392.
29. Marrazzo JM, Coffey P, Bingham A. Sexual practices, risk perception and
knowledge of sexually transmitted disease risk among lesbian and bisexual
women. Perspect Sex Reprod Health. 2005;37(1):6–12.
30. Miles MB, Huberman AM. Qualitative Data Analysis: An Expanded
Sourcebook. California, USA: SAGE Publications; 1994.
31. Fenton KA, Johnson AM, McManus S, Erens B. Measuring sexual behaviour:
methodological challenges in survey research. Sex Transm Infect. 2001;77(2):
84–92. doi:10.1136/sti.77.2.84.
32. Hoffman ND, Freeman K, Swann S. Healthcare preferences of lesbian, gay,
bisexual, transgender and questioning youth. J Adolesc Health. 2009;45(3):
222–9. doi:10.1016/j.jadohealth.2009.01.009.
33. Institute of Medicine (US) Committee on Lesbian Gay Bisexual and
Transgender Health. Conducting Research on the Health Status of LGBT
Populations. The Health of Lesbian, Gay, Bisexual, and Transgender People:
Building a Foundation for Better Understanding. Washington: National
Academies Press (US); 2011.
34. Roberts SJ. Health care recommendations for lesbian women. J Obstet
Gynecol Neonatal Nurs. 2006;35(5):583–91.

Submit your next manuscript to BioMed Central

and we will help you at every step:
• We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research

Submit your manuscript at

www.biomedcentral.com/submit