In the last four decades the medical student population in the UK has diversified in terms of gender and ethnicity [1–3] but less so in relation to social class[4]. Since 1965 the proportion of women entering UK medical schools has risen from 20% to almost 55% [5]. It is predicted that the proportion will plateau at 60%–65%, similar to many other caring professions [5], even rising to 70% at some universities [6]. In a parallel development, in some medical schools nearly 40% of the student intake is from ethnic minorities [3]. White men now comprise just 26% of all UK medical students [1]. While Asians comprise 7% of the UK population, they accounted for 19% of medical students in 2001[1] although Black Caribbeans remain under-represented [1, 7]. In London the proportion of Asian students is particularly high [8]. By 2002, medical school graduates were made up of 72% 'whites' and 28% 'non-whites' [1].

At the same time, attempts to further widen the participation of students from diverse social and ethnic backgrounds are being introduced[9]. Even so, some studies point to the existence of gendered and restricted circumscribed areas of practice, including a gendered and ethnic 'glass ceiling' (which the Oxford English Dictionary defines as 'an unofficial or unacknowledged barrier to personal advancement, especially of a woman or a member of an ethnic minority in employment') in the medical profession [4, 10–15]. Other key themes have been identified in this domain of research. In relation to gender: harassment, [16–19] the lack of female role models, [15, 19–21], gender stereotyping [14, 15] and less access to patients during obstetrics and gynaecology attachments for male medical students and for female medical students during surgery clerkships [22, 23] have been highlighted. In relation to ethnicity: poorer performances in some ethnic groups than in 'white' students in clinical examinations, despite good written examination results has been found [24] as well as differences in communication style during objective structured clinical examinations [25].

Against this background we have conducted a qualitative study in one medical school in the UK to examine students' experiences and perceptions in relation to gender and ethnicity. We have extended our earlier work on the hidden curriculum of the medical school [26]. This has been defined as 'the processes, pressures and constraints which fall outside ... the formal curriculum, and which are often unarticulated or unexplored' [27].

Qualitative data were collected in one-to-one semi-structured interviews (by HL), and conducted in a private room in the medical school (see Appendix 1), and each interview took an average of 50 minutes. Following the transcription of the interviews, we conducted a qualitative thematic analysis [28] with simple counting[29], supported by the Nvivo and Concordance software programmes. Themes were identified by reading and discussing the data and formulating a coding scheme that was used to assign passages of text to the themes. In our report for this paper we describe each 'theme' (for example, 'interest in people' as a reason for studying medicine), state how many interviewees expressed it (e.g. 28/36) and, for more important themes, provide an illustrative quotation. Our counting procedures are designed to give a rough idea of the prevalence of the themes in the data. They are not intended to provide a basis for statistical generalisation to a population.

The interview schedule (see Appendix 1) contained a number of key questions in relation to women in medicine, all of which were asked of each student. In reference to ethnicity, no specific questions were included, but relevant information provided by students voluntarily from all parts of the interviews was used in the analysis. We sought to improve validity by close examination of the plausibility of the accounts in the experience of the authors, asking for further clarification and examples of key points during the interviews, and by searching and accounting for negative instances in the data analysis [30].

Each medical student provided written informed consent prior to their study participation. Formal permission for conducting the interviews was obtained from the education committee of the medical school where the study was undertaken. The full requirements for the ethical conduct of the research, as set out by the British Sociological Association, were strictly complied with.

As this study is based on the observations and perceptions of medical students in a single medical school in the UK, the extent to which the findings are generalisable to other medical schools remain to be established. Elsewhere [29] we have reviewed various approaches to generalising from case studies such as this. The qualitative aspect of our accounts is designed to produce a degree of contextual detail that will help readers assess the degree to which our findings are transferable to other settings. We recognise though, that empirical generalisability can only firmly be established by doing similar research in other medical schools.

In this medical school, at the point of entry, there were clear gender differences in students' reasons for this career choice, with men more interested in 'external' factors related to the expectations of others for example, and women in 'internal' issues such as an interest in finding out more about the functioning of the body. The influences of mainly male 'informants' on career decisions reflects the reality of a profession still dominated by males in most senior positions.

'Non-white' students' difficulties in achieving independence from parents reflect cultural differences in family relationships. Some 'non-white' students identified a discrepancy between the ethnic composition of the medical student intake and the ethnic composition of the medical profession, suggesting that this could have negative implications for their own career prospects, as alluded to in previous publications. [10, 11] The experiences of black (African) female students may be different again, and this requires further research with a targeted sample of this minority population in medical schools. Accounts by female Muslim students implied that religious beliefs were not always acknowledged by teachers. The ethnic diversity of medical students was welcomed by some students among all ethnic groups.

When asked in general terms about gender differences in their medical training, most female and male students claimed that these did not exist, which supports the finding of Gjerberg who showed similar proportions of male and female students intended to enter surgery [15]. Nevertheless the data analysis in relation to more specific aspects of their training did reveal several ways in which such gender effects were identified by the students. Self-contradiction of this sort is quite normal in qualitative interviews (as it is in everyday life) and is worth noting in itself as a finding. Interestingly, having denied that such gender differences exist, when male and female students did disclose examples of such gender-related distinctions in their accounts, they did so in terms that were to a large extent consistent with traditional gender stereotypes, as has been shown in previous reports [12–16].

Almost a third of students (more females than males) insisted that the future of women in medicine, but not for men, consists of a harsh choice between home responsibilities and career. This does not bode well for the future family life of these doctors. Finally, the perceived reluctance by female patients to be examined by or in the presence of male medical students during gynaecology outpatient clinics, identified by a fifth of participants, has been confirmed by other reports [31].

Many of the issues raised by these students are somewhat sensitive and may be especially difficult to challenge. The key themes identified in this paper in relation to ethnicity and to gender have important implications for medical educators and for those concerned with professional development [16, 32, 33]. The results suggest a need to open up these relatively covert aspects of student culture to scrutiny and debate. This will mean taking an explicitly wider view of the influence of what has sometimes been called the hidden curriculum [26, 27] upon the training of medical professionals, and upon the practice of medicine [34, 35].