Let’s talk about sex: college students’ attitudes towards Sexuality Education in a Spanish University

Latent class analysis

Latent Class Analysis (LCA) is an extremely useful statistical tool for differentiating between various groups of people who follow the same behavioural pattern in a given situation. Paul Lazarsfeld first described this type of statistical analysis in 1950. At the beginning, it was used only for dichotomous variables, but in light of its possibilities from 1974 on its use became increasingly popular [14]. LCA has numerous applications in social science research due to its usefulness when classifying groups of individuals by attitudes, perceptions, or opinions, since it allows new groups of individuals to be established based on less obvious characteristics [15]. Interest in this statistical tool is increasing thanks to the availability of software such as Mplus [16].

As a model of multivariate analysis, LCA studies different variables at the same time. It allows the definition of a series of latent classes or profiles, represented by their response patterns to different variables. Thus, new underlying (latent) variables can be discovered and studied, though not directly observable they may influence the relationships between the variables being studied. By means of this analysis, new groups of individuals (latent classes or profiles) are defined and the probability that each observation, or variable, belongs to one or the other can be calculated [17]. Thus, LCA focuses on grouping different individuals, in this case university students, unlike other statistical techniques such as factor analysis, which focuses on associating different variables [15].

Study design

This is a cross-curricular, descriptive study, with an observational design. It was carried out on University of Valladolid (UVa) students during the 2020-21 academic year.

The UVa has sites in 4 Spanish cities: Valladolid, Palencia, Segovia, and Soria. During this year a total of 18,556 students were enrolled. To achieve a 99% confidence interval, a sample of 641 students was required [18].

In order to attract the largest possible number of participants, a user-friendly questionnaire was drawn up that was adapted to the diversity at UVa and the language of its students. It was reviewed by a group of 10 experts in sexual education and 11 university students, who provided suggestions for preparing the final version. It was distributed primarily through social media and was available online from December 2020 to April 2021.

Participants and ethical considerations

Participation was voluntary and anonymous. The people included in the study were students enrolled in one of the 133 degree courses offered by UVa for the 2020-21 academic year, adults, who expressly accepted written informed consent and who filled out the questionnaire in its entirety.

This research was approved by the Ethics Commission of Eastern Health Area of Valladolid (CEIm Valladolid), code PI 20-1833.

Data analysis

The questionnaire consisted of 2 parts, a first section that explores sociodemographic data such as age, type of studies, years at university, gender, sexual orientation, religious affiliation, and type of family and housing; and a second part that gathers opinions and attitudes towards Sexuality composed of two Likert scales: 8 opinions and 7 myths (Table 1), healthier attitudes would correspond to showing greater agreement with the opinions and greater disagreement with the myths. Two scales revealed acceptable internal consistency: result of Chronbach test were 0.752 and 0.788 respectively).

Statistical analyses were performed using MPLUS software [16]. Several latent class analyses were carried out to characterise different student profiles in accordance with their opinions and attitudes towards sexuality, and the Chi-squared technique was used to calculate the influence of sociodemographic factors on the composition of the profiles. Statistical significance was set at p<0.05. In this analysis, 8 no-woman no-man people was excluded in order to adjust Chi-squared test norms.

Sociodemographic characteristics

The final sample was made up of 1,028 UVa students.

The sample participants were mostly undergraduate students (93%), with an average age of 21 (a range between 18 and 49 years-of-age). 22.9% of the sample were in their first year of university, 67% were women and there were 8 participants who did not define themselves as either man or woman. In terms of sexual orientation, the majority defined themselves as heterosexual (80%), 12% were bisexual and 3.4% were homosexual. A predominance of Christian university students (51%) and those who have not yet become independent (53%) was observed.

Attitudes towards sexuality

In the first part, Opinions, no participant disagreed with “talking about sex”. 90% of the university students showed a favourable opinion to talking about the topic, masturbation, and other forms of sexual relations. 43.7% agreed with the consumption of pornography, and 13.3% with polyamory (62.2% were indifferent in this section). Furthermore, 56.8% were in favour of sexual fantasies, 13.2% of the consumption of drugs to improve sexual experience and 27.3% were in favour of sexual relations between the elderly (Fig 1).

• Download figure
• Open in new tab

On the second scale, myths, the percentage of students who chose to answer ‘don’t know’ or ‘indifferent’ increased in all sections, being the options preferred by between 46.4% and 70% of the students. 23% agreed that good ideas can be taken from porn, 30% that condoms reduce sensitivity in sexual relations and only 7% that jealousy expresses true love. The percentages of disagreement with or disbelieving the proposed myths were from 20% to 37% (Fig 2).

• Download figure
• Open in new tab

Latent class analysis

After making the appropriate adjustments, 3 student profiles were obtained from Opinions (Table 2) and 5, finally reduced to 4, from the analysis of myths about sexuality (Table 3). The variables that most influenced the composition of the different profiles were gender, sexual orientation and type, and religious orientation.

In the case of Opinions on Sexuality, three student profiles were identified: unfavourable/indifferent (17%), intermediate (68%) and favourable (15%). The sociodemographic factors that were significant in the composition of the profiles were gender (p 0.026), sexual orientation (p < 0.001) and religious situation and orientation (p < 0.001 in both cases) (Table 4). A higher probability of women, heterosexuals, and people with religious feelings of a Christian orientation was observed in the unfavourable profile, with a greater proportion of men and non-heterosexual and agnostic/atheist students among the components of the favourable profile.

Attitudes towards sexuality

Initially, showing healthier attitudes would mean agreeing with the opinions and rejecting the myths about sexuality. The results analysed show a large percentage of students agreeing with the opinions, in many cases exceeding 50%, while the number of students against myths about sexuality drops to only 20-30% of our sample. These results are in line with those observed by Beaumont in a study on SE in the European Union, where it is confirmed that a large part of young Spaniards’ attitudes towards SE are conditioned by stereotypes, myths, and erroneous beliefs [28].

What is striking is the large percentage of students in our sample who chose not to answer or appear indifferent in this section on attitudes; up to a third in the part concerning opinions and almost half on some of the proposed myths. If we add together the students who choose these two options, between Don’t Know and Indifferent, we find that around half of the students are not engaged with or do not have a clear opinion on the different aspects of sexuality raised in the study. The reasons for this lack of attitude were not analysed, although it could be attributed to passivity, ignorance, or to the influence of the confinement and general despondency caused by the Covid-19 pandemic.

The general predisposition to talk about sex is worthy of mention. This finding shows how widespread the presence of sex, sexuality or sexual relations is, whether formally or informally, in our sample. This fact was studied a few years ago in the United Kingdom, when it was observed that female students talked about sex about 13 times a day [29].

One of the first tools developed to understand behaviours, attitudes, and knowledge regarding sexuality among university students was the SKAT [30]. This questionnaire was implemented during the 1970s and revealed that up to three-quarters of students believed masturbation to be healthy, but that 16% of a sample of future doctors believed that masturbation could cause mental illness. A few years after this research, 90% of UVa students in our sample were in favour of masturbation.

Only 12% of our sample was against the consumption of pornography and 20% against taking drugs during sexual relations. Some studies have already indicated the large percentage of students who have had regular contact with porn and substance use [31,32].

In the last section, Myths, from 7% to 30% of the students accepted the proposals. The most commonly debunked myths were “jealousy represents true love” and “sex is more important for men than for women”, while the most enduring ones turned out to be “condoms reduce the sensitivity of sexual relations” and “good ideas come from porn.”

One of the myths debunked was that condoms reduce sensitivity in sexual relations, with which a third of our sample agreed. This could be wielded as an excuse not to use this method of contraception and protection against STIs in a sexual relationship. However, it was not among the reasons registered in research carried out with a sample of university students, when they were asked about reasons for not using a condom [33]. The responses were that the students were using other forms of contraception, participants had got carried away by the moment, the condom had been forgotten due to the influence of drugs or alcohol, a condom was not available at the time, and other reasons, the least frequent being that they are looking to procreate or that the relationship was between people of the same sex and there was no risk of pregnancy.

20% of the UVa students agreed with the controversial phrase that “all men should be circumcised”, incidentally, since 2007 the WHO has recommended voluntary circumcision in certain African countries as a part of an HIV prevention strategy [34], these being countries with high prevalence of HIV and difficult access to health programs. Additionally, there is the justification that some religions claim that the procedure is necessary, even if there are no medical reasons for performing it. The arguments in favour of the need for this surgery would not affect our sample, since they are neither African students nor do they practice any of the religions that are in favour of circumcision.

Latent class analysis

The subject studied by our sample did not influence the characterisation of the profiles calculated by means of the attitudes shown towards sexuality. However the opposite was observed in a study on prejudices and myths in sexuality, here students of Engineering and degrees related to Earth Sciences corresponded proportionally more to a profile that accepted stereotypes of diversity; the intermediate class was made up of more students taking health subjects and Social and Biological Sciences; while those who specialized in Humanities and Philology had a greater probability of being included in the profile with more positive and healthier attitudes [19].

Gender greatly influenced the characterisation of the profiles observed in Attitudes towards Sexuality, as in the study by Yu, whose LCA confirmed a greater presence of men in the class that has a higher probability of showing more unfavourable opinions or not getting beyond stereotypes [35]. A Hungarian study analysed the influence of gender, sexual orientation and religious situation on the attitudes shown, describing two student profiles, conservative and adventurous, with a higher percentage of women among the former and a greater probability that those seeking new experiences were men [36].

As described above, the LCA of Affirmations about sexuality revealed 3 student profiles regarding their opinion on the proposed affirmations: unfavourable/no engagement, intermediate and favourable.

What is striking about this analysis is the number of students that make up Class 2, more than half of the sample, which may imply a reduction in the power of the analyses or the existence of a non-response bias [37], but taking the results as a whole, what the appearance of this profile suggests is that a large percentage of students have not formed an opinion or are indifferent towards SE [35]. The causes and consequences of this lack of engagement should be addressed in greater depth in the future since there is currently hardly any research on the matter.

In Class 1, with more unfavourable opinions towards sexuality, a greater probability was observed of finding women, heterosexual students, and Christians. The healthier Class 3 was made up of a higher percentage of men, of people of non-heterosexual orientation and those without religious feelings (agnostics or atheists), which concurs with what some studies have observed [20,39,36,40] and is contrary to what has been found in research with Spanish university students, where gender did not influence the attitudes shown towards sexuality [41] or if it did, women were the ones who showed healthier opinions [23].

Regarding myths in sexuality, it has been observed that, on many occasions, students are not aware of the prejudices they show towards topics in the field of SE [19]. After the analyses of the rejection or acceptance of the proposed myths, five student profiles were observed, which were finally deduced to four. Class 1 was indifferent, Class 2 agreed with sexuality myths, Classes 3 and 4 were intermediate, and Class 5 showed a greater likelihood of disbelieving the proposed myths.

In our sample, it was observed that Class 1, characterised by a tendency towards indifference, was made up of a higher percentage of women and students without religious sentiment. In some research such as that of Evcili [37], it is women who are more likely to show greater prejudices towards Sexuality.

As for the members of Class 2, tolerant of the myths and in agreement with them, they are a minority. The characteristics of this profile, completely in favour of consuming pornography, are in line with the positive relationship observed between the acceptance of porn, myths, and stereotypes concerning sexuality [42]. This class was made up of a greater proportion of men, homosexuals, and those of a Christian orientation. These findings are in line with studies confirming that religious university students are the most likely to accept myths about sexuality [42].

Students in Class 5, healthier since they show greater discordance with the proposed stereotypes, make up a fifth of our sample. These results are quite promising, contrary to what some reports on SE in Europe state [28], where it is argued that the attitudes of Spanish students are conditioned by myths and erroneous beliefs. The university students who made up this group were, with higher probability, students of bisexual orientation and a large percentage of women compared to men [23].

Finally, the intermediate classes, 3 and 4, intermediate in terms of accepting some myths and rejecting others, were made up of a higher percentage of students of heterosexual orientation. The difference between both classes lies in religion; in Class 3 there is a greater probability of non-religious students and in Class 4, there is a greater proportion of students who follow some religion.

The objective of SE programs would be to get the members of class 1 engaged with something that, ultimately, is a part of their life, and for the members of intermediate classes 3 and 4, as well as class 2, to become part of Class 5, thus showing favourable attitudes towards sexuality.

Strengths and weaknesses

An online survey such as the one carried out in our research may imply a bias in sample selection, but in the case of the university population, this situation is greatly attenuated since it is a group characterised by good Internet connection and considerable use of social media in general [43]. Furthermore, the anonymity it provides allows participants to talk about sensitive topics such as sexuality or mental health, thereby reducing the risk of stigmatisation or that the answers are not real [38].

This study is based on a cross-disciplinary study, making it difficult to establish causes or triggers and effects or consequences of the findings. More research is needed to verify the effect, for example, of the confinement that took place in the months after the start of the field work, and its implications.

Although all questions in the questionnaire were mandatory, it was observed that, in some cases, more than half of the students chose Don’t Know or Indifference. This lack of response or engagement may lead to bias, but far from being a limitation, some studies have already shown that not responding also implies a response [37]. The reasons for this lack of attitude were not analysed, although it could be attributed to ignorance, lack of information, or insecurity [38]. More research is particularly needed to confirm and clarify such indifference.

Finally, our research analyses students at a Spanish public university, but our findings should not be generalised to other universities or to the general Spanish population of that age range. Future studies are called for, since there is little in the literature which would allow the contrasting of results and conclusions.