The six specific objectives of the study were to 1.

describe the sexual violence experiences of students in tertiary educational institutions;

This study was hinged on the Theory of Planned Behaviour (TPB). The theory describes that behavioural intention of an individual to engage in a particular behaviour is a product of three determinants: the individual’s attitude (opinions of oneself towards the behaviour); the subjective norms (opinions of others towards the behaviour); and perceived behavioural control (self-efficacy towards the behaviour) ( Ajzen, 2012). For an individual to have intention to blow the whistle for reporting sexual violence act, he/she must have a positive attitude towards whistleblowing as an effective mechanism. Such individual must also consider that, his/her action will be acceptable by significant others and that he/she has the capacity and the wherewithal to blow the whistle. However, aside from the three determinants, background variables such as demographic factors have been identified to indirectly influence behavioural intention through the three determinants and consequently the behaviour ( Menozzi et al., 2015). It is expected in this study that the socio-demographic factors of the students and staff such as academic level, course of study, gender, age, sexual violence experiences, tribe, religion, designation and length of service will influence their intention to internally blow the whistle for reporting sexual violence cases. In cases of sexual violence, being a culturally sensitive issue, a reprisal attack on the whistleblower is inevitable. Therefore, it is essential that protection mechanisms are established to protect the whistleblower ( Figure 1).

Institutions where the study was conducted were those owned and controlled by the Federal Government of Nigeria and must either be a polytechnic, college of education or university. The students were those undergoing full-time traditional face-to-face programmes and were undergraduates. Staff participants were those who work full-time mode in the selected institutions. Contract or part-time staff were excluded from the study.

Using Cochran’s survey sample size formula: n = z ²pq/e ² where n = sample size; z = standard normal deviate set at 1.95 which corresponds to 95% confidence level; p = 0.5 (set at 50% for unknown prevalence of whistleblowing intention); q = 1-p = 0.5; e = permitted error of margin set at 2.5% (in order to get a large representative sample), a sample size of 1,521 students was obtained per institution. Adding non-response rate of 10%, a final sample size of 1,673 was obtained per institution giving a total of 30,114 students’ respondents for the 18 institutions. Using the same Cochran’s formula: n = z ²pq/e ² and same parameters for the staff except for the permitted error of margin set at 5%, a sample size of 418 was obtained per institution after adding 10% non-response rate. Therefore, a total of 7,524 staff (academic & non-academic) was selected and distributed proportionately in all the 18 institutions and into equal proportion of male and female staff.

The number of participants in non-commercial focus groups is recommended to be between six and 10 in order to be able to control the group ( Kabir, 2016). Therefore, we proposed to have eight members (four males and four females) of the Student Union Government in each of the institutions to participate in the study making a total of 144 students for the 18 institutions. Also, two representatives (a male and a female) from each of the Staff Unions in each institution were selected to participate in the FGD. The Unions represented are listed according to the type of school: •

University: Academic Staff Union of Universities (ASUU); Senior Staff Association of Nigerian Universities (SSANU); National Association of Academic Technologies (NAAT); and the Non-Academic Staff Union of Universities (NASUU).

Therefore, for Universities, there were eight Executive Officers (four males, four females) from each institution giving a total of 48 participants for the six Universities. For Colleges of Education and Polytechnics, there were six officers (three males and three females) from each institution giving a total of 72 participants for the 12 institutions. Overall, a total of 120 staff formed the sample size for the staff FGD participants.

Multistage stratified sampling technique was adopted in this study. Nigeria is already stratified into six geo-political zones and in each zone, government-owned higher institutions were stratified by University, Polytechnic and College of Education. In each of the stratum, one institution was selected using simple random sampling technique (balloting) thereby giving a total of three schools from each zone ( Figure 2 - produced by the authors). A random selection of departments from each of the schools were done followed by the students’ and staff stratification into male and female gender and also stratification into academic and non-academic staff. A random selection was finally used to select the respondents by proportion as shown in Table 1. For the qualitative strand, the Presidents of each of the unions were contacted to nominate members of their cabinet to participate in the FGD putting into consideration equal gender in the nomination. However, it was observed that there were more males in many of the cabinets than females hence, more male officers participated in the FGD than females.

Two instruments were used for data collection: Sexual Violence Whistleblowing Intentions Questionnaire (SeVWIQ) and a Focus Group Discussion Guide. The SeVWIQ was in two forms. The first was administered to students and it elicited information on their attitude to whistleblowing, subjective norm, self-efficacy, whistleblowing intentions, sexual violence experiences and demographics. The second was for the staff and it elicited information on their attitude to whistleblowing, subjective norm, self-efficacy, whistleblowing intentions, and demographics. The questionnaire had 10 attitudinal questions on a 3-point Likert scale of Not True (1), Moderately True (2) and Very True (3). The subjective norm questions were nine on a 3-point Likert scale of Not Proud (1), Slightly Proud (2) and Very Proud (3). The self-efficacy questions were six on 3 point-Likert scale of Not Likely (1), Moderately Likely (2) and Very Likely (3). The questions on the attitude, subjective norm and self-efficacy were adapted from the work of Zakaria et al. (2016). The sexual violence experiences component was adapted from the Sexual Experiences Short Form Victimization (SES-SFV) scale developed by Koss et al. (2006). The SES-SFV scale assessed sexual violence victimization in the past 12 months and the number of occurrence. The whistleblowing intention component of the questionnaire were formulated using three vignettes which were adapted from the work of Ahmad (2011). The vignettes represented cases of rape, attempted rape and sexual harassment and were rated on a scale of 1 to 5. The Focus Group Discussion Guide (FGDG) which was developed from the review of literature, explored information on sexual violence cases on the school campuses, institutional laid down strategies for reporting sexual violence, factors that may predict the staff and students’ intention to blow the whistle, strategies for reporting and protecting whistleblowers, and the effectiveness of internal whistleblowing strategies.

The pilot study was carried out in a University of Technology in South West Nigeria. Using the recommendation of Connelly, (2008) that 10% of main sample size can be used for pilot study, therefore, 167 students (10% of 1,673) and 42 staff (10% of 418) were selected to participate in the pilot study. Eight members of the Student Union Government and eight members of the leadership of the four Staff Unions of the school participated in the two separate FGD sessions (one for the students and the other for staff). During the pilot study, the questionnaire was subjected to face and content validity and internal consistency, while trustworthiness of the FGD data was ensured. Part of the findings from the pilot study has been published in Ogunfowokan et al. (2023a). Also, the questionnaire’s internal consistency was further assessed during the main study to further evaluate its reliability. The Cronbach alpha results for both the pilot and the main studies are stated in Table 2.

A field supervisor was identified in each of the schools except the three Northern regions where only one field supervisor was identified and he worked with the investigator from the North to collect the data. The supervisors were responsible for the logistics of sampling and the data collection process. They were trained twice using the virtual mode between December 2020 and January, 2021. The principal investigator was the chief trainer supported by other investigators. The field supervisors then appointed field workers in their respective schools and were trained in the administration of the questionnaires. Also, the two mentees who were postgraduate students as required in the funding call served as the research assistants for the study. The mentees have been part of the writing process for the grant proposal and they were further trained alongside the supervisors.

Institutional board approval for the study was obtained on 11 ^th December, 2019 from the Institute of Public Health, Health Research and Ethics Committee (IPH HREC) of the Obafemi Awolowo University with approval number IPH/OAU/12/1460. An extension of the approval was obtained on 5 ^th May, 2021 at the expiration of the initial approval when data collection has not been completed. Informed consent of the students and staff was obtained using the informed consent form formatted after the IPH HREC form. The informed consent was obtained by the field staff after the purpose of the study has been explained to participants and concerns clarified. However, due to the sensitivity of the questions, many participants declined appending their signatures on the consent form but preferred verbal consent without recording, which was granted them. In addition, written permission to collect data was also obtained from the administrators of the respective study institutions.

Due to the sensitivity of the study and the non-appealing nature of electronic data collection to many Nigerians as previously observed ( Tella, 2015), we decided to use the paper-based questionnaire administration methods. The selected students were contacted by the field workers in their various lecture rooms and questionnaires were administered to them and same were retrieved immediately after completion. Some of them requested to return the questionnaire the following day. Also, some class representatives assisted in retrieving the questionnaires from those who made such request. The questionnaires for the staff were administered to them in their various offices. Some of them returned the questionnaires immediately after completion while others requested to return them after some days. Some staff also dropped the questionnaires in their pigeon holes for pick up. In addition, two FGD sessions (one for students and the other for staff) were conducted in each of the schools giving a total of 35 FGD sessions (the staff in a particular school refused to participate in the FGD). For the students’ FGD, they were carried out at the Student Union Government building except for one of the schools in which the school management insisted it should be conducted at the Students’ Affairs building. The staff FGD were conducted at a designated venue agreed-on between the field supervisors and the participants. The principal investigator facilitated the sessions in support of other team members for the Southern regions. For the Northern regions, the field supervisor and the investigator from the North facilitated the FGD sessions. Recordings of the sessions were taken using Android phones while field notes were taken by other support staff.

Out of the 30,114 students’ questionnaires and 7,524 staff questionnaires administered, a total of 21,937 students’ questionnaires and 3,108 staff questionnaires were completely and properly filled and were adequate for data analysis giving a response rate of 72.8% for students and 41.3% for staff. There were eight data entry clerks who were recruited and trained in June, 2021. Two data analysts were also employed for the quantitative strand and these analysts trained the data entry clerks on how to enter the coded data from the questionnaires into Epi Info software version 3.1. Hands-on training was done using pre-coded demo-questionnaires. All the questionnaires were numbered and coded for data entry. The clerks received different quantities of questionnaires on different days for entry into Epi info software. A standby research office assistant was also employed to assign the questionnaires to the clerks and also ensured that correct quantities of the questionnaires were returned after data entry. Data entry was done within a period of seven months. Following the entry of the data and the merging of the data files from the clerks by the analysts, the data were then exported to SPSS version 27 ( IBM Corp. Released, 2020).

The quantitative dataset ( Ogunfowokan et al., 2023b) consists of 44 variables for the students and 42 variables for the staff that are grouped into six: (i) attitude to whistleblowing; (ii) subjective norm; (iii) self-efficacy; (iv) whistleblowing intention; (v) sexual violence experience; and (vi) socio-demographic characteristics. However, the staff dataset does not contain the sexual violence experience data. The list of all the variables and their corresponding response options are found in Tables 3 and 4. The qualitative dataset contains transcripts on responses to the six probing questions used for the FGD which are: (i) sexual violence cases in the school; (ii) institutional reporting systems; (iii) factors influencing intention; (iv) preferred whistleblowing strategies; vi) protection of whistleblower; and (vii) effectiveness of internal whistleblowing.

During sampling of the schools, some areas in the Northeast and Northwest regions were avoided because of insurgency. After sampling, two schools were dropped because the field workers could not gain access to the schools due to ongoing insurgency in the areas. Also the study was conducted in government owned schools thereby excluding higher institutions owned by private individuals, organizations and State Governments. Other higher institutions like diploma schools of nursing, technical and specialized higher education schools among others were also not captured in the study. Hence, generalization of the findings from this study should only be for Universities, Polytechnics and Colleges of Education that are Federal government-owned. Also, FGD for staff could not be conducted in a particular university because the executive officers for the staff unions did not give their consent to participate despite all pleas and provision of information about the study.

Quantitative dataset

The questionnaires that were brought from the field were assessed for completeness by the research office assistants and two other young people employed for the job. All incompletely filled questionnaires were discarded. Consistency checks were done by ensuring that the data on each of questionnaires were consistent. For example, on the staff questionnaire, an academic staff who reported to be a Senior Lecturer later ticked the non-academic status had the non-academic status corrected. Duplicate and outliers on the dataset were also observed and compared with the original raw data on the corresponding questionnaires and corrections were made where necessary. Missing data were attended to by using mean substitution as documented by Kang (2013). However, we were aware of the bias related to self-reported data collection most especially on a sensitive issue as sexual violence and this may impact on the quality of the dataset. Also, we could not do external validation of the data set as we could not lay our hands on a similar dataset on any of the variables.

Qualitative dataset

A most profound validation for the qualitative dataset that we carried out is the data member checking. During and after the interview, participants were regularly instructed to confirm their statements as true reflections of their opinions by listening to the recordings. Facilitators were also encouraged to clarify participants’ opinions from the statements that they made to avoid biases. Peer debriding was also done among the investigators, supervisor and the research assistants immediately after concluding an FDG session to discuss the findings in proper perspective and to also avoid biases. All these efforts were to ensure the trustworthiness of the qualitative data ( Stahl and King, 2020).

There are six investigators and two research mentees in the research team. The team consists of individuals who are academics from four universities spread across three regions in Nigeria. The research mentees were Ph.D. students (a male and female) in community health nursing, of whom one of them focused on adolescent boys’ sexual violence perpetration in her Ph.D. project while the other focused on work-related stress among nurses. Both mentees were trained in the use of qualitative and quantitative data collection methods for their M.Sc. and Ph.D. projects. The six investigators (three males and three females) had Ph.D. certificates in diverse specialised areas, which include health promotion, civic education, gender and policy studies, demography and population studies, medical sociology, and community health. Some of the investigators were trained in phenomenology studies, while others were trained in quantitative data collection. The strengths of the investigators were identified and made use of during data collection. The principal investigator, who is a community health nurse expert, facilitated the FGD sessions in the southern regions. The assigned field supervisor for the Northern region, who is a sociologist and a demographer, facilitated the FGD for the Northern region in conjunction with the investigator from the North, who is a medial sociologist. In order to eliminate any form of bias, information about the researchers was also given alongside the research information at the point of obtaining informed consent from the participants.

Some schools in the northern part of the country were intentionally avoided as a result of the insurgency going on in those areas at the time of conducting the study. Also, two schools that were sampled initially were replaced at the point of data collection when the field workers could not gain access to the schools due to insurgency. Other higher institutions, like diploma schools of nursing technical schools and specialised higher education schools, among others, were not captured in the study. Hence, generalisation of the findings from this study should only be made for students in universities polytechnics, and colleges of education. The staff in one of the institutions refused to participate in the FGD despite all efforts to convince them that the study is solely for research purposes with no implication for any individual.