Questionnaire development
The questionnaire comprises of four sections which are the information, motivation, behavioural skills and behaviour sections. For the information section, its items were adapted from the knowledge sections of some previous studies [6,11,15]. Items of the motivation and behavioural skills sections were developed from modifications of the relevant sections of the IMB questionnaire on diabetes mellitus self-care by Osborn et al. [16]. The questionnaire was first developed in English language, after which it was translated through the process of translation and adaptation of instruments outlined by the WHO [17].
It was at first forwardly translated to Hausa language by a senior University academic staff of the Hausa Language Department. This translation was then assessed by a native Hausa Public Health specialist, and then backwardly translated into English by a different translator of the same qualification. The two English versions (original and back-translated) were then compared by another Public Health specialist. This was followed by questionnaire testing
Questionnaire structure
Section A
This section had a total of 45 questions assessing the different domains of knowledge on malaria in pregnancy, which were: transmission, symptoms, complications, and prevention of malaria during pregnancy. There were three options for each question: ‘Yes’, ‘No’, and ‘I don’t know’. The total maximum obtainable information score was 45 points, while the minimum obtainable score was zero (0) points.
Section B
This section assessed participants’ level of motivation for sleeping under an ITN and taking IPTp. It had a total of 12 items, and comprised of two sub-sections (one on personal motivation, and the other on social motivation). The first four items on personal motivation asked of the participants’ perception of the level of goodness or otherwise of practicing those malaria preventive measures. These questions had response options on a five-point Likert scale thus: ‘very bad’, ‘somewhat bad’, ‘neither bad nor good’, ‘somewhat good’ and ‘very good’, which were scored 1, 2, 3, 4, and 5 points respectively. The next four questions on personal motivation assessed participants’ perception of the level of pleasantness or otherwise of practicing these preventive measures. These also had response options on a five-point Likert scale thus: ‘very unpleasant’, ‘somewhat unpleasant’, ‘neither unpleasant nor pleasant’, ‘somewhat pleasant’ and ‘very pleasant’, which were scored 1, 2, 3, 4, and 5 points respectively.
There were four questions on social motivation, which assessed how truly or not, their significant others thought they should comply with those malaria preventive measures. This section had response options on a six-point Likert scale thus: ‘very untrue’, ‘mostly untrue’, ‘untrue’, ‘true’, ‘mostly true’, and ‘very true’, which were scored 1, 2, 3, 4, 5, and 6 respectively. The total maximum obtainable motivation score was 64 points, while the minimum obtainable score was 12 points.
Section C
This section assessed participants’ levels of behavioural skills. It had a total of seven items and two sub-sections. The first sub-section had three items which assessed how hard or easy it was to comply with ITN and IPTp. Responses to this section were on a four-point Likert scale, thus: ‘very hard’, ‘hard’, ‘easy’, and ‘very easy’, which were scored 1, 2, 3, and 4 respectively. The second sub-section which assessed the level of effectiveness with which the participants could execute certain tasks relating to ITN use, had four items. This section had responses on a four-point Likert scale too, thus: ‘very ineffectively’, ‘ineffectively’, ‘effectively’ and ‘very effectively’, which were scored 1, 2, 3, and 4 respectively. The total maximum obtainable behavioural skills score was 28 points, while the minimum obtainable score was 7 points.
Section D
This section assessed their frequency of ITN use during pregnancy, that is, the number of days in a week in which they slept under an ITN. Frequency of ITN use was categorized as: Never, Seldom (once to twice weekly), Sometimes (thrice to 4 times a week), Often (5-6 times a week) and Almost always. These categories were scored as: 1, 2, 3, 4 and 5 respectively. This section also asked whether or not they had received any IPT, and the number of doses they had received.
Ethics approval and consent to participate
Ethical approval was obtained to carry out the research, from the Ethics Committee for Research Involving Human Subjects of the Universiti Putra Malaysia (UPM) (UPM/TNCPI/RMC/1.4.18.2 (JKEUPM). Permission was also obtained from the Ethics Committee of the State Specialist Hospital, Maiduguri (SSH/GEN/64/Vol.1). All the respondents were first taken through the respondent’s information sheet in Hausa language, after which informed verbal consent was obtained from them. This was due to the low literacy rates in the study location [18], and it had been approved by the JKEUPM.
Questionnaire testing
This was done in stages thus: content validity by experts, face validity by 20 pregnant women, test of construct validity by on different pregnant women, and finally test of reliability by 63 out of the 190 initial respondents.
Content validity
Content validity was assessed using an expert group [19] who went through the questionnaire to ensure that the wordings of its items were clear, and that they represent their content domain. The assessment team comprised of three Public Health specialists and an Obstetrics and Gynaecology specialist.
Face validity
Twenty antenatal care attendees were conveniently selected from a secondary-level health centre in Maiduguri, north-eastern Nigeria, to assess the questionnaire. The criteria for selection was to be a fluent Hausa speaker, and also be at their first antenatal visit for their index pregnancy. The questionnaires were administered to them by interviewers, following which they were asked to assess each section of the questionnaire, based on order of its questions, language clarity, and whether the questions under each construct appropriately measured the respective constructs. Order of questions was scored as Good order, Average order, or Poor order; language clarity was scored as Clear, Average or Confusing; while appropriateness of construct measurement was scored as Good, Average, or Poor.
Construct validity and reliability
A further cross-sectional study was conducted at the same antenatal clinic, a week after the face-validity study. A hundred and ninety respondents were conveniently selected using the same criteria as for face validity. They were similarly made to complete the questionnaire, and the data obtained was then analysed in IBM SPSS version 22. Internal consistency was measured using the Cronbach’s alpha. The motivation and behavioural skills constructs of the questionnaire were subjected to an exploratory factor analysis (EFA) to determine how properly the items of each of their respective sub-sections correctly fitted. The assumptions for conducting an EFA had been met, as the data for these two constructs were collected on an interval scale, and there were also positive correlations between all the items. Items with factor loadings less than 0.3 were suppressed, and for the rotation, the oblique method (promax) was chosen due to some high correlations among some items.
Two weeks after the first questionnaire administration, it was re-administered to 63, out of the 190 respondents. These 63 respondents were randomly selected from the complete list of the initial 190 respondents using the random function in Microsoft Excel 2013. The Cohen’s kappa was measured to determine the reliability between the answers at first and second administration, for items of the information section, since the responses were in a nominal form. For the motivation, behavioural skills and frequency of ITN use sections, the Krippendorff’s alpha were measured to determine reliability.