Respondents Background and Sampling Strategy

Background. – The respondents for this research consisted of 175 males and 122 females. They were categorized by age into young adults (ages 20-34 years, n=255), middle-aged adults (ages 35-49 years, n=37), and older adults (aged older than 50 years, n=5). Background data relating to educational attainment is assessed as 1) secondary/vocational school, n=39; and college/university students, n=258. Here, it is important to note that a large number of respondents were also students at the time of the survey research. Their religions comprised of 1) Catholic, n=71; 2) Protestant, n= 157; 3) Vodou, n=21; 4) others, n=20; and no religion, n=28.

Of the 297 respondents, only 8 reported that they did not receive RE during their upbringing, and one respondent remained neutral. This finding reflects Haitians’ schooling experiences as majority of schools in Haiti are religious/sponsored by religious organizations. In other words, a significant number of Haitians had been somehow received RE throughout their lifetime/schooling. Note this RE was in most instance confessional RE (i.e. Bible teaching or Catechism). The respondents came from the 3 main regions of Haiti, including 1) the big north, n=141; 2) the big south, n=43; and the transversal, n=113.

Overall, the 297 respondents represent a small population sample. A larger representative sample size (i.e. in terms of age, education, and religion) may have led to the research offering an increased or significant set of findings. However, a brief review of the background information shows that a majority of the respondents fell into the 20-34 age category and had benefitted from college or university education. This reality may help to explain that in Haiti, young people are more likely to be familiar with technology, easier to reach and communicate with, as well as be more willing

166

to fill out both online questionnaires and paper and pencil questionnaires than older people. Indeed, filling the questionnaire on Survey Monkey requires certain internet skills and reading skills which are indicative of the younger generation as opposed to older people (Katz & Rice, 2002).

It indicates that half of the respondents identified their religion as Protestant; this is an error of the research method given that the intended technique did not work as planned (see appendix D).

Indeed, the reality also is that all the 57 institutions that were contacted (via email) in Haiti did not reply to the request of the survey research. Therefore, it was necessary to adopt a different strategy for data collection. The new strategy required the researcher to send the survey hyperlink to respondents through personal contact systems on social media, including Facebook, Messenger, and via email. I believe that this strategy heavily influenced the data collection system.

Subsequently, it was considered important to undertake a visit to Haiti so that more data could be collected, hence the inclusion of a survey that involved paper and pencil. This approach allowed the researcher to access Muslim and Vodouist respondents who, it appeared, undertook little activities online.

As a result of the Haiti’s trip, the study collected data from 10 respondents who identified their religion as Muslim and 20 who identified themselves as Vodouist. While both male and female respondents were representative for the Vodou, the 10 Muslims who completed the survey questionnaire were only males. As a parenthesis, these 10 Haitian male Muslim respondents could not have been reached without a trip to the Mosque in Port au Prince, Haiti. Based on my observation at the Mosque yard, many males came to converse with me but the women. The Haitian female Muslim went straight up inside without making eye contacts. Additionally, another

167

observation was when I was asked to partake in the worship service. The mosque is designed to segregate male and female. A large area made up the podium, the flat-and-open space for males and a designated place (closed) for the Muslim women through which you can see the upper part of their bodies only. This was very remarkable to observe as a first time visitor (both male and female first visitors sat near the doorway on the flat and open space area with the Muslim men).

Though it was not a requirement for first time visitor females to join the Muslims women, but I felt double segregated (i.e. a feeling of non-appurtenance and brain-teaser experience).

Sampling. – For this study, students from secondary and higher education institutions in Haiti was selected as the targeted population. Such selection was undertaken due to the fact that this sampling has the potential knowledge background about the education system in Haiti; so, they are at a favorable position to opine (or foresee) on the type of contribution teaching RE in Haiti schools would add to the society. In other words, they would more likely be able to reflect on the strength and weakness of the status quo i.e. confessional RE in Haiti’s schools.

Simple random sample technique was intended to use in this research, and this was for two main reasons: to reduce the potential of human bias and to make generalizations from the sample to the population. The technique that actually used to draw a sample from school and higher education institutions in Haiti was a convenient sampling because no name list was provided by any institutions. In fact, the fifty-seven institutions that were contacted (via email) in Haiti did not reply to the request of the survey research. Thus, the proposed random sampling method (i.e. Neuman, 2011, 2006, 2003) did not work as intended. This reality made it necessary for the study to use a convenient sampling method. This method involved sending the survey hyperlink to respondents

168

through personal contact systems on social media, including Facebook, Messenger, and via email.

The result heavily influenced the nature of sampling population – it misrepresents the Haitian population – (i.e. half of the respondents identified their religion as Protestant). Indeed, Neuman was right when he stated that when we select cases based on convenience, our sample can seriously misrepresent features in the entire population (p.242).