In the early 2000s James Pfeiffer, Anthropology and Global Health Professor at UW, conducted research on Pentecostalism in Mozambique and the impact of the rapidly growing religion on public health efforts. The extensive mixed methods data collection yielded multiple publications and important insights about the healthcare and economic incentives Pentecostalism offered to the poorest communities and especially to women. Pfeiffer, a member of the QUAL Steering Committee, recently shared his project as part of the QUAL speaker series.
The emergence of a research puzzle
In the 1990s, Pfeiffer was conducting research with a colleague from the Anthropology and Global Health departments, that took them to Chimoio, Mozambique. They noticed that Pentecostalism – which originated in the United States and had multi-ethnic appeal here – was rapidly spreading in the African nation. Historian Mike Davis described the religion as “…arguably the largest self-organized movement of urban poor people on the planet.”
As an anthropologist, Pfeiffer was interested in what made this religion, with its many varying forms including the Neo-Pentecostals who taught a Prosperity Gospel, so appealing to the poorest residents of Chimoio. As a public health scholar, he was also fascinated by the role the religion played in healing – health workers he collaborated with told him the pastors in these myriad Pentecostal churches were their best allies when it came to reaching people about cholera vaccines or other public health interventions.
Method follows question
Pfeiffer told the room-full of graduate students and faculty at the QUAL talk that he secured funding for a grounded theory exploratory question: Why are Pentecostal and AIC (African Independent Churches) churches growing so quickly in this region of Mozambique?
Graduate students attending the QUAL speaker series asked a lot of questions about Pfeiffer’s participant observations and interview sampling protocol.
When designing a research study the research question should determine what methods one employs, Pfeiffer said. Under the many questions he wanted to explore, Pfeiffer had quantitative inquires and the kernel of a possible hypothesis – following a civil war that ended in 1992, inequality was growing. It was both tangible to an observer and came up often in local narratives, so Pfeiffer began to wonder what the connection of growing inequality was with conversions to Pentecostalism and AIC, and in turn the reliance on pastors’ healing and public health interventions vs. traditional healers.
Pfeiffer chose a mixed methods approach and had several data collection tools:
- He conducted a survey of 616 households to find out who was joining the Pentecostal and AIC churches and how many churches there were in Chimoio.
- In 2002, he also conducted in-depth interviews with 80 recent converts to churches and recorded their illness narratives. In 2003, 30 follow up interviews were recorded.
- To those, he added 50 key informant interviews with pastors, prophet healers, traditional healers, health workers, government personnel, NGO workers, and community leaders.
- Lastly, he added participant observations – 25 church services, fasting rituals, and baptisms.
Pfeiffer hired two local research assistants to accompany him to all interviews and assist with the multi-lingual transcription and translation. To code and analyze the vast quantities of qualitative data collected, he used ATLAS.ti.
Methodological challenges
One of the first challenges came during the sampling process of identifying individuals for illness narrative interviews and again when establishing trust with the interviewees.
“Your qualitative work is only as good as your ability to get people to open up,” Pfeiffer said.
He and his research team abandoned an early idea of recruiting interviewees randomly from the households surveyed. Instead, they relied on snowball sampling and on pastors to introduce them to their congregations and help recruit illness narrative interviewees. Despite the shortcomings – such as self-selection bias – this approach was best for collecting rich interview data, Pfeiffer said.
[pullquote]Your qualitative work is only as good as your ability to get people to open up.[/pullquote]Transcription proved to come with its own slew of challenges. Some people did not want to be recorded and Pfeiffer had to rely on the research assistants’ detailed notes. Others, who agreed to be recorded, may have been more guarded in how they described their experience of joining the church. Interviews were conducted in Portuguese and a myriad of local languages – thankfully, the research assistants Pfeiffer hired locally could help navigate the multi-lingual transcription process.
Beyond the multi-lingual challenge, meaning interpretation of the interviews added another layer of complexity. Here, relying on his cultural insider research assistants was key, Pfeiffer said.
Conclusions
Two key themes that emerged from the inductive research on Pentecostalism and health in Chimoio were about economic incentives and women’s agency. Pfeiffer concluded that women were more likely to take the first step and turn to pastors for help with reproductive health concerns because traditional healers cost money and often put blame for reproductive challenges with the women.
Pfeiffer credited his participant observations and the trust he built with congregation members for allowing him to gain deep insights about the motivations for conversion and for embracing of the Pentecostal pastors’ healing rituals and healthcare advice at the expense of traditional healers.Valentina Petrova