Traditional Birth Attendants (TBAs) deliver more than 85% of the infants born in rural Sierra Leone. These midwives, who learn their skills from practice or a predecessor, are powerful women in the pervasive, secret Sande Society. Formal training programs have been established to strengthen the maternal and infant practices of these TBAs. Designing and implementing a quantitative research study to examine the effectiveness of two methods of TBA training required a strong sociocultural basis. The study was undertaken in a predominantly Mende ethnic area of Sierra Leone. I describe the significant ways in which the values and traditions of the Mende influenced the choice of research methods. Finally, I present critical questions researchers should ask to design a study that fits with the sociocultural fabric of their target population.