Dissertation Abstracts

Methods Effects in Mixed Methods Quasi-Experimental Investigation of Gendered Choice of School Subjects in Rwandan Secondary Schools

Author: Nzabonimpa, Jean Providence , njeanprovidence@gmail.com
Department: Sociology and Social Anthropology
University: Stellenbosch University, South Africa
Supervisor: Dr. Heidi Prozesky
Year of completion: In progress
Language of dissertation: English

Keywords: methods effects , mixed methods , concurrent and seque , educational choice
Areas of Research: Logic and Methodology , Women in Society , Education


Mixed methods research has gained momentum over the last decade, offering additional empirical avenues to deal with social inquiry. Despite this ascent in stature in social science research methodology circles, mixed methods is still approached with apprehension and labelled as insufficiently rigorous (Bergman, 2011). The methods effects of one methodological paradigm or data collection tool do not cancel out the effects brought in by the other (Bergman, 2011) but may likely increase the methods effects, thus producing different results. When data are collected using various data collection instruments, originating from two different research paradigms traditionally labelled incompatible, methods effects need to be assessed – something which has not thus been empirically investigated for sequential mixed methods design. It is against this background the research evolves: methodological in nature, yet embedded in the issue of gender in education as an empirical emporium to gather relevant data.

The study investigates whether three data collection tools, applied to the same sample to collect data on the same constructs in a sequential-convergent parallel mixed methods design, produce comparable, complementary, or conflicting results as shaped and dictated by methods effects. The three data collection tools chosen for this purpose are: semi-structured, personal interviews, autobiographic, personal narrative science-related essays, and structured, self-administered questionnaire. Despite an increasing list of mixed methods studies, little is known about methods effects in mixed methods concurrent and sequential methods designs as understood and used in real-life research to inform mixed methods inquiry.

In this study, it is hypothesised that the effects of each data collection instrument would surface in the course of the first phase of data collection, and that mixed methods effects, if any, would be (re)surface in a second phase of data collection, leading to different results. More specifically, it is hypothesized that phase-two respondents will be improving on rather than contradicting phase-one responses. It is also hypothesized questionnaire results will not significant differ from qualitative essay and interview results.

The empirical topic chosen to conduct the methodological investigation referred to above, is whether Rwandan students in high schools view their enrolment (or non-enrolment) in science subjects as the result of their own free choice, occupational goals, and/or of exogenous factors. It is hypothesised that respondents who have already chosen science subjects would have different views and experiences compared to respondents who have not chosen these subjects, either due to their actual or perceived particularistic or generalistic belonging to respondents’ prime reference group. This distinction will also be maintained at a methodological level with both particularistic subjectivity and generalisable evidence examined by merging quantitative and quantitised qualitative data from phase one and phase two.

We use own or third-party cookies to improve your user experience. If you allow the installation of cookies or continue to use this site, you agree to the use of cookies..

Read more