Thanks for the clarification. Actually, I had assumed that you would have collected both qual and quan information as you later mentioned. That bit didn’t worry me. (But José is right to make us clarify and articulate as many of our assumptions as possible. Just imagine the viva; how will you respond to similar questions? That image is how I try to conduct my work.)
My question to you was a little subtler than how I had phrased it. Please let me restate it with your expanded information. Let’s take a simplified example, but one that may be plausible within your study.
Let’s say that you find in the quantitative data that 34% more black students drop out than white students. And you also find in a similar data set over the same time period that black students’ academic records (as seen in their GPAs) is 45% lower than whites’. At first glance, there seems to be a strong correlation between the two figures. Some researchers may actually stop at this point and argue that the dropout rate is a result of the low GPA. Of course, this interpretation is incomplete. We need to ask why the GPA is low. So we collect some qual data, that is, we interview some students and ask about their GPA. If we are not careful, we may receive some very biased reasons. For example, if we ask directly about GPA (e.g. by asking, ‘Why do you think your GPA is low?’), participants may talk about their lack of understanding of lecture material, or low self-esteem that inhibits their involvement in the classes, or the fact that the study materials focus on white men’s achievements too much thus de-engaging the black students’ motivation, and so on. All of these reasons seem plausible, and I can imagine a report being generated on that basis that argues for a better understanding of the experience at college of black students.
However, let’s also say that the reality is that most black students (in this fictitious example) are poor and work much more than whites. They come to school tired, have less time to do homework and self-study, and their non-involvement in extracurricular activities means that they do not develop the degree of academic enculturation that white students have. These reasons would not have been captured in the first report.
The ‘biased reasons’ in the earlier paragraph refer to the kind of information retrieved by a simplistic and direct question. These reasons resulted in a study that was incomplete as the following paragraph demonstrated. Moreover, there may be even further reasons for the low GPA and the high dropout rate; for example, the university’s affirmative action policy allowed lower high school achieving black students entry into higher education, or the physical location of the university (combined with the local cost of accommodation/transport) inadvertently favoured the local white population due to its proximity to a white neighbourhood thus making black students poorer (due to costs) and the associated problems that that brings, or something else.
However, my point is that all of this post has been generated from just two quantitative data points, the GPA and the dropout rate. A problem was assumed because of these figures, and on the basis of the belief that a problem actually existed, a qualitative study (i.e. the interview on the GPA) and a further set of data collection (i.e. the amount of homework, the time travelled to school, the amount of extracurricular activities) was taken. Yet, all of this activity may have been based on an unsupportable assumption. This, to me, is one of the dangers of triangulation. How will you attempt to limit such biases?