IB Docs (2) Team
Bias in qualitative research
HL Paper 3: Avoiding bias
This section focuses on the Paper 3 question: Discuss how the researcher in the study could avoid bias. First, we will look at the different types of bias (participant, researcher, sampling) and examine the ways that bias may be avoided in qualitative research.
Upon completion of this section, you should be able to discuss the following concepts:
- Acquiescence bias
- Ascertainment bias
- Epistemological reflexivity
- Inter-coder reliability
- Participant bias
- Peer review
- Personal reflexivity
- Sampling bias
Presentation
Read through the following presentation. There is a lot of terminology in this presentation. Be sure to take good notes that will help you to answer the quiz that follows below.
Checking for understanding
The following quiz tests your understanding of terminology with regard to generalizability, credibility, and bias. Choose the "best" response to each question.
A team of researchers carries out a set of interviews with students from four different IB schools to test their level of optimism about the future. Interviews are carried out during their final year of the IB program and then again, four years later. Which of the following is not true about this study?
The only method used is interviews. Even though more than one interview is carried out with each participant, this is still only one method.
Teachers are asked to carry out "student feedback questionnaires." The students are asked to answer questions about the planning of the course, the learning activities, the assessment, and the general feeling about the course. The teacher passes out the questionnaire toward the end of the class period and gives students 15 minutes to fill in responses to the five questions. Most students are done in five minutes and have written very little detail. What most likely happened here?
The students may have lacked motivation for this task and simply felt that they should "fill it in and get this over with!" Especially if students are asked to fill this in for all of their teachers and the standard is to have them fill it in at the end of the period, fatigue/boredom may influence their willingness to put much thought or effort into the responses. This ends up biasing the findings as those that really have an issue are more likely to write more.
During the Coronavirus pandemic, there were many interview studies carried out with students about their thoughts for the future. If you look at many of the studies, although they interview "students", the studies tend to only focus on university students. Students in trade schools - gastronomy, mechanics, nursing, or arts schools - were noticeably absent from the studies. What type of bias is this?
This is a clear example of ascertainment bias. A specific part of the population of students is either intentionally or unintentionally excluded from the sample.
A university wants to do a questionnaire to find out students' attitudes toward online education. In order for the results to be representative, they want to make sure that 40% of the questionnaires come from the humanities department, 30% come from the science departments, 20% comes from the health and human services department, and 10% comes from the technology department. To do this, they will accept the first 20 questionnaires received from the humanities department, 15 from science, 10 from health and human services, and 5 from technology. What type of sampling is this?
This is an example of quota sampling. The sample proportionally represents the population from which it is drawn, but the process for selection is not random. A stratified sample would be done in a similar manner, except the participants would be randomly selected.
A researcher carries out a study of worker stress at a large hospital in New York. Questionnaires, as well as physiological markers (blood pressure and cortisol levels), were taken from doctors and nurses. The researchers made sure that staff from every department and every shift (day and night). What type of generalizability were the researchers trying to achieve?
The researchers were attempting to make the sample representative of the population from which the sample was drawn. The goal is that they will be able to generalize the findings back to the entire staff of the hospital.
Which of the following is a way to control for social desirability bias?
Social desirability bias can be avoided by giving questionnaires where the researcher does not know which participant filled in which questionnaire.
Having more than one researcher carry out the interpretation of the data from a set of interview transcriptions is a way to control for which bias?
When several researchers interpret the same data and come to the same conclusion, we can say that the interpretation is credible and most likely bias has been avoided.
At the end of a published report when a researcher thinks about how the choice of research method may have influenced the outcome of the study, this is an example of which technique?
There are two types of reflexivity. Personal reflexivity is when the researcher considers how his or her own biases or personal experience may have influenced the findings of their study. Epistemological reflexivity considers how the choice of research method may have influenced the findings - for example, using a one-on-one interview style where all of the interviewers were American and the interviewees were Rwandans. The cultural differences may have had an effect on the willingness to disclose information to the interviewer.
What of the following statements reflects a phenomenological approach?
A phenomenological approach argues that an individual's thoughts, feelings, and personal experiences are essential to understanding their behaviour. This approach works with the participant to understand their behaviour, rather than seeing them as "subjects to be studied by experts." The other three statements represent "a positivistic approach."
Which of the following is not a strategy for ensuring credibility?
A double-blind control is important for establishing validity in quantitative research, but it is not a strategy used in qualitative research.
Continue to final assessments
Twitter
Facebook
LinkedIn