Questions & explanations
1. What ethical issues does a researcher face when studying a public online forum for teachers?
Even if the forum is public, members may not expect researchers to watch them. The researcher should consider asking permission or at least telling the group. Another issue: usernames might not hide identity if someone uses a real name. The researcher must anonymize data, like changing names and details. Also, posts might contain personal stories about students or schools. The researcher needs to protect privacy. A key question: do they need consent from each member, or from the forum owner? Many researchers get consent from the community if it is closed. It is important to be transparent about the research purpose and let people opt out. Ethical digital ethnography respects the online space as a community, not just data.
2. Compare Bayesian meta-analysis and frequentist meta-analysis in terms of handling heterogeneity.
Both methods handle heterogeneity, which is variation in effects across studies. In frequentist meta-analysis, heterogeneity is modeled with a random-effects distribution, typically a normal distribution with variance tau-squared. Bayesian meta-analysis also uses a random-effects model but can incorporate prior information on tau-squared, which is often hard to estimate precisely. Bayesian methods can provide more stable estimates of tau-squared when studies are few. The Bayesian approach yields a distribution of tau-squared, reflecting uncertainty. In contrast, frequentist methods give a point estimate and confidence interval. Bayesian meta-analysis can also include additional random effects or covariates more flexibly.
3. What is 'positionality' and how does it connect to power in research?
Positionality means the social position of the researcher—things like their race, gender, age, class, and identity. This position affects how participants see them and what data they get. For example, a young female researcher asking teenage girls about dating may get more open answers than an older male researcher. The researcher's position gives them either power or shared experience. If a researcher is an outsider (different culture), participants may feel less trust. If they are an insider (same group), participants may assume the researcher already understands them. Researchers should reflect on their positionality and how it shapes the research. Being aware helps reduce bias and builds better relationships.
4. How can we check if the regression discontinuity assumption is valid?
The main assumption is that students just above and just below the cutoff would have similar outcomes without the treatment. This is called the continuity assumption. One way to check is to see if the number of students is smooth around the cutoff; a sudden jump suggests people might be manipulating their scores. Researchers also test if other factors, like gender or prior grades, also jump at the cutoff. If these factors change smoothly, the assumption is more believable. Another check is to look at outcomes that should not be affected by the treatment, such as height, and see if they change at the cutoff. If they do, there might be a hidden problem. These checks help make sure the RD estimate is trustworthy.
5. How can a researcher reduce power imbalances when studying a community of parents who have little formal education?
The researcher can start by visiting the community and building relationships before asking questions. They should use simple language and avoid academic terms. Letting parents choose where and when to talk (e.g., at a community center or their home) gives them control. Offering to explain the research in their native language or with a translator shows respect. The researcher can also involve a trusted community member as a co-researcher or liaison. Most importantly, the researcher should listen more than talk and never correct or judge parents' views. Sharing results with the community and asking for their feedback also shares power. These steps make parents feel respected and more willing to share honestly.
6. How can researchers avoid over-reliance on meta-analysis while still using systematic review methods?
Researchers can focus on describing the range of results rather than only pooling them. They should explore heterogeneity through subgroup analyses or meta-regression. Instead of producing a single effect estimate, they can present forest plots with individual study results clearly. They can also use methods like qualitative synthesis or vote counting when meta-analysis is inappropriate. Sensitivity analyses can show how robust the conclusions are. Researchers should interpret meta-analysis results cautiously, especially when studies are few or of poor quality. Finally, they should consider other types of evidence, such as large trials or observational studies, and not rely solely on the meta-analytic average.
7. How can a researcher collect data in a digital ethnography of an online course discussion board?
The researcher first gets permission from the course instructor and students. Then they download all discussion threads, save timestamps, and note who posts. They might also take screenshots of interesting exchanges. They can use software to organize posts by theme or date. Over time, they add field notes: what patterns they see, like which posts get many replies or which students never post. They might also interview some students about why they did or did not participate. All this text and notes become the data set. The researcher looks for themes like collaborative learning, confusion, or social presence. Digital ethnography combines careful saving of digital traces with the researcher's own observations.
8. Compare a study where the researcher is a school principal interviewing teachers, with a study where an outside researcher does the same. How do power dynamics differ?
When the researcher is the principal, teachers may fear that their answers affect their job or evaluations. They may avoid criticizing the school or themselves. The principal-researcher has direct authority over teachers, so the power gap is large. An outside researcher has no authority over teachers' jobs, which makes teachers more likely to speak honestly. However, teachers may still see the outsider as an expert and try to give 'correct' answers. Both situations have power issues, but the principal has a bigger power advantage. An outside researcher can build trust by explaining that participation is voluntary and anonymous. The principal should probably not research their own staff to get honest data.
9. How does action research in Indigenous education differ from Western action research?
Western action research often comes from universities and focuses on individual teachers solving problems. Indigenous action research is community-based and values the whole group. The research questions come from the community's needs, not from outside researchers. Indigenous methods include storytelling, talking circles, and land-based learning. The goal is not just to improve test scores but to strengthen cultural identity. Western research tends to be written and published; Indigenous research shares knowledge orally and through practice. Also, Indigenous action research sees knowledge as connected to the land and ancestors. Both use cycles, but Indigenous versions center on relationships and respect.
10. How does a selection model adjust the meta-analysis result?
A selection model adjusts the meta-analysis by incorporating a weight for each study that reflects how likely it was to be published given its results. Studies with very significant results get lower weights because they are overrepresented in the literature. Studies with non-significant results get higher weights because they are underrepresented. The model estimates the selection function, often using a step function based on p-value thresholds. Then, the pooled effect is re-estimated with these weights. The adjusted result is usually smaller than the original if publication bias is present. The model also provides a corrected confidence interval that accounts for the selection uncertainty.
11. What is one problem with arts-based research that researchers need to think about?
One problem is that art can be hard to interpret. A researcher might see a sad face in a drawing, but the student might have just liked drawing that way. To avoid mistakes, researchers always ask participants to explain their own art. Another problem is that some people feel shy about drawing or acting—they may not want to share personal feelings through art. Researchers need to make sure participation is comfortable. Also, artistic skill can affect expression: a student who draws well might show more detail than someone who is not confident. Researchers should focus on meaning, not beauty. So careful methods like asking follow-up questions are needed to make the findings trustworthy.
12. Compare systematic reviews for social interventions versus environmental interventions.
Both social and environmental reviews follow the same basic steps: search, select, assess quality, and summarize. Social reviews often study programs for people, like education or crime prevention. Environmental reviews study actions on nature, like pollution control or habitat restoration. Social reviews often have more subjective outcomes (like well-being) that are hard to measure. Environmental outcomes (like species count) are often easier to measure with numbers. Also, social programs can be tested with random assignment, while environmental interventions often use natural experiments. Despite these differences, both kinds of reviews produce reliable evidence for decision-makers.