MRKT 600 Week 5 DB Forum 3 Example

Discussion Board Forum 3: Thread and Two Replies

MRKT 600 Applied Market Research, Discussion Board Forum 3 (thread and two replies)

Thread by [Student Name]

The prompt and the classmates' posts are composites standing in for the course's own forum; the textbooks and published studies are real; the requirements follow Liberty's public course guide for MRKT 600.

Thread

My survey's target population is adults who have selected a new primary care doctor during the last two years, since they can describe a real recent decision rather than a hypothetical one. A probability sample of that population would require a sampling frame, a list from which every member has a known chance of selection, and no such list is available to a student. Alreck and Settle (2004) explain that probability samples support statistical generalization because selection chances are known, while nonprobability samples, including convenience samples, are chosen for accessibility and cannot be generalized in the same way. For this course, a convenience sample is the honest choice. Pretending otherwise would invite exactly the overconfident conclusions the textbooks warn against, and it would weaken the analysis due in Module/Week 7 rather than strengthen it.

I plan to recruit about sixty respondents through three channels: my church's weekly bulletin, two workplace newsletters and a community Facebook group in my county. A screening question at the start, asking whether the person has picked a new primary care doctor recently, will route ineligible people out. Each channel brings its own bias, and naming them now will make the limitations section easier to write later. Church and workplace channels may overrepresent older, employed and more educated adults, and a Facebook group may overrepresent people comfortable online. That matters for my topic, because Terlutter et al. (2014) found that users of physician-rating websites were younger, more often women, more highly educated and more likely to have chronic conditions than nonusers.

To reduce imbalance, I will set rough quotas by age group, aiming for at least fifteen respondents under 35 and fifteen over 55, and I will report the sample's age, gender and education next to census figures for my county so readers can see how it differs. Creswell (2014) advises that survey designs state the population, the sampling method and its limits clearly, and my results section will say that findings describe these respondents, not all adults.

Classmates: what bias do your recruitment channels bring, and how will you show readers where your sample differs from the population?

Reply to Monica

Monica, your survey on how college students choose banking apps is a good fit for a convenience sample, and your questions about fees, security and peer recommendations look clear and well ordered. I noticed, though, that you plan to recruit through your daughter's high school parents' association and your own LinkedIn network.

Alreck and Settle (2004) stress that a sample must come from the population the study describes, and neither channel reaches college students directly; parents and professionals may answer about students rather than as students. Creswell (2014) suggests stating the population and how participants were reached so readers can judge the fit. You might ask a campus student organization or a professor to share your link, or post in a student group, and add a screening question confirming current enrollment.

Even a small sample of real students would serve your question better than a larger sample of people describing students. Parents might be a useful second group later, if you want to compare how students and parents see the same banking decisions, but the comparison only works if both groups answer for themselves. Do you have a contact at a nearby campus, perhaps a former professor, who could share the survey?

Reply to Andre

Andre, your plan to survey 200 shoppers outside two grocery stores on Saturday mornings about private-label brands is ambitious, and in-person intercepts can produce a strong response rate. Your thread described the sample as representative of the region's grocery shoppers, though, and I would be careful with that word.

Alreck and Settle (2004) explain that intercept samples are nonprobability samples, so they cannot be generalized with known accuracy, however large they grow. Saturday-morning shoppers at two suburban stores may differ from weekday evening shoppers, online grocery customers and people who shop at discount chains. Creswell (2014) recommends reporting the limits of the sampling method openly, which strengthens rather than weakens a study in the eyes of careful readers.

You could improve coverage by adding a weekday evening shift and a third store in a different neighborhood, and by describing the results as patterns among these shoppers. Recording the time, store and approximate age of everyone who declines would also let you report response rates by group, which tells readers whose views are missing. A short question about where else respondents buy groceries would show how many of them also shop online or at discount chains. Would the store managers allow intercepts on a weeknight as well?

References

Alreck, P. L., & Settle, R. B. (2004). The survey research handbook (3rd ed.). McGraw-Hill/Irwin.

Creswell, J. W. (2014). Research design: Qualitative, quantitative, and mixed methods approaches (4th ed.). SAGE Publications.

Terlutter, R., Bidmon, S., & Röttl, J. (2014). Who uses physician-rating websites? Differences in sociodemographic variables, psychographic variables, and health status of users and nonusers of physician-rating websites. Journal of Medical Internet Research, 16(3), Article e97. https://doi.org/10.2196/jmir.3145