BUSI 820 Week 7 Integration of Faith and Learning Main Thread Example

Integration of Faith and Learning: Main Thread

BUSI 820 Quantitative Research Methods, Integration of Faith and Learning Main Thread

Thread by [Student Name]

The prompt and the bank study are composites standing in for the course's own; the books and the published research are real; the week, points and readings follow Liberty's public course guide for BUSI 820, which keeps the forum's detailed instructions inside the course.

For six weeks I have been analyzing an illustrative version of our bank's turnover survey, and it is easy to forget what the data set is. Each of its 186 rows records a teller or personal banker's answers to questions such as whether they often think about leaving their job. Behind the summated scales and the skewness values are people deciding whether the work they do every day is worth continuing. Keller and Alsdorf (2012) help me see why that matters for how I do statistics, not only for why I study turnover.

The first part of Keller's book presents work as built into creation from the start, and as dignified because the people who do it carry God's image. If the people in my data set have that dignity, then the data are not merely raw material. They are disclosures people made in trust, and the first obligation of the analyst is to protect them. In practice, that means reporting no result for a group smaller than five, so a branch manager cannot guess who wants to leave, and storing the data without names.

Keller also treats ordinary jobs as the main channel through which people care for one another, so that doing one's work well is itself a way of loving neighbors. Analysis is work, and it serves two groups of neighbors. The first are the decision makers who will rely on it; if I overstate a finding, they will spend money on the wrong solution. The second are the employees whose working lives will be shaped by those decisions. A careless claim that hybrid schedules cause turnover, for example, could take flexibility away from people for no good reason. My chi-square test in Forum 4 found no evidence for that claim, and serving my neighbors means saying so clearly.

The third part of Keller's book says the gospel recalibrates a worker's moral bearings, and that the recalibration shows up in routine decisions more than in dramatic ones. In statistics, the ordinary decisions are where integrity is tested. Simmons et al. (2011) demonstrated that a handful of choices researchers make without disclosing them, such as the point at which a researcher stops recruiting, the cases set aside as outliers, the controls added and the measures left out of the write-up, can push the chance of a false positive far above the 5 percent a p value promises. John et al. (2012), surveying thousands of academic psychologists, found that substantial shares admitted to such practices. The temptation in research is rarely to invent data; it is to keep adjusting honest data until it says what we hoped. My turnover study is vulnerable in the same ways. I have already seen that supervisor support has a small but significant association with intention to leave, and the tellers-versus-bankers difference narrowly missed significance. It would be easy to try a few covariates until that p value dipped below .05.

The integration leads me to four commitments for the study. First, I will write down my research questions, variables and analyses before collecting the real data, and report any departure from that plan. Second, I will report every planned test, including the nonsignificant ones, with effect sizes and confidence intervals. Third, I will treat outliers by a rule set in advance, not by their effect on significance. Fourth, I will protect respondents' confidentiality in every table. Keller and Alsdorf (2012) argue that grace frees people from needing their work to justify them; for a researcher, that freedom includes not needing a significant result. I would welcome classmates' views on whether these commitments are realistic under a dissertation timeline.

References

John, L. K., Loewenstein, G., & Prelec, D. (2012). Measuring the prevalence of questionable research practices with incentives for truth telling. Psychological Science, 23(5), 524-532. https://doi.org/10.1177/0956797611430953

Keller, T., & Alsdorf, K. L. (2012). Every good endeavor: Connecting your work to God's work. Dutton.

Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22(11), 1359-1366. https://doi.org/10.1177/0956797611417632