DBAX 9103 Week 8 Results Number Audit Example

Course	DBAX 9103 Business Research Project Stage 3
Week	Week 8
Paper type	Number audit
Length	About 770 words, 5 pages
Format	APA 7 student paper
School	Walden University
Program	Doctor of Business Administration (DBA)
Updated	October 2026

Free sample paper for DBAX 9103 Week 8

Numbers Drift Between Drafts: A Number Audit of a Caregiver Retention Results Chapter, Tracing Every Figure Back to the Analysis Output and Recording the Seven That Had Changed Without Anyone Noticing

Student Name

Walden University

DBAX 9103: Business Research Project Stage 3

Instructor Name

Month Day, Year

What this page is doingThe title reports the audit's result, seven drifted numbers, which shows why the audit was worth doing.

Why Audit the Numbers

Numbers drift between drafts. A mean is fixed in a table while the paragraph above it keeps the old value; a sample size changes after an exclusion and survives in the abstract; a percentage is computed from an earlier version of the data. Nuijten et al. (2016), checking statistical results in tens of thousands of psychology articles, found that about half contained at least one p value inconsistent with its test statistic, and that one in eight contained an inconsistency large enough to change a conclusion. Most such errors are not fraud; they come from copying, rounding and revision. This audit traces every number in my results chapter back to the analysis output.

What this page is doingThe audit opens with how drift happens, so the reader sees that errors arise from ordinary revision rather than carelessness.

How the Audit Was Done

I listed every number in the chapter, 94 in all, including those in tables, text, the abstract draft and figure captions. For each, I recorded where it appeared, the value stated and the value in the current analysis output. Where a number was derived, such as a percentage, I recomputed it from the counts. For each odds ratio, I checked that the p value was consistent with the interval by recovering the standard error from the interval's width.

What the Audit Found

Table 1

Numbers That Did Not Match the Current Output

Location	Stated	Current output	Cause \|
Abstract draft	n = 576	n = 531	Written before the 14-day window change \|
Abstract draft	odds ratio 1.41	1.33	Same; the biased 30-day estimate \|
Text, sample paragraph	69.8 percent completed onboarding	72.3 percent	Computed on 576 rather than 531 \|
Text, research question 2	66.4 percent retained among completers	70.3 percent	Same \|
Table 1, check-ins mean	1.40	1.64	Before 211 scheduling-only entries were reviewed again and some restored \|
Table 2 note	14 supervisors	22 supervisors	Count of supervisors with any new hire was confused with count in one region \|
Discussion draft	"about a third higher odds"	1.33	Acceptable as a paraphrase, but replaced with the number \|

All other 87 numbers matched the current output.

Reading the Errors

Five of the seven errors trace to one event: the decision to change the check-in window from 30 to 14 days. That decision changed the sample, and numbers written before it survived in places that were not regenerated: the abstract draft and two sentences of text. The check-in mean changed because the scheduling-only coding was revisited after the agreement check, and the paragraph was written from an earlier file. The supervisor count was simply wrong. None of the seven errors would have changed a conclusion, but the abstract errors would have reported the biased estimate as the main finding.

What this page is doingEach error is traced to its cause, since the cause shows where the next error is likely.

Consistency Checks

All six odds ratios in the model table had p values consistent with their intervals. For example, the check-in interval of 1.10 to 1.61 implies a standard error of about 0.10 on the log scale, which with an estimate of 1.33 gives p of about .003, as reported. Cumming (2014) notes that an interval and a p value computed from the same model carry the same information, which is why one can be checked against the other.

Preventing Further Drift

Every number in the chapter now comes from the analysis script, which writes the tables and a file of in-text values that the chapter draws from. The abstract will be written last, after every other number is final. Appelbaum et al. (2018) recommend that reported numbers be reproducible from the stated analysis; generating them from the script makes that a property of the document rather than a promise.

Who Else Checked

After my audit, I asked a classmate to pick twenty numbers at random from the chapter and trace each to the output file without my help. The classmate found all twenty and agreed with each. A second pair of eyes cannot prove there are no errors left, but it shows that the trail from output to text can be followed by someone other than the author, which is what a committee member will try to do.

What the Audit Cost

The audit took about six hours. Most of that time went into finding where each number appeared, which a single source file will make faster next time. The remaining time went into recomputing derived figures by hand as a check on the script itself. Set against the risk of a committee member finding a wrong figure during the defense, six hours is a small price.

References

Appelbaum, M., Cooper, H., Kline, R. B., Mayo-Wilson, E., Nezu, A. M., & Rao, S. M. (2018). Journal article reporting standards for quantitative research in psychology: The APA Publications and Communications Board task force report. American Psychologist, 73(1), 3-25. https://doi.org/10.1037/amp0000191

Cumming, G. (2014). The new statistics: Why and how. Psychological Science, 25(1), 7-29. https://doi.org/10.1177/0956797613504966

Nuijten, M. B., Hartgerink, C. H. J., van Assen, M. A. L. M., Epskamp, S., & Wicherts, J. M. (2016). The prevalence of statistical reporting errors in psychology (1985-2013). Behavior Research Methods, 48(4), 1205-1226. https://doi.org/10.3758/s13428-015-0664-2

The Walden DBAX 9103 Week 8 instructions

Week 8 asks you to audit every number in your results chapter. Explain how numbers drift between drafts. Describe your audit method: list every number, record where it appears, compare it with the current analysis output and recompute derived figures. Check that p values agree with their intervals or test statistics. Present every mismatch in a table with its location, stated value, correct value and cause. Trace the errors to their causes and explain which would have affected conclusions. Describe changes that will prevent further drift. Report any independent check by someone else, and the time the audit took.

Inside the DBAX 9103 Week 8 example

The audit opens with Nuijten and colleagues on reporting errors. A section describes listing 94 numbers and checking each, including intervals against p values. A table lists seven mismatches: sample size and estimate in the abstract, two percentages, the check-in mean, the supervisor count and a paraphrase. A section traces five errors to the window change. A section confirms consistency of all six p values, citing Cumming. A section describes generating every number from the script and writing the abstract last, citing Appelbaum and colleagues. A section reports a classmate tracing twenty numbers. A closing section reports six hours spent. The mismatch table gives the cause of each error in a few words, which makes the pattern behind them easy to see.

Where the marks sit in the DBAX 9103 Week 8 rubric

Audits are judged on completeness and on finding causes. Listing every number and comparing it with output carries the most weight. A mismatch table with causes shows diagnosis rather than correction alone. Checking p values against intervals demonstrates statistical understanding. Identifying which errors could have changed conclusions shows judgment. Preventive changes, such as generating numbers from scripts, show learning. An independent trace by another person shows rigor. Audits that check only tables, correct errors without causes or skip derived figures do not meet the course standard. Reporting how long the audit took, and why, helps classmates plan their own. Treating a paraphrase that could mislead as an error, even when the number is close, shows a high standard for accuracy.

DBAX 9103 Week 8 help: mistakes that cost marks

Make a spreadsheet with one row per number in your chapter, including the abstract and captions. Add a column for the current output value and mark every difference. For each mismatch, ask when it entered the text; most trace to a single change you made earlier. Recompute percentages from counts and check that p values match intervals. Then fix the cause, not only the number, so it cannot drift again. Many students check tables carefully and forget the abstract, which is where old numbers survive longest. Others correct errors silently. Have a classmate trace a random sample. Ask the desk if you want an audit sheet built around your own chapter's figures.

Get a DBAX 9103 Week 8 example written to your instructions

Attach the Week 8 prompt and rubric and say how long your results chapter runs and how many displays it holds. An audit table with internal checks and a corrections log is ready within 24-48h, with no fee on the first. It audits a stand-in chapter, since checking real numbers means comparing them with output that stays in your hands.

More DBAX 9103 papers and DBA samples

DBAX 9103 Week 8 questions, answered

Where can I find a free DBAX 9103 Week 8 sample paper?

Scroll down: the number audit and its mismatch table sit lower on this page, free for every visitor.

How many numbers were audited in DBAX 9103 Week 8?

94, of which seven had drifted.

What caused most of the errors?

Shortening the window for counting check-ins from a month to the first two weeks.

Which error was most serious?

The abstract reported the biased 30-day estimate as the main finding.

How will drift be prevented?

Every number will come from the analysis script, and the abstract will be written last.