| Course | GB815 Applied Research Topics |
|---|---|
| Unit | Unit 5 |
| Paper type | Data source audit |
| Length | About 870 words, 6 pages double-spaced |
| Format | APA 7 student paper |
| School | Purdue University Global |
| Program | MBA |
| Updated | October 2026 |
Free sample paper for GB815 Unit 5
Six Sources, One Question: A Unit 5 Data Source Audit for a Study of Franchise Loan Outcomes
[Student Name]
Purdue University Global
GB815: Applied Research Topics
Unit 5 Data Source Audit
[Instructor Name]
[Date]
The trial-download counts and Harrowgate Community Bank are composite and illustrative; the data sources and studies named are real.
The Question the Data Must Answer
For new businesses that received SBA 7(a) loans during a set window, how did the write-off rate for franchise units compare with that for similar independent firms? To answer it, the data must identify each loan, its approval date, size and industry, whether the borrower was a franchise unit and a new business, and whether and when the loan was charged off.
Source One: SBA 7(a) Loan Data
Holder: the Small Business Administration, which publishes loan-level data on approved 7(a) loans. Content: for each loan, the borrower's name and location, approval date, gross amount, guaranteed portion, industry code, a franchise code and name where the lender reported one, a field indicating whether the business was new or existing, and the loan's current standing, paid in full or written off, with a charge-off date and amount. Access terms: public download, no registration. Cost: none. Time lag: updated periodically; loans approved a decade ago have mostly reached resolution. Main limitation: the franchise code depends on lenders' reporting, and a franchise unit whose lender left the field blank appears as independent.
A trial download for loans approved in fiscal years 2010 through 2014 to borrowers in Missouri, Kansas, Iowa and Nebraska returned 6,840 loans, of which 2,610 were flagged as new businesses. Among those, 540 carried a franchise code. A check of 100 independent-coded new-business loans against franchise system names in the borrower name field found 7 that were plainly franchise units, suggesting a misclassification rate of several percent. That rate is low enough to work with but must be reported, and a name-matching step will be added to correct obvious cases.
Source Two: State Franchise Disclosure Documents
Holder: states that register franchise offerings, some of which post filings online. Content: each franchise system's disclosure document, including tables of openings, closures, transfers and terminations by year. Access: online search for some states, written request for others. Cost: none, but staff time is heavy. Main limitation: system-level totals, not outcomes for individual units, so it cannot be linked to loans. Use: to describe the health of the franchise systems in the loan sample, for example the closure rates of systems with many SBA borrowers.
Source Three: State Business Entity Records
Holder: the secretaries of state of the four states. Content: each registered business's status, such as active, dissolved or forfeited, and filing dates. Access: public search by name; bulk data available in some states by request. Cost: small fees for bulk data in two states. Main limitation: a business can close without formally dissolving. Use: a secondary outcome, business survival, to compare with loan charge-off. Kalnins and Lafontaine (2013) showed that state administrative records, in their case Texas sales-tax permits, can track how long establishments last at a scale no survey could match.
Source Four: Census Business Dynamics Statistics
Holder: the Census Bureau. Content: counts of business births, deaths and survival rates by firm age, size, industry and state. Access: public. Main limitation: no franchise flag and no firm-level records. Use: a benchmark for overall survival rates, so the loan sample's outcomes can be placed against the broader population of new firms.
Source Five: Commercial Franchise Database
Holder: a commercial data provider, licensed to some university libraries. Content: franchise system characteristics such as founding year, unit counts and fees. Access: through the library license, if the university holds one; the candidate's library does not. Cost: a personal license would be several thousand dollars. Verdict: not needed, since disclosure documents supply the system characteristics the study requires.
Source Six: Interviews With Loan Officers
Holder: loan officers at community banks, reachable through the state SBA lenders' association. Content: how lenders classify franchise borrowers and handle charge-offs. Access: depends on introductions. Use: to interpret the loan data, particularly how the franchise code is entered, not as a primary data source.
Linking the Sources
The loan data can be linked to state entity records by borrower name and address, and doubtful pairs will be checked by hand. Franchise system names in the loan data can be linked to disclosure documents. Linking fails where borrowers operate under a trade name different from their registered name, which the trial suggested affects about one loan in ten; those cases will be matched by address where possible and otherwise excluded and counted.
The best source is free and public, and its one weak field, the franchise code, can be tested before the study depends on it.
Verdict
Source one is sufficient for the main question and has passed a trial. Sources two and three add context and a secondary outcome at modest cost. Source four provides a benchmark. Source five is not needed, and source six should be used only for interpretation. Glennon and Nigro (2005) and DeYoung et al. (2008) both worked with SBA loan records of this kind, which supports their use for research on default.
References
DeYoung, R., Glennon, D., & Nigro, P. (2008). Borrower-lender distance, credit scoring, and loan performance: Evidence from informational-opaque small business borrowers. Journal of Financial Intermediation, 17(1), 113-143. https://doi.org/10.1016/j.jfi.2007.07.002
Glennon, D., & Nigro, P. (2005). Measuring the default risk of small business loans: A survival analysis approach. Journal of Money, Credit and Banking, 37(5), 923-947. https://doi.org/10.1353/mcb.2005.0051
Kalnins, A., & Lafontaine, F. (2013). Too far away? The effect of distance to headquarters on business establishment performance. American Economic Journal: Microeconomics, 5(3), 157-179. https://doi.org/10.1257/mic.5.3.157
Reading the GB 815 Unit 5 instructions
The fifth GB815 assignment asks you to audit the data sources your leading direction would use. State the question the data must answer and the fields it requires. Then examine each source in turn, recording its holder, content, unit of record, access terms, cost, time lag, identifiers and main limitation. Run a trial on the most important source and report what it showed, including any data quality problem you can measure. Give each secondary source a specific use or a reason for excluding it. Explain how the sources would be linked and where linkage fails. Close with a verdict on whether the data can support the study. Present the audit in a single table with columns for each attribute, followed by a paragraph on each source. Seven pages is common, and published studies that used the same data strengthen the case.
How the GB 815 Unit 5 example is put together
The audit states the question and the fields it needs. The SBA loan data section lists approval date, amount, industry, franchise code, new-business flag and loan status, then reports the trial: 6,840 loans, 2,610 new businesses, 540 franchise codes and 7 misclassified independents in a check of 100, with a note on correcting them. Disclosure documents give system-level context, state entity records give a secondary survival outcome supported by Kalnins and Lafontaine, Census statistics give a benchmark, a commercial database is judged unnecessary and loan officer interviews are kept for interpretation. Linkage by name and address is explained with its failure rate, the highlighted sentence summarizes the finding, and two published default studies built on the same SBA records close the audit.
Where the points sit in the GB815 Unit 5 rubric
GB815 instructors grade the data audit on whether it would let a committee judge feasibility. Each source should be described with the attributes that decide feasibility: holder, terms, cost, lag and limitation. A trial on the main source is weighted heavily, especially when it measures a data quality problem rather than assuming one. Secondary sources earn credit when each is given a specific use or excluded with a reason. Linkage should be explained with its expected failure rate. The verdict should follow from the evidence. Citing studies that used the same data shows the sources are established for research. Clear tabular presentation with explanatory paragraphs is expected, and figures from the trial should be reported plainly.
GB815 Unit 5 help from the desk
Many GB815 data audits list datasets that sound relevant without checking their fields or access terms. Others assume key variables are accurate, though lenders, agencies and respondents enter them with varying care. Trials are rare, so feasibility rests on hope. Secondary sources are often collected without a clear use. Linkage between sources is assumed to be straightforward. Students sometimes include costly commercial data that a free source could replace. Define the fields the question needs, audit each source on the same attributes, run a trial, measure any data problem, give each source a use, explain linkage and its failures and reach a verdict. Tell us the sources you plan to use, and our writers can audit them with you.
Get a GB815 Unit 5 example written to your instructions
List what your direction needs to measure, the sources identified so far and any replies from data holders, with the Unit 5 prompt and rubric attached. In the audit prepared for you, holders and terms are recorded exactly, each source gets a trial or a described one, and the linkage problem is faced directly. First sample free, 24-48h. The paper above is an original model document written by our desk, not a submitted student paper and not an official Purdue University Global document.
More GB815 papers and MBA samples
- GB815 Unit 1: Discussion Board Post
- GB815 Unit 2: Direction Feasibility Memo
- GB815 Unit 3: Evidence Mapping Exercise
- GB815 Unit 4: Study Quality Appraisal
- GB815 Unit 6: Seminar Reflection
- GB815 Unit 7: Scope Narrowing Exercise
- GB815 Unit 8: Practitioner Relevance Brief
- GB815 Unit 9: Comparison Matrix
- GB815 Unit 10: Research Prospectus
- GB794 Unit 5: Measurement Validity Review
- GB793 Unit 5: Levels of Analysis Exercise
- GB792 Unit 5: Capital Structure Analysis
- GB791 Unit 5: Disclosure Review
GB815 Unit 5 questions, answered
What does a data source audit add beyond a list of datasets?
It records the conditions that decide feasibility: who holds the data, what agreement or fee applies, how current the records are, what unit each record describes and how sources join. A list says where data might be; an audit says whether this candidate can obtain and use them for this question, which is what a committee needs to know.
Is publicly available data acceptable for doctoral applied research?
Often it is preferable. Public data can be obtained without negotiation, allow others to verify results and avoid confidentiality obligations. The challenge is usually fit, since public records were collected for someone else's purpose. The audit's task is showing how public sources can be made to answer the question and where they fall short.
Should the audit include interviews as a data source?
If the direction relies on them, yes. Interviews have holders too, in the sense of gatekeepers who can introduce or block access to participants. Recording who would make introductions, on what terms and with what likely response belongs in the audit, and any human-subjects review is noted as a requirement still to be met.
Where can I find a free GB815 Unit 5 sample paper?
This page holds a full GB815 Unit 5 data source audit for a study of franchise and independent SBA loan outcomes, with a trial download and six sources rated.
Does SBA loan data identify franchise borrowers?
The SBA's public 7(a) records include a franchise code and name entered by the lender. Because lenders enter it, some franchise units may be coded as independent, so the field should be checked before use.