This drill gives you 12 original CBDA research questions practice items for Domain 1, Identify the Research Questions. The set covers how a business problem becomes research questions, who counts as a stakeholder from a data point of view, how to reshape and reframe a vague problem, how hypothesis testing sharpens a hunch, and how the future state defines success. Each answer cites a Guide section.
How this CBDA research questions set is built
The set targets Domain 1 of the CBDA blueprint, worth 20% of the exam. The 12 questions are original (no exam dumps) and mirror the real exam: a short scenario, then what the analyst should do first, best or next. Two need a small calculation. Difficulty sits slightly above typical question banks; many distractors belong to a later domain.
Ideas tested from the Guide to Business Data Analytics:
- 2.1.1 Define Business Problem or Opportunity. A business problem is broader than a research question and can produce several of them. See how a business problem turns into a research question.
- 2.1.2 Identify and Understand the Stakeholders. Take a data view: who creates the data, who is exposed to it, and who acts on the insights.
- 2.1.4 Define Future State. Success criteria depend on business context.
- 2.1.5 Formulate Research Questions. Questions use plain business language and need stakeholder consensus.
- 3.15 Problem Shaping and Reframing. Shaping refines a problem until a solution process applies; reframing restates it in another context. Tools such as the 5 Whys narrow a problem, while reframing looks for a different formulation.
- 3.16 Stakeholder List, Map, or Personas. Records interests, biases and analytics literacy; see stakeholder lists and personas.
- 3.12 Hypothesis Formulation and Testing. Turns a hunch into a claim data can verify.
| Sub-topic | Questions |
|---|---|
| Business problem to research questions | 1, 7, 11 |
| Stakeholders from a data view | 2, 6 |
| Problem shaping and reframing | 3, 5, 9, 11 |
| Hypothesis formulation and testing | 4, 8, 12 |
| Future state success criteria | 10 |
CBDA exam at a glance
| Fact | Detail |
|---|---|
| Format | 75 multiple-choice, scenario-based questions in 2 hours |
| Domain 1 weight | Identify the Research Questions, 20% |
| Exam fee | US$405 (members US$250), first year of membership included |
| Passing score | Not published by IIBA |
Last verified: October 2026
Practice Questions
Question 1
After interviews, the Harbour Grocers analyst recorded the business problem: repeat online orders fell 18% in two quarters. The sponsor wants data extraction to start with that statement as the brief.
What should the analyst do NEXT?
- A. Build a dashboard that tracks the decline by week
- B. Have the data team profile the order data quality
- C. Derive research questions with measurable success standards
- D. Hold a workshop to confirm the decline is a real issue
Question 2
Northfield Health is studying rising readmissions. Its stakeholder list holds the chief medical officer, ward managers and the quality director, who will act on findings. (Choose two.)
Which two groups should the analyst add FIRST to get a data view of stakeholders?
- A. Ward clerks who enter discharge dates and diagnosis codes
- B. The press office that handles media enquiries
- C. Facilities staff who maintain ward equipment
- D. The data warehouse team that maintains readmission tables
- E. Finance auditors who review the annual accounts
Question 3
Velora Travel's commercial director asks the team to "find our best customers" for a premium campaign. Marketing means customers who spend most per trip; finance means customers who stay for many years.
What is the BEST action for the analyst?
- A. Rank customers by last year's spend, the easiest measure
- B. Reframe the request with stakeholders to agree what "best" means
- C. Request a churn model that compares both views
- D. Ask the sponsor to choose between the marketing and finance views
Question 4
Copperline Logistics funds a new dock only if average loading time exceeds 40 minutes. The analyst runs a lower-tailed z-test of H0: mean is at least 40 minutes (5% significance, critical value -1.65).
| Sample size | Sample mean | Sample standard deviation |
|---|---|---|
| 49 trucks | 38.6 minutes | 4.2 minutes |
What should the analyst conclude?
- A. Fail to reject H0, because z of about -0.33 is inside the range
- B. Fail to reject H0, because the mean is within one standard deviation
- C. Reject H0, because z of about -2.33 supports the funding case
- D. Reject H0, because z of about -2.33 is below the critical value
Question 5
A Brightwell Energy analyst reframed "too many billing calls" into a draft problem statement. It describes the problem, lists affected groups, and outlines a diagnostic approach using call and billing data.
What should the analyst add to complete the reframed problem statement?
- A. A comparison with an alternative formulation of the problem
- B. A detailed schedule for the data collection activities
- C. A list of charts the final dashboard will contain
- D. The names of the data scientists who will build models
Question 6
Tamsin Insurance plans a workshop on slow claims. Operations wants faster payouts, finance fears leakage, the claims director blames an old system, and several attendees are new to analytics.
What is MOST useful to capture in stakeholder personas before the workshop?
- A. Preferred dashboard layouts and chart types
- B. Data volumes each department stores
- C. Vested interests, biases and analytics knowledge
- D. Budgets approved for the next financial year
Question 7
For Lindqvist Retail's stockout problem, a data scientist drafts: "Fit a gradient-boosted regression on SKU-store-week features to minimise forecast RMSE." The operations director cannot tell what it answers.
What should the analyst do BEST?
- A. Keep the draft and add a glossary of technical terms
- B. Restate it in plain business language and confirm it with stakeholders
- C. Replace it with a list of metrics the stores already report
- D. Ask the director to take a short statistics course first
Question 8
Pelican Telecom marketers believe a new onboarding email raises usage. After sampling and getting a p-value of 0.07, they propose a 10% significance level so the result counts.
What is the MAIN problem with this approach?
- A. A t-test cannot compare average usage between two groups
- B. The sample should exclude customers who received the email
- C. A p-value of 0.07 proves the email has no effect on usage
- D. The significance level should be set before collecting data
Want more than 12? Create a free account on certificate.tips to unlock more CBDA practice questions, including the full 75-question CBDA exam simulator with exam and practice modes. certificate.tips is our product.
Question 9
Orchard Bank's head of lending reshaped a falling-approvals problem alone and wants analysis to start. Risk and sales managers, not consulted, see it differently.
What is the BEST way to reduce the risk in this reframing?
- A. Run a brainstorming workshop to reframe it with all three groups
- B. Proceed with her version, since she owns the initiative's budget
- C. Have the data team test every framing in parallel
- D. Log her version as an assumption and revisit it after the analysis
Question 10
Fenwick Mutual defines fraud-model success as lowest total cost: a missed fraud costs $4,000, a false alarm $150 in review. Two models were tested on 10,000 claims.
| Model | True positives | False positives | False negatives | True negatives |
|---|---|---|---|---|
| A | 80 | 40 | 20 | 9,860 |
| B | 92 | 160 | 8 | 9,740 |
Which model better meets the success criterion?
- A. Model A, because its precision of about 67% is much higher
- B. Model A, because its accuracy of 99.4% beats Model B's 98.3%
- C. Model B, because its total cost is about $56,000 versus $86,000
- D. Model B, because it raises more alerts for the review team
Question 11
Kestrel Furniture asked for analytics on late deliveries. Reframing from the customer's view, the analyst finds most late orders were placed after a 2 p.m. cut-off the website never mentions.
What should the analyst do?
- A. Continue with the predictive model the sponsor approved
- B. Present the non-analytics fix to stakeholders as a valid option
- C. Record the cut-off finding as out of scope in lessons learned
- D. Limit the research question to orders placed before the cut-off
Question 12
At Sundial Foods, a test at 5% significance showed higher average sales with a new shelf layout. The draft summary says: "The test proves the layout will raise sales in every store."
How should the analyst BEST revise the summary?
- A. Remove the test and show only raw sales by store
- B. Add the test statistic and formula for executives to check
- C. Replace "proves" with "suggests" but keep "every store"
- D. State the result with its significance level, without certainty
Ready for a full run? Create a free account to try more questions in the CBDA exam simulator: a full 75-question set with exam and practice modes.
Answer Key and Explanations
1. Correct answer: C. A business problem is a higher-level description than a research question. The Guide says it may lead to one or more research questions, refined until each has a measurable success standard. Source: Guide to Business Data Analytics, 2.1.1 Define Business Problem or Opportunity Why the others are wrong: A reports the symptom only; B belongs to Source Data; D re-checks an already evidenced problem.
2. Correct answers: A and D. The Guide's data view asks who creates the data, who is exposed to it, and who decides on it. Clerks create the data; the warehouse team holds it. Source: Guide to Business Data Analytics, 2.1.2 Identify and Understand the Stakeholders Why the others are wrong: B, C and E neither create, hold nor use the data.
3. Correct answer: B. The Guide's reframing example: "high-value" customers may mean lifetime value, highest spend, or interest in high-value purchases. Agree the term first. Source: Guide to Business Data Analytics, 2.1.7 Select Techniques for Identify the Research Questions Why the others are wrong: A ignores the conflict; C jumps to a solution; D skips the analyst's facilitation role.
4. Correct answer: D. Standard error = 4.2 / √49 = 0.6, so z = (38.6 − 40) / 0.6 ≈ −2.33, below −1.65, so H0 is rejected. As in the Guide's bank example, the test challenges an experience-based belief. Source: Guide to Business Data Analytics, 3.12 Hypothesis Formulation and Testing Why the others are wrong: A omits the standard error; B uses the raw deviation; C misreads rejection as support.
5. Correct answer: A. The Guide lists four parts: problem statement, impacted stakeholders, high-level solution or key needs, and a comparison to an alternative formulation. The draft lacks only the comparison. Source: Guide to Business Data Analytics, 3.15 Problem Shaping and Reframing Why the others are wrong: B is planning; C belongs to reporting; D does not describe the problem.
6. Correct answer: C. In Domain 1, stakeholder analysis captures vested interests, inherent biases, current-state views and analytics knowledge. This helps facilitate conflicting needs. Source: Guide to Business Data Analytics, 3.16 Stakeholder List, Map, or Personas Why the others are wrong: A matters at reporting; B describes systems; D does not help facilitation.
7. Correct answer: B. Good analytics questions avoid technical language and reach stakeholder consensus. The mathematical restatement comes later, in Perform Data Analysis. Source: Guide to Business Data Analytics, 2.1.5 Formulate Research Questions Why the others are wrong: A keeps the jargon; C is not a question; D shifts the burden to stakeholders.
8. Correct answer: D. The Guide's process sets the significance level and decision rule before collecting samples. Moving the threshold after seeing the p-value bends the rule to fit the result. Source: Guide to Business Data Analytics, 3.12 Hypothesis Formulation and Testing Why the others are wrong: A is false, the t-test is a named variant; B misses the main flaw; C confuses non-significance with no effect.
9. Correct answer: A. The Guide warns that reframing can carry an influential stakeholder's personal bias. It recommends a group exercise such as brainstorming or a stakeholder workshop. Source: Guide to Business Data Analytics, 3.15 Problem Shaping and Reframing Why the others are wrong: B keeps the bias; C lacks stakeholder input; D risks solving the wrong problem.
10. Correct answer: C. Cost A = 20 × $4,000 + 40 × $150 = $86,000; cost B = 8 × $4,000 + 160 × $150 = $56,000. The Guide notes that when success is actual cost, precision is not the right metric. Source: Guide to Business Data Analytics, 2.1.4 Define Future State Why the others are wrong: A ignores missed fraud; B is inflated by rare fraud; D is not the agreed criterion.
11. Correct answer: B. The Guide lists uncovering simpler non-analytics solutions as a strength of reframing; not every business need leads to analytics. Source: Guide to Business Data Analytics, 3.15 Problem Shaping and Reframing Why the others are wrong: A ignores a cheaper fix; C drops the core cause; D hides it.
12. Correct answer: D. The Guide says test results are probabilistic and must be stated with confidence intervals and significance levels. It also advises simple examples when explaining tests to stakeholders. Source: Guide to Business Data Analytics, 3.12 Hypothesis Formulation and Testing Why the others are wrong: A discards evidence; B suits analysts, not executives; C still overclaims.
How to score this set
One point per question; Question 2 counts only if both answers are right.
- 11–12: move on to mixed sets across domains.
- 8–10: review misses by sub-topic in the table above.
- Below 8: reread Guide sections 2.1.1–2.1.5, 3.12, 3.15 and 3.16.
For a timed run across all domains, create a free account and use the CBDA exam simulator.
What candidates say about research questions on the CBDA
The certificate.tips author who passed, in our CBDA certification overview, flagged about 20 of 75 questions and found the exam harder than question banks, with more statistical reasoning than definitions. Other candidates stress method: Mufaddal Haidermota recommends per-domain mock questions, full-length tests and mind maps. LN Mishra of Adaptive US advises daily practice and reviewing every wrong answer, and Techcanvass's candidate "Ivan" credits mind maps and two to three hours of daily study.
FAQ
What does Domain 1 of the CBDA exam cover?
It covers defining the business problem, understanding stakeholders, current and future state, formulating research questions and planning the analytics approach. It carries 20% of the exam.
What is the difference between a business problem and a research question?
A business problem is broad, such as falling repeat orders. A research question is narrower, plain-language and answerable with data against a measurable standard. One problem can yield several.
Do I need to calculate statistics on the CBDA exam?
The Guide treats framing hypotheses and interpreting results as the analyst's core skill; the data team runs tests. A quick z statistic still helps check an answer.
How is problem reframing different from root cause analysis?
The 5 Whys narrows a problem to its causes. Reframing restates it in another context and can reveal simpler, even non-analytics, fixes.
Sources
- Certification in Business Data Analytics (CBDA) — IIBA, checked October 2026
- Certification Fees — IIBA, checked October 2026
- CBDA Certification Handbook — IIBA, checked October 2026
- 7.1.3 What are statistical tests? — NIST/SEMATECH e-Handbook of Statistical Methods, accessed October 2026
- Business Data Analytics Certification — Mufaddal Haidermota on Medium, January 2023
- CBDA Certification Journey: Ivan — Techcanvass, 2024
- Mastering the CBDA Exam — LN Mishra, Adaptive US, October 2024