Insurance Research & Evidence-Based Practice 4 — Questions and Answers
Question 1: An actuary applies the Bornhuetter-Ferguson method to estimate ultimate losses. What two inputs does this method blend?
- Market share and combined ratio
- A priori expected losses and development-pattern-based projections (Correct answer)
- Reinsurance ceded and net retained losses
- Premium volume and expense ratio
Correct answer: A priori expected losses and development-pattern-based projections
The Bornhuetter-Ferguson method combines a prior estimate of ultimate losses with emerging loss development data for a credibility-weighted projection.
Question 2: What is 'moral hazard' in the context of insurance research and behavioral economics?
- The tendency of insurers to act unethically when setting rates
- The increased risk-taking behavior that occurs because a party is protected by insurance (Correct answer)
- A systematic bias introduced when researchers have conflicts of interest
- Fraudulent misrepresentation on an application
Correct answer: The increased risk-taking behavior that occurs because a party is protected by insurance
Moral hazard describes how insurance coverage can reduce incentives to prevent losses, since the insured bears less of the financial consequence.
Question 3: In a systematic literature review for insurance topics, what is the primary purpose of the PRISMA checklist?
- To standardize premium calculation across jurisdictions
- To provide transparent reporting standards for systematic reviews and meta-analyses (Correct answer)
- To rank insurers by financial strength
- To outline regulatory filing requirements
Correct answer: To provide transparent reporting standards for systematic reviews and meta-analyses
PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) ensures comprehensive and transparent documentation of the review process and findings.
Question 4: What does 'adverse selection' mean in insurance underwriting research?
- An insurer rejecting all high-risk applicants
- Higher-risk individuals disproportionately seeking or retaining insurance coverage (Correct answer)
- An adjuster selecting adverse facts to deny claims
- A reinsurer choosing unfavorable treaty terms
Correct answer: Higher-risk individuals disproportionately seeking or retaining insurance coverage
Adverse selection occurs when people with higher-than-average risk are more likely to purchase or maintain insurance, distorting the insured pool.
Question 5: A researcher computes Cronbach's alpha for a multi-item insurance knowledge test. What does a high alpha (e.g., 0.85) indicate?
- The test items are highly correlated and measure a consistent underlying construct (Correct answer)
- The sample size is large enough for generalization
- The test predicts future claim behavior accurately
- Respondents answered at an above-average rate
Correct answer: The test items are highly correlated and measure a consistent underlying construct
Cronbach's alpha measures internal consistency reliability; a value near 1.0 indicates that items on the scale are strongly interrelated.
Question 6: Which research methodology would be most appropriate for exploring WHY policyholders decide to lapse their life insurance policies?
- Large-scale quantitative regression analysis of lapse rates
- In-depth qualitative interviews with former policyholders (Correct answer)
- Randomized controlled trial comparing two policy designs
- Actuarial loss development triangles
Correct answer: In-depth qualitative interviews with former policyholders
Qualitative interviews capture nuanced motivations, attitudes, and experiences that quantitative methods cannot easily quantify.
Question 7: What is the purpose of a 'holdout sample' (or test set) in building an insurance predictive model?
- To reserve a portion of premium for regulatory surplus requirements
- To evaluate model performance on data not used during training (Correct answer)
- To hold claims payments while litigation is pending
- To set aside questionable policies for manual underwriting review
Correct answer: To evaluate model performance on data not used during training
A holdout sample allows unbiased assessment of how well the model generalizes to new, unseen data after training.
Insurance Research & Evidence-Based Practice 4
An actuary applies the Bornhuetter-Ferguson method to estimate ultimate losses.
What two inputs does this method blend?