Chapter 30 Quiz: HR Analytics and People Data

Instructions: Answer all 20 questions. For multiple-choice questions, choose the single best answer. For True/False, write True or False. For short-answer questions, write 2–4 complete sentences. Attempt all questions before consulting the answer key.

Section A: Multiple Choice (Questions 1–10)

Question 1

A company has 400 employees at the start of 2023 and 420 at the end. During 2023, 78 employees left (all separations combined). What is the approximate annual turnover rate?

A) 18.1%, using starting headcount B) 19.0%, using average headcount C) 19.5%, using ending headcount D) Both A and B are technically valid depending on which denominator convention you use

Question 2

You are building a salary analysis report for the VP of HR. A department has 4 employees in the "Senior IC" job level. What is the correct action?

A) Report the 4 salaries anonymized with letter codes (A, B, C, D) B) Suppress or combine the group with an adjacent level before reporting, as 4 is below the minimum group size C) Report the median and mean but hide individual salaries D) Report all metrics since the group size is only slightly below 5

Question 3

What does a compa-ratio of 0.79 indicate for a group of employees?

A) Their average salary is 79% of the company-wide salary B) Their median salary is 21% below their pay band midpoint — potentially underpaid relative to the band C) 79% of the group is paid at or above the median for their level D) The group's turnover risk is 79% higher than average

Question 4

Priya finds that 46% of South warehouse 2023 separations occurred in the 0–6 month tenure bucket. What does this most strongly suggest?

A) The pay in that role is below market rate B) There is a problem with early-tenure experience — onboarding, management, or unmet expectations C) The physical demands of warehouse work cause early attrition universally D) The department is undergoing a performance management program

Question 5

Which pandas method is best for computing the 25th and 75th percentile of salaries within a grouped DataFrame?

A) df.groupby("department")["salary"].mean() B) df.groupby("department")["salary"].agg(p25=lambda x: x.quantile(0.25), p75=lambda x: x.quantile(0.75)) C) df.groupby("department")["salary"].describe() D) df.pivot_table(index="department", values="salary", aggfunc="median")

Question 6

A company's headcount projection assumes 5% annual growth and 18% annual attrition. Starting from 250 employees, which best represents the number of new hires needed in the first quarter?

A) About 5 (5% of 250 divided by 4) B) About 11 (5% growth + 18% attrition, divided by 4, applied to 250) C) About 45 (total annual need) D) About 14 (only the attrition replacements, ignoring growth)

Question 7

Why is the median preferred over the mean for compensation analysis?

A) Median is always larger than mean, making the company look more competitive B) Median is required by law in compensation reporting C) Compensation distributions are right-skewed by high earners; median is less sensitive to those outliers D) Mean requires a larger sample size to be statistically valid

Question 8

What does FTE (Full-Time Equivalent) measure that simple headcount does not?

A) The legal employment status of workers B) A standardized workforce size that normalizes part-time employees to a full-time equivalent unit C) The total cost of workforce including benefits D) The number of employees available on any given day

Question 9

A seasonal turnover index of 145 for July means:

A) 145 employees left in July B) July saw 45% more separations than the average month C) 14.5% of the workforce turned over in July D) July's turnover is 145 days above average

Question 10

You want to identify which departments have the highest early-tenure attrition risk. Which combination of analyses is most appropriate?

A) Company-wide turnover rate + department headcount B) Department turnover rate + tenure-at-separation distribution by department C) Compa-ratio analysis + absence rate by department D) FTE count + median tenure by department

Section B: True or False (Questions 11–15)

Question 11

A company with 1,000 employees and a 14% turnover rate has lower retention risk than a company with 50 employees and a 22% turnover rate.

Question 12

In HR analytics, it is acceptable to report a department's median salary when the department has only 3 employees, as long as you round to the nearest $5,000.

Question 13

Voluntary turnover and involuntary turnover require different organizational responses: voluntary turnover points to culture, management, or compensation issues, while involuntary turnover reflects performance management decisions.

Question 14

A compa-ratio above 1.15 always indicates that an employee or group is overpaid and should have their salary reduced.

Question 15

The absence rate formula (total absence days / (headcount × working days) × 100) produces a number that is always below 100% in a healthy organization.

Section C: Short Answer (Questions 16–20)

Question 16

Explain the minimum group size principle for HR reporting. What number does this chapter recommend, and why does the specific number matter? What should you do when a group falls below the minimum — specifically, what are two acceptable approaches and one approach that should be avoided?

Question 17

A headcount projection shows that Acme Corp needs to hire 58 people in the next 12 months to maintain and grow its workforce. The VP of HR says, "That's more than one a week. We need to dramatically improve our recruiting capacity." What assumptions underlie the projection, and how should those assumptions be communicated when presenting this number to leadership?

Question 18

Describe two different patterns that might appear in a tenure-at-separation analysis, and explain what organizational problem each pattern might indicate: 1. A pattern where the 0–6 month bucket is by far the largest (45%+ of separations) 2. A pattern where the 5+ year bucket is unexpectedly large (30%+ of separations)

Question 19

In Section 30.5, the chapter recommends presenting pay equity analysis at the job level (e.g., median salary for "Senior IC" by demographic group) rather than as a company-wide average gap. Explain in your own words why a raw company-wide average gap between demographic groups is potentially misleading, and what would need to be controlled for to make a meaningful pay equity comparison.

Question 20

Priya's South warehouse analysis found: 34% turnover rate (vs. 26% company warehouse average), 46% of separations in 0–6 month tenure bucket, July–September seasonal spike, and 71% voluntary resignation rate. She recommended an enhanced onboarding program and a 90-day retention bonus. A colleague questions this, saying: "Those are just correlations — you don't know the actual cause." How would you respond to this critique? What additional data would strengthen Priya's causal argument, and what limitations should she acknowledge even with that additional data?

Answer Key

Section A: Multiple Choice

Answer 1: D Both A and B are technically defensible. The most common convention is to use average headcount (start + end / 2 = 410) as the denominator: 78 / 410 = 19.0%. Using starting headcount (78 / 400 = 19.5%) or ending headcount (78 / 420 = 18.6%) are less common but not wrong. The chapter notes that average headcount is the more rigorous approach because it accounts for the fact that headcount changes throughout the year. In practice, many organizations use a single snapshot for simplicity.

Answer 2: B Suppress or combine the group. With 4 people in a level, reporting median salary narrows down which individual earns what — especially if the group knows who is in it. The correct action is to combine "Senior IC" with an adjacent level (e.g., aggregate with "Individual Contributor" for reporting purposes) or suppress the cell entirely. Option A (anonymized codes) still allows people to triangulate salaries if they know who the four people are. Option D is wrong — the minimum is a firm threshold, not a suggestion.

Answer 3: B Compa-ratio = employee salary / pay band midpoint. A ratio of 0.79 means the group's median is 79% of the midpoint, or 21% below it. This falls clearly in the "potentially underpaid relative to the band" range (below 0.85). It does not directly mean the pay is below market — the pay band itself might be set correctly or incorrectly — but it flags that the group is below where the organization intended to pay them.

Answer 4: B When nearly half of separations happen in the first 6 months, the problem is almost certainly with the early employee experience: poor onboarding, unmet job expectations, management quality, or simply misalignment between what was promised and what the job actually involves. Option C is tempting but wrong — the analysis shows South warehouse has a higher early-tenure rate than other warehouse locations, not just an industry-wide pattern.

Answer 5: B The named aggregation syntax with lambda functions is the correct approach for computing multiple percentiles in a single groupby.agg() call. Option C (describe()) does compute percentiles but returns a wider format that is harder to work with in subsequent analysis. Options A and D do not compute percentiles.

Answer 6: B Quarterly growth need = 250 × 5% / 4 = 3.125 ≈ 3. Quarterly attrition replacements = 250 × 18% / 4 = 11.25 ≈ 11. Total quarterly hires needed ≈ 3 + 11 = 14. The closest answer is B (about 11), which represents just the attrition component. A more precise answer is 14 total, but among the options given, B is the most defensible since it correctly captures the scale. Note: in a real headcount projection, both components are included.

Answer 7: C Compensation distributions are right-skewed because a small number of very highly paid executives and senior employees pull the mean upward. If a 1,000-person company has 5 executives earning $2M each, those 5 salaries add $10M to the total payroll and significantly inflate the mean, making it unrepresentative of what the median employee earns. Median is not affected by these outliers.

Answer 8: B FTE standardizes mixed workforces. A company with 100 full-time and 50 part-time employees does not have 150 full units of workforce capacity — the part-timers contribute less capacity. FTE calculation converts part-timers to equivalent full-time units: if part-timers work half the hours, 50 part-timers = 25 FTE, giving a total of 125 FTE from 150 headcount.

Answer 9: B A seasonal index is calculated as (actual count / average monthly count) × 100. An index of 145 means that month had 45% more separations than the average month (an index of 100 = exactly average). It is a relative measure, not an absolute count.

Answer 10: B Department turnover rate tells you which departments are losing people fastest. Tenure-at-separation distribution by department tells you when in employees' tenure those departures are happening. Together, they answer "which departments have early-tenure risk specifically." Options A and C are missing the tenure-at-separation component, which is essential for the "early-tenure" focus.

Section B: True or False

Answer 11: False Turnover rate (percentage) is what matters, not absolute headcount. The first company's 14% rate means it loses 140 employees per year — more absolute separations but proportionally lower risk than the second company's 22% rate. The company with 22% turnover is losing nearly 1 in 4 employees annually, which is more severe from a retention perspective regardless of company size. The statement conflates absolute numbers with rates.

Answer 12: False Rounding does not solve the identification problem. With only 3 employees in a department, anyone who knows approximately what two of them earn can deduce the third even from a rounded median. The minimum group size principle exists precisely because small groups are identifiable even with statistical measures. The correct action is to suppress the cell or combine the group with adjacent data.

Answer 13: True This is exactly the distinction the chapter makes. Voluntary resignations signal that employees are choosing to leave — indicating pull factors (better opportunities elsewhere) or push factors (poor culture, management, compensation). Organizations respond by addressing root causes, improving culture, adjusting pay, or improving management. Involuntary terminations reflect the company's decisions about performance or restructuring — they require different responses like reviewing performance management processes or handling restructuring thoughtfully.

Answer 14: False A compa-ratio above 1.15 indicates above-midpoint pay, but it does not automatically mean the pay should be reduced. An above-band employee might be a top performer whose contribution justifies the pay premium, a long-tenured employee who has grown beyond their job level classification (suggesting a promotion may be overdue), or a critical-skill employee in a competitive market. The compa-ratio flags a situation that deserves review — not a decision to reduce pay.

Answer 15: True The absence rate formula produces a percentage of scheduled time lost. Even in highly disruptive circumstances, an absence rate approaching 100% would mean employees are almost never present — which would mean the business has effectively ceased to function. In normal organizations, rates typically range from 1–5%. For the rate to approach 100%, virtually every employee would have to be absent for every working day, which is not a realistic scenario for an operating business.

Section C: Short Answer

Answer 16 The minimum group size principle means that HR statistics (especially compensation statistics) should only be published for groups containing at least a specified minimum number of employees. This chapter recommends a minimum of 5 employees.

The specific number matters because with fewer people, statistical measures like median salary can allow indirect identification: if you know that a department has only 3 people and their median salary is $72,000, you can often narrow down which person earned which salary especially if you know the other two people's approximate range.

Two acceptable approaches for small groups: (1) Combine the small group with an adjacent group (e.g., aggregate "Senior IC" with "Individual Contributor" for a combined salary summary), making the combined group large enough to report. (2) Suppress the cell entirely — display it as "N/A" or "-" with a note explaining the suppression reason.

One approach to avoid: reporting the data anyway but rounding or masking individual values. Rounding to the nearest $5,000 or showing a range instead of a point estimate does not adequately protect small groups because inference is still possible from context.

Answer 17 The headcount projection assumes constant rates: specifically, that the 5% growth rate and 18% attrition rate will be maintained uniformly throughout the year. This is almost never true in practice — most organizations have seasonal hiring patterns, planned restructuring events, or project-based workforce changes that create uneven quarterly distributions. The model also assumes that all attrition is replaceable at the same pace, which may not hold for specialized roles.

When presenting the projection to leadership, the communication should include: (1) the explicit assumptions — "This assumes 5% growth (approved in our plan) and 18% attrition (based on 2022–2023 actuals)"; (2) a range rather than a point estimate — "The actual need may range from 48 to 70 hires depending on attrition rate"; (3) a note that this is a projection, not a forecast — "This is a planning tool, not a prediction. We will update it quarterly as actual data comes in."

Answer 18 Pattern 1 (0–6 month bucket dominant): A very high proportion of separations in the first 6 months indicates that new employees are not successfully integrating with the organization. Possible causes include: (a) poor onboarding — new employees feel unsupported, unclear about their role, or not integrated into the team; (b) unmet expectations — the job is materially different from what was described in recruiting; (c) management quality in the department — a manager who struggles with new employees particularly in the critical first weeks; (d) compensation that looks competitive on the offer letter but is not when compared to competing employers who are actively recruiting.

Pattern 2 (5+ year bucket unexpectedly large): A disproportionate number of separations among long-tenured employees is unusual and concerning. This could indicate: (a) a wave of retirements among a cohort of employees hired at a similar time (a "retirement cliff"); (b) a cultural shift or leadership change that is making long-tenured employees — who joined under a different company culture — feel misaligned; (c) a restructuring or layoff affecting more experienced (and therefore higher-compensated) employees; (d) poaching by competitors who are specifically targeting experienced employees. This pattern poses a serious institutional knowledge risk because the exiting employees carry deep organizational and customer knowledge.

Answer 19 A company-wide average pay gap between demographic groups (e.g., Group A earns 12% more than Group B on average) does not by itself indicate pay inequity, because it does not control for the many factors that legitimately influence salary. These include job level or grade (if one group is concentrated at more senior levels due to historical hiring patterns, their average will be higher even with identical pay for identical work), years of experience and tenure (which affect pay progression under most compensation systems), geographic location (employees in high cost-of-living markets earn more), and functional role (technical roles typically command higher market pay than administrative roles).

A meaningful pay equity comparison controls for these factors, comparing employees in the same job level, at similar tenure, in the same location, performing similar work. When those controls are applied, a remaining gap is more meaningfully attributable to pay inequity rather than job mix. Without those controls, the company-wide average gap is a starting point for investigation, not a conclusion.

Answer 20 The critique that "correlation is not causation" is technically correct but not very useful at the operational decision level for HR. Priya can respond: the combination of four independent patterns pointing in the same direction — elevated turnover, concentrated in early tenure, with a seasonal spike, and predominantly voluntary — creates a coherent hypothesis that the early-tenure experience is the root cause. No single data point makes the case; the convergence of multiple signals makes it credible.

Additional data that would strengthen the argument includes: exit interview data (if available) showing what departing South warehouse employees said were their reasons; onboarding satisfaction surveys from new hires (comparing South to other locations); comparison of the South warehouse manager's tenure to other locations' managers (newer managers correlate nationally with higher team turnover); customer service metrics or productivity data from the first 90 days to see if new hires in South are struggling at a higher rate.

Limitations Priya should acknowledge even with that additional data: the synthetic data does not control for differences in the local labor market in the South region; there may be facility-specific factors (equipment age, safety issues, physical conditions) that drive attrition independently of onboarding; and any intervention (onboarding program, retention bonus) should be monitored for effectiveness rather than assumed to have worked. The analysis identifies where to act — it does not guarantee the recommended action will succeed.