peopleanalyst

library / lib66a32037a0a72cf7

Statistics for Compensation

In a sentence

A practical, applied guide that teaches compensation and HR professionals the descriptive statistics and model-building techniques needed to analyze pay data and make sound, defensible decisions.

Statistics for Compensation translates the math of uncertainty into the everyday work of compensation professionals, showing how to turn raw pay and HR data into decisions that help organizations attract, retain, motivate, and align the right people. Through a fictitious company (BPD) and a series of running case studies, John Davis walks readers from basic notions of percent and compound interest through frequency distributions, measures of location and variability, and on to the heart of the book: building models (linear, exponential, maturity curve, power, and multiple regression) that relate pay to grade, experience, company size, and other factors. The book's distinctive philosophy—'behind every data point there is a story' and 'aggressive inquisitiveness'—reframes statistics as a tool to raise the right issues, not just produce numbers, and culminates in a complete annual workflow for analyzing salary surveys, determining market position, building market-based salary structures, and recommending salary increase budgets.

The story it tells the reader

The reader A compensation or HR professional responsible for pay decisions who wants to make sound, defensible recommendations that help the organization attract, retain, motivate, and align the right people.

External problem

They must analyze internal and external pay data, determine market position, build salary structures, and recommend budgets—often under time pressure and uncertainty—without a strong statistical toolkit.

Internal problem

They feel uncertain and exposed when executives say 'prove it to us using data,' worried their analyses are amateurish, incorrect, or easily challenged.

Philosophical problem

Pay decisions affecting people's livelihoods and an organization's success should be grounded in honest, transparent analysis rather than guesswork or copying the crowd.

The plan

  1. Adopt an aggressively inquisitive mindset and remember that behind every data point there is a story.
  2. Learn the basic notions—percent, percent difference, compound interest—and how to summarize data with distributions and histograms.
  3. Master measures of location and variability and the meaning of populations vs. samples.
  4. Use the five-step model-building process to relate pay to factors like grade, experience, and company size.
  5. Apply the right model (linear, exponential, maturity curve, power, multiple regression) and evaluate it with r-squared, SEE, correlation, and common sense.
  6. Run the full salary-survey workflow to determine market position, build a market-based structure, and recommend a budget, stating assumptions transparently.

Success

  • The reader confidently turns raw data into clear summaries, models, and executive-ready presentations.
  • They identify market position, build defensible salary structures, and recommend budgets backed by transparent assumptions.
  • They spot anomalies, traps (Simpson's Paradox, multicollinearity), and hidden issues, raising the right questions before decisions are made.
  • Their organization makes sounder pay decisions and engages employees through a fair value exchange.

At stake

  • Decisions rest on raw, misunderstood, or misleadingly presented numbers.
  • Recommendations are easily challenged, undermining the professional's credibility and integrity.
  • Statistical traps lead to flawed conclusions and inappropriate spending or inequities.
  • The organization pays inappropriately—too high, too low, or inequitably—harming its ability to attract and retain the people it needs.

Model of the world · 12 constructs · 13 relations

A framework in which design levers and contextual conditions (job grade, experience, company size, internal valuation, pay policy) are quantified through statistical models to produce psychological/behavioral and organizational outcomes (market position, structure fit, pay competitiveness) that drive employee attraction, retention, motivation, and alignment.

Design levers

  • Market Model (Fitted Pay-Factor Relationship)
  • Market-Based Salary Structure
  • Salary Increase Budget Recommendation
  • Internal Job Value (Grade/Experience Hierarchy)
  • Pay Policy (Desired Market Position)

Intermediate states & behaviors

  • Data Quality and Aggressive Inquisitiveness

Outcomes

  • Market Position (Pay Competitiveness)
  • Sound Decision Quality
  • Workforce Outcomes (Attract, Retain, Motivate, Align)

Moderators / context: Market Pay for Similar Jobs/Skills · Organization Size (Sales/Revenue/Budget) · Model Quality (Goodness of Fit and Sensibility)

Consolidated shape of the book’s model — full constructs and relationships below.

Internal Job Value (Grade/Experience Hierarchy)design lever

The organization's ordering of jobs by worth, expressed as grades from a job evaluation system or as years since degree in maturity-curve systems, reflecting required knowledge, skills, responsibility, and impact on goals.

Market Pay for Similar Jobs/Skillscontextual condition

The external compensation level for comparable jobs and skills obtained from salary surveys, summarized as weighted/unweighted average, median, or percentile, and aged to a common date for comparability.

Organization Size (Sales/Revenue/Budget)contextual condition

The scale of the organization measured by sales or revenue (for-profit) or budget (nonprofit), used as the primary predictor of single-incumbent executive pay in power models because both vary by orders of magnitude.

Pay Policy (Desired Market Position)design lever

The deliberate decision of where the organization wants its pay to sit relative to the market at a given point in time (e.g., at, above, or below market average or a target percentile), reflecting strategy and balance with benefits.

Market Model (Fitted Pay-Factor Relationship)design lever

The least-squares statistical model (linear, exponential, maturity curve, or power) that translates internal value or organization size into predicted average market pay, integrating external value with internal value.

Model Quality (Goodness of Fit and Sensibility)contextual condition

The degree to which a fitted model corresponds to reality, judged by appearance of the line through data, coefficient of determination, correlation, standard error of estimate, multicollinearity, simplicity, and common sense.

Market-Based Salary Structuredesign lever

A range of pay (minimum, midpoint, maximum) for each grade or job derived from the market model and a chosen range spread, used as an administrative control device to ensure consistent, competitive pay.

Market Position (Pay Competitiveness)outcome metric

The percentage by which an organization's total employee pay is above or below the market midpoint at a point in time, indicating how competitive pay currently is relative to the chosen market.

Salary Increase Budget Recommendationdesign lever

The final recommended percentage increase to payroll, built from catch-up to market, anticipated market movement, and pay policy, then adjusted by 'soft' organizational factors before approval.

Data Quality and Aggressive Inquisitivenessbehavioral pattern

The practitioner's verification, cleaning, and investigation of data—understanding the story behind each point and explaining anomalies—which conditions the validity of every downstream analysis and decision.

Sound Decision Qualityoutcome metric

The degree to which compensation and HR decisions are well-founded, transparent, defensible, and aligned with organizational strategy given inherent uncertainty.

Workforce Outcomes (Attract, Retain, Motivate, Align)outcome metric

The organization's ability to attract, retain, motivate, and align the kinds and numbers of employees it needs to achieve its goals, supported by competitive and fair pay and a sound value exchange.

How they connect

  • internal job value → predicts market model
  • market pay data → influences market model
  • organization size → predicts market model
  • data quality and inquiry → moderates market model
  • market model → predicts market based salary structure
  • model quality → moderates market based salary structure
  • market based salary structure → predicts market position
  • market position → predicts salary increase budget
  • pay policy → moderates salary increase budget
  • salary increase budget → influences decision quality
  • market position → influences decision quality
  • decision quality → influences workforce outcomes
  • market position → influences workforce outcomes

Possible measures & feedback loops

A candidate team / org survey built from this book’s model — exploratory operationalizations, not validated instruments. Where a construct maps to a validated measure in Principia, we’ll point to that instead.

Internal Job Value (Grade/Experience Hierarchy)

Grade number; Job evaluation point total; Years since BS degree (YSBS)

self-report suitability: low

Market Pay for Similar Jobs/Skills

Weighted/unweighted average pay; Median pay; Standard percentiles (P10-P90); Aging factor applied

self-report suitability: none

Organization Size (Sales/Revenue/Budget)

Annual sales/revenue; Budget (nonprofit); Log-transformed size

self-report suitability: none

Pay Policy (Desired Market Position)

Stated target percent vs market; Target percentile; Timing within plan year

self-report suitability: medium

Market Model (Fitted Pay-Factor Relationship)

Regression coefficients; Coefficient of determination (r-squared); Standard error of estimate; Midpoint progression

self-report suitability: none

Model Quality (Goodness of Fit and Sensibility)

r-squared; Correlation; Standard error of estimate; Correlation matrix among x-variables; Number of x-variables (simplicity)

self-report suitability: low

Market-Based Salary Structure

Minimum/midpoint/maximum by grade; Range spread percentage; Midpoint progression

self-report suitability: none

Market Position (Pay Competitiveness)

Percent from market midpoint; Budget needed to meet market; Position by grade/family

self-report suitability: none

Salary Increase Budget Recommendation

Catch-up/fallback percent; Anticipated market movement percent; Pay policy percent; Final recommended percent

self-report suitability: low

Data Quality and Aggressive Inquisitiveness

Proportion of outliers explained; Presence of source documentation; Assumption-check checklist completion

self-report suitability: medium

Sound Decision Quality

Recommendation acceptance rate; Documented assumptions present; Ability to answer challenge questions

self-report suitability: medium

Workforce Outcomes (Attract, Retain, Motivate, Align)

Turnover/retention rate; Recruiting cost and time-to-fill; Engagement survey scores

self-report suitability: medium

Preview the survey →

Frameworks & instruments in this book

  • Behind every data point, there is a story—investigate anomalies before reporting.
  • Be aggressively inquisitive: persistently ask why things are as they are.
  • Numbers raise issues; they rarely provide a single right answer ('it depends').
  • Models describe relationships between variables, not the variables themselves.
  • Predictions from regression are of averages, not exact values.
  • Keep the horse before the cart: plot the data first, then regress.

Several of these are operationalized as tools in the People Analytics Toolbox.

Topics

Related in the library