ARTICLES HR ANALYTICS
#1 Edwards & Edwards (2019) 2
Chapter 1: Understanding HR analytics 2
Chapter 2: Analytical foundations of HR measurement 3
Chapter 3: Analysis strategies 4
Chapter 4: Case study 1 Diversity analytics 5
Chapter 12: Reflection on HR analytics - usage, ethics and limitations 8
#2 Khan & Millner (2020) 10
Chapter 1: Redefining HR 10
Chapter 2: The age of data and people analytics 12
Chapter 5: Working with data 13
Chapter 6: A people analytics framework - laying the foundation through metrics,
reporting, core analytical activities 15
Chapter 7: A people analytical framework - identifying business insights from analytics16
Chapter 8: Delivering people analytics projects 17
Chapter 10: The road ahead 20
#13 Peeters et al. (2020) 23
#9 Levenson & Fink (2017) 24
#3 Cascio et. al (2019) 26
Chapter 1: HR measurements make investing in people more strategic 26
Chapter 2: Analytical foundations of HR measurement 27
#8 Heuvel & Bondarouk (2017) 31
#6 Angrave et al. (2016) 33
#5 Charmorro-Premuzic et al. (2019) 35
#7 Rasmussen & Ulrich (2015) 37
#10 C.I.P.D. (2018) 39
#11 Henke et al. (2018) 40
#12 Van der Laken (2018) 41
1
,#1 Edwards & Edwards (2019)
Chapter 1: Understanding HR analytics
● Different meanings of “predictive”
○ Predictors/drivers: identifying causal factors that explain variation (e.g., why
performance or turnover differs).
○ Predictive modelling: applying models to estimate what happens if we change
or adjust these drivers.
○ Forecasting: using models to predict how current or future employees/teams
will behave.
● Data quality & Big Data challenge
○ Success of predictive HR analytics depends entirely on good-quality people
data.
○ The issue is often not too little, but too much data (“big data”).
○ Key HR data: skills, qualifications, competencies, training, engagement,
customer satisfaction, performance reviews, pay/bonus.
● Competence gap in HR
○ Few HR professionals have the statistical skills to perform predictive
analytics.
○ HR training (e.g., CIPD, SHRM, AHRI) often downplays quantitative/statistical
training → creating a major capability gap.
● Business case & application
○ Predictive analytics helps identify drivers of performance, retention,
engagement, etc.
○ Analytics results must always be translated into a “So what?” answer →
implications for strategy and decision-making.
● HR analytics as a strategic driver
○ Predictive models can guide HR strategy, monitor intervention impact, and
improve decisions on hiring, turnover, and performance.
○ Provides HR with a more credible, evidence-based role in the organization.
● Becoming a persuasive HR function
○ Developing analytic literacy (Huselid & Becker) is the next step for HR’s
evolution.
○ Predictive HR analytics allows HR to show evidence on: who will perform
well, who may leave, which interventions work, etc.
○ Enables “what if” scenario modelling → stronger business cases and ROI
justification.
2
,Chapter 2: Analytical foundations of HR measurement
Data sources in HR analytics
● HR databases (e.g., SAP, Oracle): personal details, performance, salary, diversity,
absence, promotions, leavers.
● Employee surveys: engagement, job satisfaction, person–organization fit.
● Customer satisfaction surveys: preferences, loyalty, ratings of
branches/employees.
● Sales data: revenue, new customers, target achievement.
Operational data: call center efficiency, onboarding times, supermarket scan rates.
● Key insight: linking data across systems reveals drivers of outcomes (e.g., L&D
training + customer satisfaction).
Software for HR analytics
● SPSS: user-friendly, menu-driven, widely used in HR, good for manipulation and
reporting.
● Minitab: very simple, menu-based, but less flexible.
● Stata: command-driven, strong for economists, more complex.
● SAS: powerful for big data, command-driven, steep learning curve.
● R: free, flexible, but requires coding skills. Excellent for advanced methods.
● JASP: new, free, easy GUI; still developing.
SPSS environment
● Data view: each row = a case (employee), each column = a variable (e.g., age,
gender), each cell = a value.
● Variable view: define name, type (numeric, string, date), labels, missing values,
measure level (nominal, ordinal, scale).
● Important: coding categorical data numerically (e.g., Gender: 0 = Female, 1 = Male)
enables richer statistical analysis.
Data preparation
● Import Excel/CSV files into SPSS, copy-paste data, define variables before analysis.
● Merge datasets with a unique key (Employee ID). Sorting by key variable ensures
successful merges.
● Recode text (e.g., “Marketing” → 1, “Sales” → 2) into numeric values for analysis.
Big Data in HR
● Defined by the 3 Vs: volume, velocity, variety.
● Database types:
○ Relational (SAP, Oracle) – structured, table-based.
○ Multidimensional (OLAP, Cognos) – fast for reporting/warehousing.
3
, ○ Object-oriented (Workday) – flexible, stores “objects” like Employee with
rules.
○ NoSQL (Hadoop) – handles massive, unstructured datasets.
● Business benefits: transparency, experimentation, segmentation, better
decision-making, fraud/error reduction.
● Challenge: shortage of people with deep analytical skills
Chapter 3: Analysis strategies
From descriptive to predictive HR analytics
● Descriptive: reports, averages, dashboards (historical).
● Predictive: identifies drivers and forecasts outcomes.
● Descriptive is useful but limited (snapshot view, no causality). Predictive adds
strategic value.
Statistical significance
● Hypothesis testing: null (H0) vs. research (H1).
● p-values: <0.05 (95% confidence), <0.01, <0.001 (stronger evidence).
● Significance ≠ importance; effect size matters.
Types of data
● Categorical:
○ Binary (Yes/No), Nominal (e.g., gender, department), Ordinal (e.g., salary
bands, performance ratings).
● Continuous:
○ Interval (e.g., temperature), Ratio (e.g., tenure, salary, height).
● Data type determines which test is appropriate.
Variables
● Independent variable (X) → predictor.
● Dependent variable (Y) → outcome.
● Example: Training (X) → Customer satisfaction (Y).
Which test to use?
● Chi-square: categorical X × categorical Y.
Example: Gender → Promotion.
4
#1 Edwards & Edwards (2019) 2
Chapter 1: Understanding HR analytics 2
Chapter 2: Analytical foundations of HR measurement 3
Chapter 3: Analysis strategies 4
Chapter 4: Case study 1 Diversity analytics 5
Chapter 12: Reflection on HR analytics - usage, ethics and limitations 8
#2 Khan & Millner (2020) 10
Chapter 1: Redefining HR 10
Chapter 2: The age of data and people analytics 12
Chapter 5: Working with data 13
Chapter 6: A people analytics framework - laying the foundation through metrics,
reporting, core analytical activities 15
Chapter 7: A people analytical framework - identifying business insights from analytics16
Chapter 8: Delivering people analytics projects 17
Chapter 10: The road ahead 20
#13 Peeters et al. (2020) 23
#9 Levenson & Fink (2017) 24
#3 Cascio et. al (2019) 26
Chapter 1: HR measurements make investing in people more strategic 26
Chapter 2: Analytical foundations of HR measurement 27
#8 Heuvel & Bondarouk (2017) 31
#6 Angrave et al. (2016) 33
#5 Charmorro-Premuzic et al. (2019) 35
#7 Rasmussen & Ulrich (2015) 37
#10 C.I.P.D. (2018) 39
#11 Henke et al. (2018) 40
#12 Van der Laken (2018) 41
1
,#1 Edwards & Edwards (2019)
Chapter 1: Understanding HR analytics
● Different meanings of “predictive”
○ Predictors/drivers: identifying causal factors that explain variation (e.g., why
performance or turnover differs).
○ Predictive modelling: applying models to estimate what happens if we change
or adjust these drivers.
○ Forecasting: using models to predict how current or future employees/teams
will behave.
● Data quality & Big Data challenge
○ Success of predictive HR analytics depends entirely on good-quality people
data.
○ The issue is often not too little, but too much data (“big data”).
○ Key HR data: skills, qualifications, competencies, training, engagement,
customer satisfaction, performance reviews, pay/bonus.
● Competence gap in HR
○ Few HR professionals have the statistical skills to perform predictive
analytics.
○ HR training (e.g., CIPD, SHRM, AHRI) often downplays quantitative/statistical
training → creating a major capability gap.
● Business case & application
○ Predictive analytics helps identify drivers of performance, retention,
engagement, etc.
○ Analytics results must always be translated into a “So what?” answer →
implications for strategy and decision-making.
● HR analytics as a strategic driver
○ Predictive models can guide HR strategy, monitor intervention impact, and
improve decisions on hiring, turnover, and performance.
○ Provides HR with a more credible, evidence-based role in the organization.
● Becoming a persuasive HR function
○ Developing analytic literacy (Huselid & Becker) is the next step for HR’s
evolution.
○ Predictive HR analytics allows HR to show evidence on: who will perform
well, who may leave, which interventions work, etc.
○ Enables “what if” scenario modelling → stronger business cases and ROI
justification.
2
,Chapter 2: Analytical foundations of HR measurement
Data sources in HR analytics
● HR databases (e.g., SAP, Oracle): personal details, performance, salary, diversity,
absence, promotions, leavers.
● Employee surveys: engagement, job satisfaction, person–organization fit.
● Customer satisfaction surveys: preferences, loyalty, ratings of
branches/employees.
● Sales data: revenue, new customers, target achievement.
Operational data: call center efficiency, onboarding times, supermarket scan rates.
● Key insight: linking data across systems reveals drivers of outcomes (e.g., L&D
training + customer satisfaction).
Software for HR analytics
● SPSS: user-friendly, menu-driven, widely used in HR, good for manipulation and
reporting.
● Minitab: very simple, menu-based, but less flexible.
● Stata: command-driven, strong for economists, more complex.
● SAS: powerful for big data, command-driven, steep learning curve.
● R: free, flexible, but requires coding skills. Excellent for advanced methods.
● JASP: new, free, easy GUI; still developing.
SPSS environment
● Data view: each row = a case (employee), each column = a variable (e.g., age,
gender), each cell = a value.
● Variable view: define name, type (numeric, string, date), labels, missing values,
measure level (nominal, ordinal, scale).
● Important: coding categorical data numerically (e.g., Gender: 0 = Female, 1 = Male)
enables richer statistical analysis.
Data preparation
● Import Excel/CSV files into SPSS, copy-paste data, define variables before analysis.
● Merge datasets with a unique key (Employee ID). Sorting by key variable ensures
successful merges.
● Recode text (e.g., “Marketing” → 1, “Sales” → 2) into numeric values for analysis.
Big Data in HR
● Defined by the 3 Vs: volume, velocity, variety.
● Database types:
○ Relational (SAP, Oracle) – structured, table-based.
○ Multidimensional (OLAP, Cognos) – fast for reporting/warehousing.
3
, ○ Object-oriented (Workday) – flexible, stores “objects” like Employee with
rules.
○ NoSQL (Hadoop) – handles massive, unstructured datasets.
● Business benefits: transparency, experimentation, segmentation, better
decision-making, fraud/error reduction.
● Challenge: shortage of people with deep analytical skills
Chapter 3: Analysis strategies
From descriptive to predictive HR analytics
● Descriptive: reports, averages, dashboards (historical).
● Predictive: identifies drivers and forecasts outcomes.
● Descriptive is useful but limited (snapshot view, no causality). Predictive adds
strategic value.
Statistical significance
● Hypothesis testing: null (H0) vs. research (H1).
● p-values: <0.05 (95% confidence), <0.01, <0.001 (stronger evidence).
● Significance ≠ importance; effect size matters.
Types of data
● Categorical:
○ Binary (Yes/No), Nominal (e.g., gender, department), Ordinal (e.g., salary
bands, performance ratings).
● Continuous:
○ Interval (e.g., temperature), Ratio (e.g., tenure, salary, height).
● Data type determines which test is appropriate.
Variables
● Independent variable (X) → predictor.
● Dependent variable (Y) → outcome.
● Example: Training (X) → Customer satisfaction (Y).
Which test to use?
● Chi-square: categorical X × categorical Y.
Example: Gender → Promotion.
4
