A common question I frequently get is “What are books should the aspiring People Analyst read?” This post is a personal view on 5 that made an impact on me. Reading them won’t turn you into a great analyst, but I think all analysts will be better after reading them.
The case study: Work Rules by Lazlo Bock
Google was seen as an early pioneer in using data to understand and manage the workforce, and employees’ experiences. The case studies mentioned in this book are mostly well known by now – Google seems to work hard at controlling the message of what it is doing and to this analyst it feels like the really valuable stuff they’re doing is probably tightly hidden behind closed doors. This is hardly surprising given the ability to use People Analytics to build competitive advantage for a firm.
The part that jumped out for me when I read this book was not the analysis that they were doing but their philosophy of running experiments to test changes. Unfortunately this is rare. Well conducted experiments remain the gold-standard for good reason.
The stats book: Computer Age Statistical Inference: Algorithms, Evidence, and Data Science by Bradley Effron & Trevor Hastie
A recent publication but an absolute gem of a book. This isn’t the book to turn to if you’ve never done statistics before (but arguably if you’re that new to the field you’re probably not an aspiring People Analyst). However if, like me, your education in statistics was pre-computing this is the book which will catapult you into the 21st century.
Part philosphy and part technique Effron and Hastie quickly move through Frequentist, Bayesian and Fisherian inference and 300 or so pages they’re discussing deep learning and SVM. The graphics are well done and colour, the equations are there but well explained.
There are lots of good stats books, especially if you want to go for topic-specific ones. For general statistics knowledge Doing Bayesian Analysis by John Kruschke is a good start or even Statistics in a nutshell by Sarah Boslaugh if you want a good, easy read. The MOOCs do stats really well.
The application in business: Data Science for Business by Foster Provost & Tom Fawcett
If you want to understand how to integrate People Analytics into your work this is a very good place to start. Provost and Fawcett explain the key topics and themes of machine learning / analysis in a (relatively) simple manner, highlighting what to look out for. It’s probably best if you need to manage analysts rather than teaching you how to do analysis but of course an analyst will probably learn a lot from it, not least how to communicate what they do in a clear and concise manner.
There are numerous books on a similar theme. This one stands out to me as having a lot less hype and more practical. It appeals as one written by folks who actually do analysis rather than just want to show how exciting / cool it can be.
Developing the mental models: Personnel Economics in Practice by Edward Lazear & Michael Gibbs
Having domain knowledge is critical to do high-quality analysis. Today we’re seeing lots of great analysts coming from a variety of non-traditional backgrounds into People Analytics and there is a need for these people to quickly get knowledge of how organizations & workforces work. This rich book does that with clarity.
Personnel Economics is the application of econometrics & economics into workforce issues. It tends to be a micro-economics based approach but also covers topics from game theory to information asymmetry. This book will help you to understand a broad range of HR related topics, but do so in a manner that makes analysis of them with your data simpler.
Data visualisation: Signal by Stephen Few
I don’t agree with everything that Stephen writes but that doesn’t stop his books – especially Signal, Now You See It and Show me the Numbers being essential reading for analysts. They are beautifully published books, bristling with great, practical advice. Signal, the most recent one is a great addition – I particularly like the discussion on understanding uncertainty and his discussion on funnel plots (p46 – 53) and parallel coordinates plots (p148 – 152).
I would treat his ideas – and they’re passionately made – as starting points. We do a lot of usability testing of visualisations and some of the recommended approaches don’t seem to work in practice as well as they might be expected to. The other aspect which isn’t given much discussion in his work is the role of interaction in modern visualisation. Interaction adds other goals to a visualisation as well as clarity of understanding. The balance between these needs is a delicate manner.
Very few of these books will give you the immediate tools or examples to doing analytics. However, I think they are more useful as they provide the insight to how to approach an analytic problem, how to structure it, how to decide how you’re going to tackle it. All require thought into how to apply what they discuss but in the end this will likely make you a better analyst than something which encourages you to follow a few recipes.
For more HR Analytics-specific books have a look at iNostix’s list of 29 HR analytics Books from 2015.