Don't fly blind. See how the power of experiments works for you. When it comes to improving customer experiences, trying out new business models, or developing new products, even the most experienced managers often get it wrong. They discover that intuition, experience, and big data alone don't work. What does? Running disciplined business experiments. And what if companies roll out new products or introduce new customer experiences without running these experiments? They fly blind. That's what Harvard Business School professor Stefan Thomke shows in this rigorously researched and eye-opening book. It guides you through best practices in business experimentation, illustrates how these practices work at leading companies, and answers some fundamental questions: What makes a good experiment? How do you test in online and brick-and-mortar businesses? In B2B and B2C? How do you build an experimentation culture? Also, best practice means running many experiments. Indeed, some hugely successful companies, such as Amazon, Booking.com, and Microsoft, run tens of thousands of controlled experiments annually, engaging millions of users. Thomke shows us how these and many other organizations prove that experimentation provides significant competitive advantage. How can managers create this capability at their own companies? Essential is developing an experimentation organization that prizes the science of testing and puts the discipline of experimentation at the center of its innovation process. While it once took companies years to develop the tools for such large-scale experiments, advances in technology have put these tools at the fingertips of almost any business professional. By combining the power of software and the rigor of controlled experiments, today's managers can make better decisions, create magical customer experiences, and generate big financial returns. Experimentation Works is your guidebook to a truly new way of thinking and innovating.

Featuring engaging examples from diverse disciplines, this book explains how to use modern approaches to quasi-experimentation to derive credible estimates of treatment effects under the demanding constraints of field settings. Foremost expert Charles S. Reichardt provides an in-depth examination of the design and statistical analysis of pretest-posttest, nonequivalent groups, regression discontinuity, and interrupted time-series designs. He details their relative strengths and weaknesses and offers practical advice about their use. Reichardt compares quasi-experiments to randomized experiments and discusses when and why the former might be a better choice. Modern moethods for elaborating a research design to remove bias from estimates of treatment effects are described, as are tactics for dealing with missing data and noncompliance with treatment assignment. Throughout, mathematical equations are translated into words to enhance accessibility.

Like other sciences and engineering disciplines, software engineering requires a cycle of model building, experimentation, and learning. Experiments are valuable tools for all software engineers who are involved in evaluating and choosing between different methods, techniques, languages and tools. The purpose of Experimentation in Software Engineering is to introduce students, teachers, researchers, and practitioners to empirical studies in software engineering, using controlled experiments. The introduction to experimentation is provided through a process perspective, and the focus is on the steps that we have to go through to perform an experiment. The book is divided into three parts. The first part provides a background of theories and methods used in experimentation. Part II then devotes one chapter to each of the five experiment steps: scoping, planning, execution, analysis, and result presentation. Part III completes the presentation with two examples. Assignments and statistical material are provided in appendixes. Overall the book provides indispensable information regarding empirical studies in particular for experiments, but also for case studies, systematic literature reviews, and surveys. It is a revision of the authors’ book, which was published in 2000. In addition, substantial new material, e.g. concerning systematic literature reviews and case study research, is introduced. The book is self-contained and it is suitable as a course book in undergraduate or graduate studies where the need for empirical studies in software engineering is stressed. Exercises and assignments are included to combine the more theoretical material with practical aspects. Researchers will also benefit from the book, learning more about how to conduct empirical studies, and likewise practitioners may use it as a “cookbook” when evaluating new methods or techniques before implementing them in their organization.

Every company's ability to innovate depends on a process of experimentation whereby new products and services are created and existing ones improved. But the cost of experimentation often limits innovation. New technologies--including computer modeling and simulation--promise to lift that constraint by changing the economics of experimentation. Never before has it been so economically feasible to ask "what-if" questions and generate preliminary answers. These technologies amplify the impact of learning, paving the way for higher R&D performance and innovation and new ways of creating value for customers.In Experimentation Matters, Stefan Thomke argues that to unlock such potential, companies must not only understand the power of experimentation and new technologies, but also change their processes, organization, and management of innovation. He explains why experimentation is so critical to innovation, underscores the impact of new technologies, and outlines what managers must do to integrate them successfully. Drawing on a decade of research in multiple industries as diverse as automotive, semiconductors, pharmaceuticals, chemicals, and banking, Thomke provides striking illustrations of how companies drive strategy and value creation by accommodating their organizations to new experimentation technologies.As in the outcome of any effective experiment, Thomke also reveals where that has not happened, and explains why. In particular, he shows managers how to: implement "front-loaded" innovation processes that identify potential problems before resources are committed and design decisions locked in; experiment and test frequently without overloading their organizations; integrate new technologies into the current innovation system; organize for rapid experimentation; fail early and often, but avoid wasteful "mistakes"; and manage projects as experiments.Pointing to the custom integrated circuit industry--a multibillion dollar market--Thomke also shows what happens when new experimentation technologies are taken beyond firm boundaries, thereby changing the way companies create new products and services with customers and suppliers. Probing and thoughtful, Experimentation Matters will influence how both executives and academics think about experimentation in general and innovation processes in particular. Experimentation has always been the engine of innovation, and Thomke reveals how it works today.

Optimize the performance of your systems with practical experiments used by engineers in the world’s most competitive industries. In Experimentation for Engineers: From A/B testing to Bayesian optimization you will learn how to: Design, run, and analyze an A/B test Break the "feedback loops" caused by periodic retraining of ML models Increase experimentation rate with multi-armed bandits Tune multiple parameters experimentally with Bayesian optimization Clearly define business metrics used for decision-making Identify and avoid the common pitfalls of experimentation Experimentation for Engineers: From A/B testing to Bayesian optimization is a toolbox of techniques for evaluating new features and fine-tuning parameters. You’ll start with a deep dive into methods like A/B testing, and then graduate to advanced techniques used to measure performance in industries such as finance and social media. Learn how to evaluate the changes you make to your system and ensure that your testing doesn’t undermine revenue or other business metrics. By the time you’re done, you’ll be able to seamlessly deploy experiments in production while avoiding common pitfalls. About the technology Does my software really work? Did my changes make things better or worse? Should I trade features for performance? Experimentation is the only way to answer questions like these. This unique book reveals sophisticated experimentation practices developed and proven in the world’s most competitive industries that will help you enhance machine learning systems, software applications, and quantitative trading solutions. About the book Experimentation for Engineers: From A/B testing to Bayesian optimization delivers a toolbox of processes for optimizing software systems. You’ll start by learning the limits of A/B testing, and then graduate to advanced experimentation strategies that take advantage of machine learning and probabilistic methods. The skills you’ll master in this practical guide will help you minimize the costs of experimentation and quickly reveal which approaches and features deliver the best business results. What's inside Design, run, and analyze an A/B test Break the “feedback loops” caused by periodic retraining of ML models Increase experimentation rate with multi-armed bandits Tune multiple parameters experimentally with Bayesian optimization About the reader For ML and software engineers looking to extract the most value from their systems. Examples in Python and NumPy. About the author David Sweet has worked as a quantitative trader at GETCO and a machine learning engineer at Instagram. He teaches in the AI and Data Science master's programs at Yeshiva University. Table of Contents 1 Optimizing systems by experiment 2 A/B testing: Evaluating a modification to your system 3 Multi-armed bandits: Maximizing business metrics while experimenting 4 Response surface methodology: Optimizing continuous parameters 5 Contextual bandits: Making targeted decisions 6 Bayesian optimization: Automating experimental optimization 7 Managing business metrics 8 Practical considerations

The Philosophy of Scientific Experimentation focuses on the identification and clarification of philosophical issues in experimental science.Since the late 1980s, the neglect of experiment by philosophers and historians of science has been replaced by a keen interest in the subject. In this volume, a number of prominent philosophers of experiment directly address basic theoretical questions, develop existing philosophical accounts, and offer novel perspectives on the subject, rather than rely exclusively on historical cases of experimental practice.Each essay examines one or more of six interconnected themes that run throughout the collection: the philosophical implications of actively and intentionally interfering with the material world while conducting experiments; issues of interpretation regarding causality; the link between science and technology; the role of theory in experimentation involving material and causal intervention; the impact of modeling and computer simulation on experimentation; and the philosophical implications of the design, operation, and use of scientific instruments.

In Era of Experimentation, Daniel Peart challenges the pervasive assumption that the present-day political system, organized around two competing parties, represents the logical fulfillment of participatory democracy. Recent accounts of "the rise of American democracy" between the Revolution and the Civil War applaud political parties for opening up public life to mass participation and making government responsive to the people. Yet this celebratory narrative tells only half of the story. By exploring American political practices during the early 1820s, a period of particular flux in the young republic, Peart argues that while parties could serve as vehicles for mass participation, they could also be employed to channel, control, and even curb it. Far from equating democracy with the party system, Americans freely experimented with alternative forms of political organization and resisted efforts to confine their public presence to the polling place. Era of Experimentation demonstrates the sheer variety of political practices that made up what subsequent scholars have labeled "democracy" in the early United States. Peart also highlights some overlooked consequences of the nationalization of competitive two-party politics during the antebellum period, particularly with regard to the closing of alternative avenues for popular participation.