Picture yourself as a teacher contemplating a curious phenomenon among high school students: the relationship between sleeping late and catching the school bus. An initial hypothesis might be, “Sleeping late causes students to miss their school bus.” This stems from a personal experience: I slept late and consequently missed the bus. However, this hypothesis might be challenged upon observing graduate students, who, despite sleeping late, consistently catch their buses.

This scenario exemplifies the complex, often misleading nature of causality. The observed association—sleeping late and missing buses—doesn’t inherently imply causation. Here, various other factors could be at play. Perhaps high school students have earlier bus schedules, or graduate students, despite sleeping late, can wake up early and not miss their transportation. It’s plausible that the initial observation of sleeping late and causing one to...

In Hellenic philosophy, ancient Greece played a pivotal role. Pre-Socratic philosophers such as Thales and Heraclitus explored the nature of change and causality. Ancient Greek philosophers contributed significantly to systematic approaches to understanding causality by examining cause-and-effect relationships. They introduced important causal concepts, including the idea that nothing comes from nothing (attributed to Parmenides). While not originating the principle of sufficient reason, their work laid the foundations for later philosophical developments. Their ideas have influenced our understanding of causation, though modern concepts have evolved significantly, incorporating insights from various traditions, scientific advancements, and mathematical frameworks.

Aristotle, however, provided a more structured approach to causality with his four causes theory:

• Material cause: The material from which something is made (e.g., the...

Beyond understanding the theoretical and historical underpinnings, one must ponder the practical necessity of causality first. We shall discuss examples of the ubiquitous application of causality across various industries. For instance, enterprises leverage causal inference techniques to gain deeper insights into customer behaviors, needs, and preferences. They employ these methods to elucidate both natural and anthropogenic phenomena. Mastery of causal inference equips you with an extremely powerful tool, rendering you an invaluable asset in any team or organizational context. Your proficiency in this domain can significantly contribute to the overarching objective of delivering value to stakeholders.

Let's discuss further why causality is not only an intellectually rewarding area but also practically indispensable.

In medical and public health arenas, causal inference is vital for assessing treatment efficacy. Randomized controlled trials (RCTs...

It’s tempting to attribute causality to superficial observations, mistaking mere associations for causation. Take, for instance, the observation that social media posts made later in the day receive fewer likes and comments, suggesting reduced engagement. One might hastily conclude that the timing of these posts is the causal factor. However, without rigorous statistical testing, such claims remain speculative. In this book, we will teach you how to conduct these necessary tests, distinguishing between simple association and true causation.

In statistics, we discuss association, causation, and correlation. While correlation is often used interchangeably with association in everyday conversations, they have distinct meanings in statistical contexts. So, what is the difference between association and correlation?

In causality, association encapsulates a general linkage between two variables, without explicitly characterizing the...

Let's walk through a study titled Inked into Crime? An Examination of the Causal Relationship between Tattoos and Life-Course Offending among Males from the Cambridge Study in Delinquent Development [1]. It’s about a study to ascertain whether or not there exists a causal connection between the presence of tattoos and the propensity for criminal behavior across one’s lifespan. Analyzing data from 411 British males, the researchers utilized propensity score matching—a statistical technique frequently used to ascertain causal inference (which we shall learn about later in this book, in Chapter 6). This approach meticulously dissects the complexity between the ink on skin and the propensity for crime, offering a more refined perspective on this age-old debate.

Rooted in the shadow of 19th-century criminological thought, specifically the theories of Lombroso, tattoos have long been cast in the dim light of criminality...

From the previous section, it is evident that causal inference involves employing observational or experimental data to establish causal links, utilizing various statistical methods and theories to measure the influence of one variable (the “treatment” or “intervention”) on another (the “outcome” or the “effect”).

From a statistical vantage point, it focuses on estimating the counterfactual, hypothesizing the outcomes in alternate scenarios where the treatment was absent. This necessitates assumptions about data and underlying mechanisms, including the exclusion of unmeasured confounders. Let’s go over these concepts one by one.

Counterfactual analysis

Counterfactual analysis involves exploring "what-if" scenarios to understand the effects of actions that didn't occur. It is used to estimate the causal impact of interventions by imagining alternative outcomes...

This chapter introduced the concept of causality and its importance across various fields. A brief historical overview acknowledged the contributions of ancient philosophers and modern statisticians in developing causal inference methods. We started by defining causal inference and distinguishing it from association and correlation with practical examples. We also touched on complex ideas such as potential outcomes, confounding variables, and Simpson’s paradox, explaining how they affect causal studies.

Finally, the chapter underscored the importance of causal inference in making informed decisions in our data-driven world. This foundation prepares you for a deeper exploration of causal inference in subsequent chapters.

1. Inked into Crime: https://www.sciencedirect.com/science/article/abs/pii/S0047235213001189.
2. Causal or Spurious: Using Propensity Score Matching to Detangle the Relationship between Violent Video Games and Violent Behavior: https://www.researchgate.net/publication/257252863_Causal_or_Spurious_Using_Propensity_Score_Matching_to_Detangle_the_Relationship_between_Violent_Video_Games_and_Violent_Behavior.
3. Imbens, G.W., Rubin, D.B. (2010). Rubin Causal Model. In: Durlauf, S.N., Blume, L.E. (eds) Microeconometrics. The New Palgrave Economics Collection. Palgrave Macmillan, London. https://doi.org/10.1057/9780230280816_28.
4. Mill’s methods, https://beisecker.faculty.unlv.edu/Courses/Phi-102/Mills_Methods.htm.
5. Peter Armitage, Fisher, Bradford Hill, and randomization, International Journal of Epidemiology, Volume 32, Issue 6, December 2003, Pages 925–928, https://doi.org/10.1093/ije/dyg286.
6. BIOS 6611 (2021, July 31) , Neyman-Pearson Approach to...