How deeply does Causal AI cover observational data?

Defines passive measurement, demonstrates group statistics, and places it in the causal hierarchy; stops short of treating association as causation.

Observational data

As explained in Causal AI

Observational data are measurements collected without the investigator assigning the treatment or intervention being studied.

When a team asks whether side-quest engagement increases in-game purchases, logged player behavior provides evidence without assigning anyone to engage more or less. This is observational data: measurements of naturally occurring behavior, contrasted with data produced by an intervention or randomized experiment. section 7.1.3

Observational data records statistical associations arising from dependencies in the data-generating process. A group comparison can estimate an effect size, but the difference alone does not establish that changing the observed behavior would cause the outcome; causal identification is needed to assess whether observational data supports that inference. section 7.1.3 section 10.0

Use observational data to:

Estimate associations when experiments cannot be run, then use causal models to simulate interventions and estimate causal effects. section 1.6
Place logged observations at level 1 of the causal hierarchy, below interventional data and counterfactual data. section 10.1.3
Check whether confounders and available causal structure permit identification before assuming that more data will solve the problem. section 10.0

Online game engagement and purchases

A database query returns 1,000 players: 482 with high side-quest engagement and 518 with low engagement. Mean purchases are $111.61 and $73.10, respectively. The $38.51 difference is a descriptive group contrast; deciding whether increasing engagement causes higher purchases requires causal analysis rather than this comparison alone. section 7.1.3

Table 7.1 Example rows from observational data on Side-Quest Engagement and In-Game Purchases

User ID	Side-Quest Engagement	In-Game Purchases
71d44ad5	high	156.77
e6397485	low	34.89
87a5eaf7	high	172.86
c5d78ca4	low	215.74
d3b2a8ed	high	201.07
dc85d847	low	12.93

Inspect representative rows pairing naturally observed side-quest engagement with in-game purchases. section 7.1.3

How Causal AI covers it

liveBook manning.com

Read Causal AI on liveBook Get the book at manning.com

Also explained in

liveBook manning.com

Causal Inference for Data ScienceObservational data are measurements collected without assigning or controlling the conditions under which observations occur.
liveBook manning.com

Delivering Machine Learning Projects: From design to deploymentObservational data is data collected from naturally occurring conditions without deliberately assigning interventions.

Compare how 3 books explain observational data →

Related concepts

causal hierarchy theorempart of · The causal hierarchy theorem states that observational, interventional, and counterfactual information are generally distinct levels that cannot be inferred from lower-level information without assumptions corresponding to the higher level.
experimentationcontrasts with · Experimentation is the deliberate testing of an idea, intervention, or alternative by observing outcomes to gather evidence or compare effects.
causal inferenceused by · Causal inference is the process of determining or estimating how causes, exposures, or interventions affect outcomes, using data and assumptions about causal relationships.
causal adjustmentused by · Causal adjustment estimates a causal effect by conditioning on variables that block noncausal pathways and averaging over their target-population distribution.

This explainer was generated by Manning's knowledge graph from the text of Causal AI. The section links point to the parts of the book it draws on.