Where does joint collide?

Here the word is an adjective, not a thing: the joint distribution, or joint probability, of two or more random variables gives the probability of their values occurring together, and it holds all the information about how the variables behave as a set, including how they depend on one another. Casella and Berger's Statistical inference develops the joint density and mass function, from which the marginal and conditional distributions are then recovered. The referent is a mathematical object over several variables at once, estimated from data, and it names a relationship rather than a physical joint in a body, a frame, or a rock face.[1]

Unit

A probability assigned to combinations of values of two or more random variables

Recovery means

Not applicable

Measured by

Estimated from paired or multivariate data as a joint density, mass function, or contingency table

A natural fracture in rock across which there has been no visible shear displacement: the two walls have pulled apart rather than slid past each other, which sets a joint apart from a fault. Joints usually form in regular sets, and their orientation and spacing record the stresses the rock has been through. Engelder's Loading paths to joint propagation during a tectonic cycle studied how such fractures open and grow through a cycle of burial, uplift, and unloading. The referent is a break in rock formed by extension, mapped in the field, and it has nothing in common with an articulation in the body, a bolted connection, or a probability over variables beyond the four letters.[2]

Unit

A natural extension fracture in rock, by its orientation, spacing, and set

Recovery means

Not applicable

Measured by

Mapped in the field by strike, dip, and spacing, and related to the stress state that opened it