collinearity
UK[ˌkɒlɪnɪˈærɪti]US[ˌkɑːlɪnɪˈærɪti]
n
a situation in statistics where two or more predictor variables in a regression model are highly linearly related.
n
the condition of points lying on a single straight line.
Morpheme Breakdown
co
linear
ity
co
together
linear
line
ity
state or condition
Etymology
the word 'collinearity' is a modern scientific formation built from classical components. the prefix 'co-', from latin, establishes the concept of togetherness. it is attached to the adjective 'linear', which itself derives from the latin 'linea' (a line, string), a word of uncertain origin, possibly related to 'linum' (flax, thread). the suffix '-ity', also of latin descent, transforms the combined concept into an abstract noun denoting a state or condition. thus, the term's literal construction, 'the state of being together on a line', logically evolved into its specialized statistical meaning describing variables that move together in a linear fashion, as well as its geometric sense for points sharing a single line.
Analysis
Structure: co (together) + linear (line) + ity (state or condition)
- co: a prefix from latin 'com-', meaning 'together' or 'with'. it indicates joint action or association.
- linear: from latin 'linearis', meaning 'pertaining to a line'. it is an adjective-forming suffix in english.
- ity: a noun-forming suffix from latin '-itas', denoting a state, condition, or quality.
Examples
high collinearity between the independent variables can make it difficult to determine their individual effects on the outcome.
in geometry, the collinearity of the three points was proven using vector methods.
the researcher used variance inflation factors to check for multicollinearity, a specific form of collinearity, in the regression model.