Skip to main content

SAGE Social Science Thesaurus

Search from vocabulary

align
annotate
download

Concept information

... > social science subjects > education > educational systems > curriculum > curriculum studies > inquiry and research > quantitative research > quantitative data analysis > general linear models > least squares criterion
... > social science subjects > social research > research methods and evaluation > quantitative research > quantitative data analysis > general linear models > least squares criterion
... > contexts > population characteristics > health > public health > epidemiologic methods > statistics as topic > regression analysis > least squares criterion

Preferred term

least squares criterion  

Definition

  • The rule that using the ∗mean to predict the scores in a distribution results in predictions that are most accurate, with “accurate” in this case indicating that using the mean yields the smallest possible sum of squared ∗errors (or squared ∗deviation scores). In ∗regression analysis, it is called ∗ordinary least squares (OLS), which is a method or criterion for calculating the ∗regression equation (or drawing the regression line) that best summarizes or fits a distribution. [Source: Dictionary of Statistics & Methodology; Least Squares Criterion]

Broader concept

Belongs to group

URI

http://data.loterre.fr/ark:/67375/N9J-W2HCJFXD-0

Download this concept:

RDF/XML TURTLE JSON-LD