Linear Regression

From Displayr
Jump to navigation Jump to search

A statistical technique used to quantify the causal relationship between variables (it assumes a causal relationship). For example, it may be believed that advertising expenditure and price cause sales which relationship can be expressed as:

[math]\displaystyle{ Sales=\beta_{Constant}+\beta_{Advertising} \times Advertising + \beta_{price} \times Price }[/math]

where [math]\displaystyle{ \beta_{Constant} }[/math], [math]\displaystyle{ \beta_{Advertising} }[/math] and [math]\displaystyle{ \beta_{Price} }[/math] are parameters to be estimated from the data (in the context of regression, the parameters are commonly known as coefficients and these parameters represent the underlying base rate of sales and the relative effects of advertising and price, respectively).