When doing regression, simple linear or any of the other
regression approaches, the analyst always must begin with the
determination/isolation of at least two key variables -- one
"dependent" and the other "independent". So, for example, I may do
a forecast of future profits that relies on sales data
(independent) and associated profit data (dependent) from the same
years. We say something like...
"Since profits "depend" on sales volume (not the other way
around), we can use one (sales) to...