What is 'Regression?'

Regression is a statistical quantity used in economics, investing and other alterations that tries to define the strength of the association between one dependent variable (usually represented by Y) and a sequence of other altering variables (recognized as independent variables). Regression assistances investment and financial managers to value assets and understands the relations among variables, such as product prices and the stocks of businesses trade in those commodities.

In statistical demonstrating, regression analysis is a statistical procedure for assessing the relationships among variables. It contains many methods for modeling and analyzing several variables, when the emphasis is on the association among a dependent variable and one or more independent variables.

regression

Regression

Simple regression is used to observe the association among one dependent and one independent variable. After execution an analysis, the regression statistics can be used to forecast the dependent variable when the independent variable is recognized. Regression goes beyond correlation by adding forecast competences.

The regression line is a plot of the predictable value of the dependent variable for all values of the independent variable.  The regression line is the one that best fits the data on a distribute plot.

Example

A company wants to know if there is an important association among its publicity expenses and its sales volume. The independent variable is publicity budget and the dependent variable is sales volume. A lag time of one month will be used because sales are predictable to lag behind actual publicity expenses. Data was together for a six month period. All facts are in thousands of dollars. Is there a important association between publicity budget and sales volume?

Independent Variable

Dependent Variable

6.1

27.1

4.2

25.0

3.9

30.4

5.7

29.7

7.3

40.1

5.9

28.8

Model: y = 9.873 + (3.682x) + error
Standard error of the estimation = 2.637
t-test for the implication of the slope = 3.961
Grades of freedom = 4
Two-tailed possibility = .0149
r-squared = .807

You might make a statement in a report like this: A simple linear regression was executed on six months of data to define if there was a important association between publicity expenses and sales volume. The t-statistic for the slope was important at the .05 critical alpha level, t(4)=3.96, p=.015. Thus, we discard the null hypothesis and conclude that there was a positive important association between publicity expenses and sales volume. Furthermore, 80.7% of the inconsistency in sales volume could be clarified by publicity expenses.

Analysis is a statistical procedure for assessing the relationships among variables. It contains many methods for modeling and analyzing several variables, when the emphasis is on the association among a dependent variable and one or more independent variables.

 

business statistics

March 09, 2018