Statistics and Time Series.

Check your regression analysis for these 7 Assumptions. A Quantitative Research Consulting Firm.

At we help clients to get statistical data straight and accurate as much as possible. No matter whether our client is a lawyer supporting a case for economic damages, or  a graduate student writing a paper contesting the counterpart’s mishandling of data. will provide technical support for a cogent understanding of the data being used as evidence in litigation or academic processes.

Before you submit your paper for which you successfully run several simple or multiple regression analysis, we at can help you out checking everything is sound.

These are the 7 assumptions your work must comply with wherever you run the Method of Least Squares:
The overall objective of checking the 7 assumptions is to gauge how close or far your estimates are from the population data. In other words, we will tell you how “accurate” your betas are.

Assumption 1:

Your model must be linear in the parameters (Note that if your data is not linear in the variables we can linearize it for you).

Assumption 2:

Your X’s values must be fixed and independent of the error term.

Assumption 3:

Zero mean value of the error term.

Assumption 4:

Homoscedasticity. Your regression must not suffer heteroscedasticity.

Assumption 5:

No autocorrelation between disturbances. This means no correlation between error terms.

Assumption 6:

The number of observations must exceed the number of parameter to estimate. You can do that by yourself.

Assumption 7:

Nature of the X variable. Neither outliers nor negative variance in variable X.


Contact us if you think we can be of any help.

2 replies »

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.