If a sample of data is being analyzed via linear regression and fails

 

If a sample of data is being analyzed via linear regression and fails to follow linear regression assumptions, the results given by the data lead to an incorrect answer. Linear regression measures variables of predictability to explain dependent variables. Linear regression assumes homoscedasticity, multivariate normality, linear relationship, no auto-correlation, and multivariate normality. For example, homoscedasticity assumes variances in error are the same for all values as a result; errors with respect to predicting outcomes are larger than anticipated. When conducting statistical analysis certain tests known as parametric statistical tests make assumptions on the parameters of those tests. The parameters are the distribution of the population by which the results are pulled. When there are violations of these assumptions the result is altered. This leads to incorrect or misleading information but often the effect of the violation depends on the level or degree of the violation.

The influence of the assumption violation on business decision-making can negatively affect a company financially. As an example, many of today’s fortune 500 companies use statistical analysis to determine future demand. Is it necessary to eliminate a product or increase its productivity? How many new customers will be gained and how many will be lost? What time of year are customers more likely to want a service, gadget? How long will it take to deliver a widget from point A to point B? These companies use statistical information to make the appropriate decisions at the right time. If assumption violations occur while conducting the statistical analysis, the false information gathered off the results will cause a company to take a hit from its bottom-line and in some instances will result in bankruptcy or a business having to close its doors.

Ismael

Reference:

What is Forecasting? (n.d.). Retrieved July 28, 2017, fromhttps://tlc.trident.edu/d2l/le/content/104339/view…

Testing of Assumptions. (n.d.). Retrieved July 28, 2017, from http://www.statisticssolutions.com/testing-of-assu…

student 2

Normally, assumptions are negative, typically they mean that there is something accepted as truth without any evidence or proof. With interpreting regression results, there are some assumptions that should not be violated. According to the text, there are multiple assumptions while interpreting the results. If these assumptions were violated “biased or variance of the estimate will be increased.” (Kannu, 2016) Any violation will lead to a misinterpretation of the data. Decisions, especially those made in businesses, need to be created on a bases of properly collected and analyzed research and data. Data is necessary to make the most efficient and effective decisions. While violating regression assumptions, a bias is created and information may end up skewed or misinterpreted. According to Osborne & Waters (2002) some of these violations are:

-Variables are normally distributed.

-Assumption of a linear relationship between independent and dependent variable(s)

-Variables are measured without error (reliably)

-Assumption of Homoscedasticity

As future business leaders, it is our responsibility to be able to determine if violations were present and if we are unaware to look for assumption violations, we will fall into the trap of making incorrect decisions.

References

Kannu, B. P. (2016, June 27). What are the consequences of violating linear regression assumptions? Retrieved July 25, 2017, from https://www.quora.com/What-are-the-consequences-of-violating-linear-regression-assumptions

Osborne, J. W., & Waters, E. (2002). Four Assumptions Of Multiple Regression That Researchers Should Always. Practical Assessment, Research & Evaluation, 8(2), 1-5. Retrieved July 25, 2017, from http://PAREonline.net/getvn.asp?v=8&n=2