revision

[Type text] [Type text] [Type text]

7

Regression Analysis Reporting II

Trident University

Teresa A. Coward/ ID M0000318024

Module 2 SLP 2

BUS520: Business Analytics and Decision Making

Professor Dr. David Fogarty

January 29th, 2018

Overview

I’m a consultant for the Diligent Consulting Group, previously completed the initial project for our client, which was comprised of developing and testing a forecasting method that used linear regression techniques. This method used monthly year one sales over a twelve-month period to forecast year two sales. The ABC Furniture Company believed that the number of patrons who visit their store during any particular month was in relation to the total number of sales for that given month in question. More specifically, the client believed that there was a positive relationship between higher customer traffic in the store and higher total sales associated with consumer commerce, i.e. the client believed that the higher the number of customers who visited the store, the higher the total sales would be.

The client had provided me with the number of customers who visited the store over the most recent twelve-month period from January to December, with the sales corresponding to each of those months. A linear regression equation was obtained using this client's collected information. The linear regression equation was then used to forecast the sales for year two. The forecast sales were later compared with the actual year two sales. In this case the comparison was meant to obtain the trend with which the performance in this docket was moving. This is an analytical move that is used in obtaining for example variances for analysis purposes and ultimately making a decision.

Statistical Evaluation

When factual information is used to scientifically examine closely data by utilizing linear, logarithmic or exponential models for representations and make for certifiable investigations. The information gathered acts as a motivator behind basic leadership decisions. In this manner, for our situation we will utilize the factoring principle where the data is concerned and negate through the research, taking a gander at all the different issues that needs to be address that are concerns of management, from those suggestions steer to a comprehension of these different factors connecting together for a solution. One of the most usual applications of statistics is describing a set of data using estimation. By anlizing thus throughly examining the raw data, we can make and draw a logical conclusion or even compare, contrast or rank of the data on the specified attribute. This helps us to make a clear analysis of the data at hand and therefore come up with clear understanding of this correlation between the two, therefore coming to a sound decision in the end accordingly. Evaluating the status of your business by considering its attributes that affect customers is a very important aspect for growth and development, of any business establishments (Walpole, 1982). As a manager or any other executive for consideration with the mandate of managing the existence and operations of the business, the understanding of the foresaid variables relationship is a crucial thing that needs not be ignored. My research will show this, as far as wanting the corporation to go far as far as performance and economic visibility are concerned.

According to Statistics How To.com; “the mean error is an informal term that usually refers to the average of all the errors in a set. In dissecting this case study, we are creating the linear equation and regression model that will give us a clear relationship between our independent and dependent variable. First, we’ll calculate in excel the mean error and then we’ll streamline to viable conclusion, as quoted from Statistics How To.com; an “error” in this context is an uncertainty in a measurement, or the difference between the measured value and true/correct value. The more formal term for error is measurement error, also called observational error. How the data relate in regard to the correlation that the two variables have, the value expected from the same correlation and the behavior of the regression line. The linear regression makes an effort to model the affiliation between supported variable and objective variable by fitting a linear equation to observed this figures. In our case the dependent variable is sale and independent variable is the consumer.

The mean percentage error (MPE) is the computed average of percentage errors by which forecasts of a model differ from actual values of the quantity being forecast.

The mean absolute percentage error (MAPE), also known as mean absolute percentage deviation (MAPD), is a measure of prediction accuracy of a forecasting method in statistics, for example in trend estimation, as it usually expresses accuracy as a percentage.

Value Calculation Forecast

Endeavoring to fit all raw data for value review, applicable information in this technique once the determination of the association between not standing more on the opposition that one variable causes the other. A linear regression line has an equation of the form, where X is the explanatory variable and Y is the dependent variable. The slope of the line is, and is the intercept (the value of y when x = 0).

The provided in the excel sheet we can see that there are two column one is sales and other one is customer. This portion of the research we’ll assume and conclude that

Dependent variable (Y) = sales, the Independent variable (X) = customers hence we have to fit regression and find scatter plot and analyze as well as interpret the data. From the regression and scatter plot the linear equation of the model is. (Excel sheet is attached)

In the equation the slope is 0.648 and the y intercept is 111.65. The interpretation of slope is for one unit change in customers will be 0.648 unit increase in sales.

Mean absolute percentage error calculation.

And for SES – MAPE for alpha = 0.15

And for SES – MAPE for alpha = 0.9

Now for overall significance test statistic follows F-distribution and for individual significance test statistic follows t-distribution.

Here P-value < alpha, Reject H0 at 0.05 level of significance. Deduction, the population slope for customers is different than 0. Or consumers are significant variable.

Concluded Recommendation

After diligent research and as your consultant for the Diligent Consulting Group, I’ve completed the analysis as well as finalized the forecasting by the two methods; fist Linear Regression (LR) and Single Exponential Smoothing (SES) to forecast sales. Therefore, I have been able to categorize the relationship between our two main identified variables in this case; consequently, my proposal is as follows:

My recommendations as I’ve come to understand through my research, is that the mean absolute percentage error is 6.620 for Single Exponential Smoothing method and the mean absolute percentage error is 17.736 for forecast method. Simply, I’ve concluded that the lowest mean absolute percentage error is better to use and suggested which Single Exponential Smoothing method.

References

CONTENT TEAM, A. (2016, July 14). Going Deeper into Regression Analysis with

Assumptions, Plots &

Downie, N. M. & Heath, R. W. (1965). Basic Statistical Methods (2nd ed.). Harper &

Row Publishers

Solutions, S. (n.d.). Assumptions of Linear Regression. Retrieved January 23, 2018, from

http://www.statisticssolutions.com/assumptions-of-linear-regression/

Statistics How To.com. (n.d.). Regression Equation: What it is and How to use it.

Retrieved January 22, 2018, from http://www.statisticshowto.com/what-is-a-regression-equation/

Walpole, R. (1982). Introduction to Statistics. (3rd ed.). Prentice Hall Publication.

(2016, January 22). Retrieved January 23, 2018, from

https://www.youtube.com/watch?v=n8J5TbbFSN4