Due Nov 10 (Friday) Submit your work in class: 20 points Name: SUMMARY OUTPUT Re

Question

Due Nov 10 (Friday) Submit your work in class: 20 points Name: SUMMARY OUTPUT Regression Statistics Multiple R 0.769489638 R Square 0.592114303 Adjusted R Square 0.56073848 Standard Emor 793089.0209 Observations 2 ) Inputs observations ANOVA SS MSF Significance F Regression 1 1.18701E+13 1.19E+13 18.871670.000795544 Residual 13 8.17687E+12 6.29E+11 Total 14 2.0047E+13 Intercept XVariable 1 Coefficients Standard Emort Stat TP-value Lower 95% Upper 95% Lower 95.0% Upper 95.0% 4833.942857 430931.3949 0.011217 0.99122 -926136.7357 935804.6214 -926136.7357 935804.6214 205896. 1071 47396.13437 4.344 154 0,000796 103502, 984 308289.2303 103502.984 308289.2303 avn itThousand 1. (2 Points) describe the two purposes of using a regression analysis in the business decisions. 2. (2 points) what can you measure if you use a regression analysis? a. Pattern b. Trend c. Seasonality d. Irregularity 3. (2 points) take a look at number 1. Interpret 0.5607 (Do not simply say it is a measure of how close the data are to a regression line, and focus on the number) 4. (2 points) take a look at number 2. What does 15 mean in the table?

Explanation / Answer

Two purposes of using regression in business decisions:

Regression Analysis determines the strength of relationship between dependent and independent variables. It can therefore be used for predicting or forecasting future trends and effects by using past data. This can help businesses plan for future requirements. Regression can help in demand analysis or in predicting effects of increasing man-hours on company revenues among other things.

It is also used for optimization and quality control of processes by analysing the dependency between various variables. It can help businessess minimize costs by developing a causal relationship between costs and performance indicators.

--------------------------------------------------------------

Pattern, Trend and seasonality can be analysed using regression

Irregularity is better handled by simulation model

---------------------------------------------------------------

R squared gives an estimate of how good a fit a model is. The value of R square increases with increase in number of effects. But the fact is not all effects improve the model. So, R square gives a highly optimistic estimation for linear regression fit which is corrected by adjusted R square. This is why you'll notice adjusted R square is always less than R square.

0.5607 (which is less than R squared value) means that some of the predictors in the models do not contribute significantly to the goodness-of-fit of the model. The value of adjusted R squared will increase if useless predictors are dropped from the model.

----------------------------------------------------------

15 gives the number of observations in the model. Simulations have shown that 10-20 observations per variable is a good rule of thumb. Any less and your results will be erratic giving misleading values of p and R square. Your sample must be large enough to get a good fit.

Related Questions

Navigate

• Browse (All)
• Browse D
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.