Please answer ASAP. Thank you Regression in Excel can be performed in 3 ways: a.
ID: 3227113 • Letter: P
Question
Please answer ASAP. Thank you
Regression in Excel can be performed in 3 ways: a. Slope and Intercept from the Formulas -Statistics part of Excel b. Data Analysis Regression c. Trendline Regression on an Excel Chart. For each of these three give advantages and disadvantages. a. Slope Intercept b. Regression in Data Analysis c. Trendline Regression Multiple regression in Excel (as in every statistics software package and in all statistics books) is multiple linear regression. If we have a situation where we believe that one of the independent variables has a nonlinear effect upon the dependent variable, how might we approach the regression modeling? (could be more than one acceptable approach) We have a set of data consisting of y values and x values, and we have reason to believe that a nonlinear model y = x/(ax - b) might be an acceptable model. How would we transform the data so that we can use a y = a + bx model? Let's assume that we perform the regression analysis for the problem 3 data and obtain values for a and b; the resulting R^2 value is 0.89. What is the meaning of the 0.89 value for R^2? Regression modeling gives much more information than does a correlation analysis. Why should we even bother with a correlation analysis? Explain how to perform an analysis for multicolinearity. What is the purpose of the Adjusted R^2 when doing multiple regression?Explanation / Answer
Q-1-a
Slope intercept form gives us instant slope and intercept but do not give standard error and other useful information of regression result.
Q-1-b
Regression in data analysis is useful as it gives all information like, R-square f-test , coefficients, standard error, t-statistics and p-value but its not as quick as part-a
Q-1-c
Trend line regression help us to show the regression line with R-square on the scatter plot simultaneously and quick glance is better while other useful results are not given like test of significance.
Q-2
First we have to create the non-linear, like square, inverse, or natural logarithm of predictors and then run regression analysis using excel
Q-3
Y= x/(ax-b) = 1/(a-b*(1/x))
Taking inverse
(1/Y)= a-b*(1/X)
So run regression of Z=1/Y on W=1/X as usual
Q-4
This means that 1/X explains 89% of the variability in 1/Y
Q-5
Correlation analysis is not causal analysis and to know causal effect we need regression analysis. Whereas correlation measure the degree of relationship ,regression tells actual relationship.
Q-6
For this run the regression and calculate the variance inflation factors. IF VIF>5 is greater than moderate multicollinearity and if VIF>10 then sever multicollinearity is present
Q-7
Adjusted R-square adjust R-square for the number of predictors and sample size and hence may reduce if irrelevant variables are added to regression model. It help to compare the regression models with same dependent variable but different predictors.
Related Questions
drjack9650@gmail.com
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.