Please note that this is a statistics forum , hnce we shall take care of the sta

Question

Please note that this is a statistics forum , hnce we shall take care of the statistics part of the question only. We cant answer the economic portion of the question

We shall enter the data in excel and then perform the regression analysis as follows by going to the data > data analysis tab

Log linear function can be formed using the log of the Q

as we can see the R square is almost the same for the linear(0.85) and the log liner models(0.84) and both the variables are significant too .

Significance F5.65865E-05

for normal

Significance F0.000107769 for log linear

and the f ratio are slightly higher too for the normal model, hence we can say that the normal linear model is performing slightly better than the log linear model

please help me to answer the question

M PR N a Market P Data Analysis 1 3.00 45586.00 2.75 274000.00 7952.00 Analysis Tools 2 3.00 37521.00 3.50 13450.00 8222.00 Exponential smoothing 3 3.00 41333.00 2.64 5415 7166.00 F-Test Two-Sample for Variances 4 6800.00 6686.00 Fourier Analysis 3.25 47352.00 2.35 Histogram 5 2.75 11245.00 7715.00 Moving Average 3.25 51450.00 Random Number Generation 6 3.25 27655.00 3.15 54500.00 6643.00 Rank and Percentile 7 3.50 30265.00 2.55 26600.00 5155.00 Regression Sampling 8 3.50 39542.00 3.00 158000.00 7127.00 t-Test: Paired Two Sample for Means 9 3.50 41596.00 2.75 22500.00 5834.00 3.75 42657.00 2.45 46150.00 5093.00 10 11 3.75 36421.00 2.89 8200.00 5828.00 12 3.75 47624.00 2.49 38500.00 6590.00 13 4.00 50110.00 3.15 1 6228.00 14 4.00 57421.00 2.80 92000.00 7218.00 15 4.00 38450.00 2.90 38720.00 5846.00 OK Cancel Help

Explanation / Answer

As a measure of performance of a model, you should look at adjusted R2, not at R2 alone. You will be able to see that in linear model, the value of adjusted R2 is 82% where as the same for log-linear model is 79.7%. Apart from this measure, you should also have a residual plot, (Just check on the residuals and residual plot in the regression option of excel), where a plot showing the residuals(error part) is shown, this plot should be within a band, it shouldn't diverge. The model for which all the residuals lie within a band that model will be better than the other. Apparently from the adjusted R2 analysis it looks like that the linear model is performing better than the log-linear model

Related Questions

Navigate

• Browse (All)
• Browse P
• Subjects

---

---

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