Project on Time Series
Project 4 Fall 2012 1. Open the data file called JCrew on Blackboard under the Assignments link. 2. Get a 4 point Moving Average for the data using Time Series Analysis. 3. Highlight the Revenue column and the 4MA column. Insert /Line. 4. Go back to the data. Time Series Analysis/ Exponential Smoothing. Use alpha of . 7. 5. Highlight Revenue and Smoothed and Insert /Line. 6. Go back to the data. Time Series Analysis/ Trendline / pick Exp Ln. Check the Scatterplot and all boxes on the right side. 7. Finally, go back to the data and choose Time Series/ Deseasonalize. Questions: 1.
Compare the 4 point moving average chart to the exponentially smoothed one. Which one shows the SECULAR trend better? Explain. The four point moving average shows the secular trend better because its values aren’t as volatile as they are in the exponentially smoothed model. 2. What is the forecasted revenue for JCrew in Quarter I of 2010 using Exponential Smoothing? 377. 388 in Q1 of 2010 Look at the Logged Model 3. What percent of the variation in Revenue is explained by Time? 84% of the variation is explained by time 4. By how much does Revenue change per quarter on average? Revenue changes by 4. % per quarter on average 5. Are there any outliers (suspicious or definite)? There is one outlier at time period 4, but it is only suspicious 6. Is Autocorrelation a problem? No because the Durbin-Watson is 2. 77 therefore reject fail to reject H0 H0: No residual correlation (p=0) H1: Positive residual correlation (p>1) 7. Does the data seem to fit the plot well? Explain. Yes it fits the plot well in general. There is one suspicious value that skews the plot. Look at the Deseasonalized Model 8. What is the secular trendline? y=10. 15x + 139. 39 9. How well does the model explain JCrew’s revenue? 94. 2% of the variation in Jcrew’s revenue is explained by the model 10. Which quarter is most prosperous for JCrew? 1st Quarter is the most prosperous for Jcrew with a seasonal index of . 898 11. Fill in the following table: |2010 |t |Predicted |SI |Forecast | |QI | 21 |352. 54 | . 898 | 316. 58 | |QII | 22 | 362. 69 | . 968 |351. 08 | |QIII | 23 | 372. 84 | . 938 | 349. 72 | |QIV | 24 | 382. 99 | 1. 196 | 458. 06 |