Lab12 - like serial correlation is present? (d) Calculate...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
Lab Exercise – Week 12 STAT 3022, section 001 11/22/2011 Computational Exercise. Perhaps the most intensively studied time series consists of annual counts of sunspots, begun in 1610 by Wolfer and continued to the present time; 200 years of the sunspot series appear in the data set found at http://users.stat.umn.edu/ ~ graalum/data/ch15/ex1509.csv (a) Construct a time plot of this series. Adjust the aspect-ratio of the plot so that runs are more apparent. (b) Transform the variable spots to the square-root scale. Construct a time plot of this series. Add a horizontal line at ¯ y , where y = spots . Does it look like serial correlation is present? (c) Construct a lag plot to show the relationship between adjacent residuals. Does it look
Background image of page 1
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: like serial correlation is present? (d) Calculate the estimate of the rst serial correlation coecient, r 1 . (e) Perform the large-sample test for serial correlation, described in Section 15.4.1. What evidence to we have that a serial correlation exists (i.e., test H : = 0 vs. H a : 6 = 0)? (f) Conduct the nonparametric runs test to see if serial correlation is present. Do your conclusions agree with those from part (e)? (g) Finally, calculate the standard error of the average in our serially correlated time series, using the adjustment described in Section 15.2.2. How does it dier from the unadjusted standard error?...
View Full Document

Ask a homework question - tutors are online