This is an ongoing listing of errors in the 5th edition. I will give credit to those who point out errors when I can. When I don't know the name of the person who pointed out the error, I am still grateful even if I can't give credit.
Because of the way I expect to distribute this list, I cannot be sure
that subscripts and Greek symbols will come across. Subscripts are usually shown in
parentheses [as in s(1)], superscripts as s^2, and Greek letters are spelled out.
| Page | Line | Correction | |||||||||
| Frontpages | This book is dedicated to my wife, Donna. I forgot to do that in the first printing, and feel really bad. | ||||||||||
| 67 | Prob 2.20 | Jaime Adcock pointed out that in part b the ")" should come after the superscript 2, not before it. | |||||||||
| 89 | Ex 3.19 | Tim Weston pointed out that a much better answer than the one I gave would be " that the T-scores do not have an exact mean of 50 and a SD of 10 because of random sampling error." I gave a correct answer to a question that is no longer there. | |||||||||
| 90 | Ex 3.22 | The sat.dat file on the CD is in error. You can get the correct file by going to http://www.uvm.edu/~dhowell/StatPages/Methods/DataMethods5/ | |||||||||
| 164 | Table 6.5 | Gary Stockdale (UC Davis) pointed out that unless society has changed far more than any of us think, the labels on the rows should be reversed. I say it correctly in the text. | |||||||||
| 224 | Line 33 | Albert Smith, at Cleveland State, made a interesting point that power is the probability of rejecting the null when a particular alternative hypothesis is true. Power can only be defined relative to some set of parameters, hence the reference to an alternative hypothesis. | |||||||||
| 252 | Footnote | The data are actually on your disk in Tab9-2.dat. Billy Bowen at the Univ. of Southern Miss. pointed this out to me. | |||||||||
| 254 | Example of radj | The formula should contain (1-.5062) instead of 1-.506)2. Thanks to Pekka Rapeli. | |||||||||
| 266 | first equation | The formula should read [SSY - SSresidual]/SSY Thanks to Dale Berger. | |||||||||
| 266 | Line 3 | It should read 1-r2 instead of r2. Thanks to Albert Smith. | |||||||||
| 277 | second from bottom | The word "slopes" should be changed to "correlations." Thanks to Albert Smith. | |||||||||
| 308 | Middle | Strike the parenthetical sentence that says "(Lines are not drawn when two judges give an object the same rank.) I don't know why I said that--if those lines don't cross other lines, it doesn't hurt to have them, and if they do you need to have them. Thanks to Mike Clark at the Univ. of North Texas. | |||||||||
| 341 | Line 15 | Change to "weaker of the two." | |||||||||
| 342 | Equation for df | The term in the denominator that
is squared should be  . Thanks to Dan
Denis at York University and Ric Luecht at UNC-Greensboro. |
|||||||||
| 355 | Middle | For two groups f' would be equal to d/2, not to d as stated there. Thanks to James Alexander, at the University of Tasmania. | |||||||||
| 399 | 5th from bottom | Change "too large" to "too slowly." Also change last formula to alpha(sub r) = alpha/(k/r) = r*alpha/k. | |||||||||
| 413 | 3rd equation | The denominator should not contain a square root. Thanks to Aaron Shurger at Princeton. | |||||||||
| 413 | Top last full paragraph | There are 25 df for error, and the critical value of F is 4.245. Thanks to Dale Berger. | |||||||||
| 417 | Exercise 12.19 | I made an error in the answer I gave to this question. I gave one from an earlier edition to referred to data that are no longer there. J. C. Pienaar at the University of Natal, Pietermaritzburg, South Africa pointed this out. The correct answer can be found at Correction for Exercise 12-19. | |||||||||
| 426 | Table 13.2 | Larry Cross pointed out that I messed up the calculations for the main effects. The formulae are correct, the answers are correct, and the (10)(5) and (10)(2) are correct. But everything within square brackets [] goes with the other calculation. Karl Wuensch pointed out that for SSC, the mean for younger folks should be 13.16. | |||||||||
| 450 | Line 4 | Change reference to Chapter 16, not 15. | |||||||||
| 431 | Figure 13.4 | The points in that figure are not plotted exactly, but the overall impression is correct. | |||||||||
| 458 | Table 13.12 | The data are correct as given, but the SPSS data file, and some of the others that come on the CD, list the variables as C, B, A when they should have been listed as A, B, C. Thanks to Vamsi Koneru at the Univ. of Miami. | |||||||||
| 463 | Last eq. in part (a) | SSAB at C1 should be SSCells AB at C1... instead of SSCells BC at C1. Thanks to Vamsi Koneru at the Univ. of Miami. | |||||||||
| 468 | Ex13.21 | This problem is right, but the answer in the back is wrong. That answer was based on a previous question that gave cell totals, rather than cell means. Multiply every SS and MS by 25 and you will come out right. Dan Denis caught this one. | |||||||||
| 477 | 9 lines from end | Change SStotal to SSerror. | |||||||||
| 521 | 15 lines from end | Insert "for each difference score" after "to be 0.0." | |||||||||
| 527 | Ex14.7 | The Between Groups covariance matrix that I give
in the complete solution is wrong. (It was based on totals rather than
means.) The correct matrix should be
Thanks to Marty Giesen at Mississippi State. |
|||||||||
| 548 | Line 9 | R should be R2. Thanks to Dale Berger. | |||||||||
| 567 | Line 5 | Gary Stockdale pointed out that what I wrote was incorrect. Andreas Klier has suggested that I mean r0(i.1) instead of r0(1.i) . He is correct, and I think that explains Gary's concern. | |||||||||
| 574-577 | I made a major error here, and I cannot, in good conscience, pretend that it was a typo. I treat the path labeled "b" as if it were the simple direct path from the mediator to the dependent variable. In reality, "b" should be the coefficient from the mediator to the dependent variable after controlling for the independent variable. In other words its value is .323, and its standard error = .323/3.041 = .106. The conclusion are unaffected.A revised Section 15.13 can be found as a pdf file at MedModCorrection.pdf. Thanks to Karl Wuensch. | ||||||||||
| 582 | Line 7 | Martin Le Voi pointed out that it should read "csupport = -15" instead of "csupport = 15." | |||||||||
| 588 | 2nd full para. | Niko Tiliopoulos, at the Univ. of Edinburgh, pointed out an error that resulted from another error. For these data Outome is actually coded 1 = better, 2 = worse. Therefore my description of the direction of prediction is backward. It is trying to predict the higher level of outcome. | |||||||||
| 611 | First matrix | The column heading AA(11) should read AB(11). Thanks to Jerwen Jou | |||||||||
| 613 | Matrix | Upper right entry should be 0.00 insteadof 1.00. Thanks again to Jerwen Jou. | |||||||||
| 651 | Ex 16.9 | Reference should be to Table 16.2, not 6.13. Thanks to Christianne Hawken at Dalhousie. | |||||||||
| 696 | Line 2 | Change to "She had reason to doubt ..." | |||||||||
| 698 | Line 13 | This should read 99.97% instead of 99.93%. thanks to Karl Wuensch. | |||||||||
| 703 | Fig. 18.6 | First there are two typos-->4 should read >= 4, and 397 should read 379. Second, Karl asked why I say 277 <= 4 when 379 + 31 = 410. Apparently it is because of rounding. A 3.5 would be recorded as 4, but is not >= 4. (I didn't write the program.) Ignore "G = 0.5". | |||||||||
| 765 | Cohen(1960) | Dan Denis pointed out that the volume for this reference is actually vol. 20 rather than vol. 10. | |||||||||
| 775 | Ex3.19 | Replace answer with "The mean and standard deviation deviate from what we expect due to sampling error." | |||||||||
| 776 | Ex6.13 | I should have said the odds of voting for civil unions are 2.66 times as high in women. | |||||||||
| 776 | Ex6.27 | Two errors. First, men are more likely to be the primary shopper (at least that's how the data are phrased). Second, it should read the odds of being the primary shopper are 14.06 times higher for men than women. Thanks to Christopher Green at York. | |||||||||
| 784 | Ex13-11 | See note about page 468. Multiply all SS and MS by 25. |
There is a major problem with the data sets. I don't know who to blame, but it
really doesn't matter. The data as found on the CD does not go beyond Chapter
13. I cannot supply complete SPSS, Minitab, etc. data sets, but I can certainly
supply the data in ASCII form, which you can read into whatever program you
have. There is a self-extracting archive of all of the
data found in DataAll.exe. Simply click on this
link and tell the resulting dialog box where you want to store the download--the
desktop is fine. After the file is downloaded, click on the downloaded file and
indicate where you want the data files to be stored. Because this is a
self-extracting archive, you do not need archiving software to open the file.
Once the data are extracted, you can safely delete the archive file itself.
Return to
Dave Howell's Statistical Home Page
Last revised: 01/28/05