The errata list is a list of errors and their corrections that were found after the product was released.
The following errata were submitted by our customers and have not yet been approved or disproved by the author or editor. They solely represent the opinion of the customer.
| Version |
Location |
Description |
Submitted by |
Date submitted |
| Printed |
Page 162
middle calculation on top of page |
Author previously tried to correct this calculation:
"...and solution should be 23.51 (not 25.73) Top of page 162, formula, first line numerator should be y_d-bar (a bar should be over the d) numbers should be 4.8-0 / 23.51/sqrt(10) = 0.646 ."
The author incorrectly used the sample variance s^2 = 23.51 in the calculation of standard error s. Looks like it should be:
=4.8 - 0 / 4.849 / sqrt(10) = 3.13.
Seems like the line "Thus, you would fail to reject the null hypothesis in this experiment." should be changed because 3.13 > 1.833 and 2.821.
|
Anonymous |
May 05, 2012 |
| Printed |
Page 168
calculated value of t |
Author or editor previously corrected the t calculation by saying the value of 0.94 should be 0.88. I believe this to still be incorrect. The value for Sd^2=0.94 is correct. But s= sqrt(0.94) = 0.97 and not 0.88 as stated. This changes the answer to 4.88.
|
Anonymous |
May 05, 2012 |
| Printed, PDF |
Page 195
2nd paragraph, after table |
The text indicates that "The computed chi-square value is 21.7 with 3 degrees of freedom". Given the data in Table 10-6, the chi-square value should be 10.128.
We can verify in R:
x <- matrix(c(10,40,30,20,12,25,50,13), nrow = 2, byrow = TRUE)
> chisq.test(x)$statistic
X-squared
10.12821
|
Titus Barik |
May 08, 2012 |
| Printed |
Page 217
1st Paragraph |
Where H = 2.26 (from page 216), the book states, "To calculate the significance of H, you compare it to [Chi-square] at the p = 0.05 level, for df = 8, which is 15.51; the null hypothesis can be rejected, since at least one team outperforms the others."
The last time I checked, 2.26 did not exceed 15.51, thus warranting acceptance of the null hypothesis.
|
E. Joesph Nolan IV |
Apr 22, 2013 |
| Printed |
Page 217
1st Paragraph |
For three cases, the book states, "...for df = 8, which is 15.51...."
This should state "df = 2, which is 5.991...."
|
E. Joseph Nolan IV |
Apr 22, 2013 |
| Printed |
Page 242
SS Within equation |
The first number entered for the SS Within formula and the resulting value are wrong:
As SS Within = SS Total - SS Between, and SS Total = 3,963.75 (line above), it should read "SS Within = 3,963.75 - 616.05", instead it reads "SS Within = 200,594 - 616.05", resulting in a faulty SS Within value of 3,963.75 instead of 3347.7
For the F-value calculation below, the correct SS Within value must have been used to get 3.31.
|
Anonymous |
Nov 21, 2012 |
| PDF |
Page 242
last paragraph |
The critical values for F1,18 are incorrect.
Text reads "Thus, the critical value for F1,18 at p = 0.05 is 3.16; as a result, you would reject the null hypothesis on this occasion. Note that the critical value for F1,18 at p = 0.01 is 5.092."
The values reported seem to be for F3,18, not F1,18. According to the NIST tables of Upper Critical Values of the F Distribution (http://www.itl.nist.gov/div898/handbook/eda/section3/eda3673.htm), the values should be 4.414 at p = 0.05 and 8.285 at p = 0.01. The true values indicate the opposite conclusion; you would _not_ reject the null hypothesis on this occasion. The p value in this case should be 0.08543, well above the 0.05 criterion.
|
Myq Larson |
May 25, 2014 |
| PDF |
Page 242
Last 2 paragraphs |
The equation for SSwithin is stated as
SSwithin = 200,594 -616.05 = 199,977.6
It should be:
SSwithin = 3863.75 - 616.05 = 3247.7
Then the F ratio calculation corresponds
Also the F-ratio critical values stated are both incorreect.
F (1,18) at p=0.05 is 4.41, not 3.16 as stated.
And at p=0.01 it is 8.29, not 5.092
Therefore the conclusion is incorrect too.
Incorrect calculation + incorrect conclusion = confused and annoyed reader.
|
Catherine Stanley |
Mar 28, 2016 |
| Printed |
Page 251
Table at top of page |
Is the "Steps" column in this table correct? It looks suspicious because 1. it is categorical data and the text on pg 250 seems to imply that we'd be counting number of steps, 2. It is an exact repeat of the "Age" column in the previous exercise and 3. I cannot get my MANOVA output to match yours using this data.
|
Anonymous |
Jul 27, 2010 |
| Printed |
Page 281
1st paragraph |
Coefficients for regression are all listed as identical (equal to 0.68), but the table above shows unique values (0.39, 0.015, -0.011, & 0.067)
|
Anonymous |
Nov 19, 2011 |
| Printed |
Page 283
1st paragraph |
A one stddev change in 'Size' misreferenced as resulting in throughput accuracy of 0.385 stddev; should be accuracy instead of size. Scanner size was referenced two times.
|
Anonymous |
Nov 19, 2011 |
| Printed |
Page 293
bottom |
Table 12-2 indicates that the data is the Pearson r correlation of the data contained in table 12-1. Calculation of this in R does not correspond to what is in the book.
intel <- data.table(reading=c(8, 5, 2, 8, 10),
music=c(9, 6, 3, 9, 7),
arithmetic=c(6,5,2,10,1),
verbal=c(8, 5, 6, 9, 10),
sports=c(5,6,8,8,5),
spelling=c(9,5,6,10,10),
geometry=c(10,5,4,6,2))
cor(intel, method = 'pearson')
reading music arithmetic verbal sports spelling geometry
reading 1.00000000 0.8210528 0.19271857 0.85495984 -0.5897678 0.8513059 0.07537784
music 0.82105281 1.0000000 0.67053162 0.61007114 -0.3707390 0.7277995 0.55506862
arithmetic 0.19271857 0.6705316 1.00000000 0.08795839 0.3422993 0.2692150 0.60066251
verbal 0.85495984 0.6100711 0.08795839 1.00000000 -0.2543852 0.9767379 -0.08941021
sports -0.58976782 -0.3707390 0.34229925 -0.25438520 1.0000000 -0.2108706 -0.15559398
spelling 0.85130588 0.7277995 0.26921504 0.97673789 -0.2108706 1.0000000 0.10780492
geometry 0.07537784 0.5550686 0.60066251 -0.08941021 -0.1555940 0.1078049 1.00000000
Manual computation of reading vs. music confirms what is shown above and does not match what is in the book.
ms <- intel$music
rd <- intel$reading
sum((rd - mean(rd)) * (ms - mean(ms)))/((length(rd)-1)*sd(rd)*sd(ms))
[1] 0.8210528
|
Keith Chewning |
Sep 03, 2015 |