Errata

When discussing the Kappa formula the 3rd paragraph says:
"Kappa has a range of 0-1: ..."

Right below this paragraph it is discussed how kappa values can be classified. The first item is for values below 0 (< 0).

The truth is that kappa has not the range of 0-1, but -1 to 1, as the number of cases in agreement do not need to be higher than the number of expected agreements.

First of all, permutations ARE all the possible WAYS... and the second, set has no order so permutations are made OF a set, not IN a set.

Note from the Author or Editor:
first sentence should read "permutations are all the possible ways..

second correction has already been made

the denominator calculation is wrong

(0.01)(0.99) is 0.0099 not 0.00099

Note from the Author or Editor:
The denominator of the second fraction should be:

0.0095 + 0.0099

When using the Greek word 'mu', the symbol for sigma is shown in parentheses immediately following.

Please tell the authors 'Thanks for an excellent book on a very important and
interesting topic'.

Note from the Author or Editor:
Change the symbol to mu (not sigma)

error:
requirement for a bar graph is that there can only be one y-value for each x-value,

correction:
requirement for a line graph is that there can only be one y-value for each x-value,

error:
bar chart, as in Figure 4-16.

correction:
line chart, as in figure 4-16.

error:
decrease the visual impact of the trend

correction:
inflate the visual impact of the trend

error:
Obesity amog U.S. adults, 1990-2002 (CDC), using a large range to inflate

correction:
Obesity amog U.S. adults, 1990-2002 (CDC), using a large range to decrease

The mean of the last normal distribution should be -2 instead of 2

Note from the Author or Editor:
correct: this line of legend for Figure 7.1 should be "mu = -2, sigma = 0.71"

The ?-value for the third graph should be -2. Not 2.

Note from the Author or Editor:
In figure 7-1, p. 127, third line of legend: mu = -2 (not 2)

I think word "variance" is meant to be "standard deviation" (as indicated by the use of the sigma symbol).

Note from the Author or Editor:
correct: the text should read "and standard deviation sigma (the Greek letter sigma)

Rather than "...a value of 10..." it should read ".... a value of 110...."

Note from the Author or Editor:
Correct: the sentence should read "Similarly, a value of 110 from this population..."

error:
Similarly, a value of 10 from this population has a Z-score of 2

correction:
Similarly, a value of 110 from this population has a Z-score of 2

You ask the reader to refer to "the later chapters relating to specific study designs" to learn about control variables. Since there are no such chapters, where in this book is this information? The index (which is really, really bad, by the way) only points to this very page 132, which is not very helpful.

While we are at it: if this book is supposed to be a text book, the explanations are way to bad and the mistakes are too many. If this is supposed to be a reference book, the index just doesn't cut it.

Note from the Author or Editor:
Remove the clause beginning "and leave the discussion.." in the second paragraph under "Independent and Dependent Variables" sothe sentence ends with "..independent and dependent variables."

The problem gives two "scores" of 190 and 175. I thought it would mean "Z-score", since "score" is not normally used in the book for values or data points. So 190 should be more extreme. But the solution gives the solution to the confusion: 190 was meant to be a data point.

Note from the Author or Editor:
Add "raw scores" to problem 2 p. 150 as follows: "Which of the following raw scores has a more extreme Z-score...

"These relations would usually be expressed as t_{0.05,20}=1.725 and t_{0.05,20}=1.725, respectively."

I'd bet, that should read "... and t_{0.01,20}=2.528, respectively." Otherwise the sentence doesn't make any sense.

Figure caption says "Comparison of the normal and t distribution for v=5, 15, and 25," but I see only one curve.(or am I just misinterpreting the graph?)

Note from the Author or Editor:
see previous errata for resolution

In the text: "s? = 0.75". In the formula, 0.75 is inserted for s, so either s=0.75 or the formula should contain the square root of 0.75.

Note from the Author or Editor:
resolution provided in other erratum

The first line of the last paragraph reads "The recruiter finds that average accuracy a=79% and s^2=0.75..." It should read "and s=0.75", not s^2.

The formula states that 24.0 = 2.4, which is a bit far-fetched.

Note from the Author or Editor:
Remove = from the second line of formula and cut third line to the right of the second line,ie,

E y 24.0
- = ---
n 10

The second line of the equation ("= sum(y)/n") should be prefaced with "mean", otherwise the equation would read (carrying down from first line) "Sum(y) = sum(y)/n", which doesn't make any sense.

Further, in the second set of equations below "The variance can then be computed as" the third line should read "59.12 - 24^2/10", not "2.4^2/10".

Note from the Author or Editor:
Change "2.4^2/10" to "59.12 - 24^2/10"

CI_0.95 = ... = 2.0 +- ...

Why 2.0?

According to Wikipedia, this should be 2.4.

Note from the Author or Editor:
see previous errata for resolution

The formula should have y bar instead of y.

.025, 9 should be written as a subscript in parentheses
t(0.025, 9).

The foruma should use 2.4 instead of 2.0 the correct values are:
2.1060163<= mu <= 2.6939837

Note from the Author or Editor:
see previous errata for resolution

The first sentence reads "Thus, mu_ballet = 87.95, mu_football = 32.38...". Since these are sample means, shouldn't this read "Thus ybar_ballet = 87.95, ybar_football = 32.38..."?

Note from the Author or Editor:
the sentence should read "Thus, y-bar_ballet = 87.95, y-bar_football = 32.38..." (using the symbols)

The equation for degrees of freedom contains an error in the denominator. s1 and s2 should both be squared inside the brackets.

Note from the Author or Editor:
Change s1 -> s1^2 and s2 -> s2^2

I would agree with the equation for the t test, but the wrong number is plugged in from what is given in the question. In every other book that I have read, variance = (std deviation)^2. The problem gives what appears to be a variance (ie. s^2 = 3.5), so the standard deviation is the 1.87 (= sqrt(3.5)). This then would change the result of the t score.

A similar error occurs for the example on page 168 when computing the t score.

Note from the Author or Editor:
on p.166 change s2=3.5 to s=3.5

on p.168, change the worked solution to:

= 1.5-0
-----
0.88 <---- this is the only change
---
sqrt(10)

= 5.39

The axes on the graph are not quite in proportion; the distance from the 0 to 10 tick on the x axis is 5.5cm; the distance from the 0 to 10 tick on the y axis is 6cm. It doesn't affect the data, but visually, it just seems subtly "off."

Note from the Author or Editor:
new figure emailed to Marlowe

error:
a = -2, which is positive, so the relationship is positive.

correction:
a = -2, which is negative, so the relationship is negative.

Note from the Author or Editor:
see previous errata for resolution

"a = -2, which is positive, so the relationship is positive".

That's the problem when you copy-and-paste text from the paragraph before: -2 is negative, so the relationship is negative.

Incidentally, my relationship with this book is beginning to become negative, too.

Note from the Author or Editor:
change "which is positive, so the relationship is positive" to "which is negative, so the relationship is negative"

The variable d (sixth column) is obviously the difference between ranks. However, there are several issues with its use here:

1. The variable is not identified as such after the defining equation on page 183.

2. The difference is ostensibly computed as [Column 5]-[Column 4], which is inconsistent with the method in which differences were computed earlier (see Tables 8-3 and 8-4).

3. The incorrect sign is used for the rank difference for Vehicle D (should be 1 instead of -1).

Note from the Author or Editor:
p. 183 2nd paragraph, 2nd sentence add "(d)" after "differences in ranks" so it readys "So, a lot of the computational effort involves working out the differences between ranks (d) for individual items..."

p. 183 just below the formula, add "Note that because we are squaring d, it doesn't matter which rank is subtracted from which, as long as you are consistent." (before the sentence starting "Let's revisit the relationship..."

p. 184 table 9.8 For vehicle D the value for d should be 1 (not -1)

Second row, second column should be Cell22 and not Cell21.

Note from the Author or Editor:
Correct: the fourth cell should read Cell22 (22 is subscript)

Currently: "In this case, 21.7 is a highly significant value ..."

It should be 45.5 (the sum), not 21.7 (single value of one cell).

Note from the Author or Editor:
the sentence should read "In this case, 45.5 is a highly significant value..."

'Player A' combined batting average given as '323', should be '0.323'

Note from the Author or Editor:
In table 10-14, third row last column the number should be 0.323, not 323

"medians for the different rates are $164,331, $199,628..." and so on till the end of the line. Where these numbers are coming from? The never used later either.

Note from the Author or Editor:
p. 234 2nd paragraph modify first sentence to read: "From the graph, you can see that there is a general rise in house prices with rising interest rates; in fact the medians for the different rates are $164,331, ..."

Problem expanding y_ij = mu + alpha_i + epsilon
with alpha_i = y_i_bar - y_i. Table 12-3 actually shows values that would be correct if we think that alfa_i = y_i_bar - mu
Also, it's not clear where epsilon values come from.

Note from the Author or Editor:
p. 236 1st formula: change y_i to y_bar

p. 236 just above Table 12-3, after "as shown in Table 12-3." add "Note that the epsilon values are simply the difference, or error, between the price as predicted from the grand and group means, and the actual, observed price."

The equation for the MH common odds ration is not correct. In general, the dividend of the sigmas is not equal to the sum of the dividends of the individual sigma terms. Perhaps the right hand side should be something more like this?:

(40x45)/150 + (60x25)/150
----------------------
(30x35)/150 + (50x15)/150

Note from the Author or Editor:
The correction is correct. There should be a single numerator and single denominator rather than two fractions added together.

You already introduced the Z-score on page 128, complete with exercises on p. 150.

Note from the Author or Editor:
Revise the beginning of first sentence of the first paragraph under "Standardized Scores" as follows:

The standardized score, also known as the normal score of the Z-score (further discussed in chapter 7), transforms..."

and drop the clause "which is discussed in Chapter 7." from the end of this sentence

In the first example under "Properties of Roots", it is possible to take the cube root of a negative number, so a and b do not need to both be greater than zero in all cases. Do you mean m and n must both be >= 0?

In the last example under "Properties of Roots", there is no variable b. Again, do you mean m and n?

Note from the Author or Editor:
On the first example, drop the phrase "where both a and b >= 0"

On the second example, it should read "where b is not equal to 0" (using the "not equals" symbol"

On the third example, it should read "if n and m are positive integers and a >= 0)"

"If a = b, then a + b = a + c" should read
"If a = b, then a + c = b + c"

On page 397, in the first bullet point under "Solving Equations" the example equation (If a = b, then a + b = a + c) is wrong. It's supposed to be an example of adding a constant to both sides of an equation, so it should be something like a + c = b + c.

Note from the Author or Editor:
correct: line should read: "If a = b, then a + c = b + c (changing a constant...."

"...if a = 5 and b = 6, then a < 5 and a < b are both true..."

First inequality is false.

Note from the Author or Editor:
Change to "then a < 6 and a < b are both true.."

The SPSS Programming and Data Management book is not out of print. The current version is always available as a PDF from http://www.spss.com/statistics/base/data_management_book.htm. You can also order a printed copy of the book.

Note from the Author or Editor:
change to "(which may be downloaded in PDF format or ordered in book format from http://www.spss.com/statistics/base/data_management_book.htm)"