Change Log for the Handbook

The initial publication of the CD of the Handbook was generated from the 6/2003 version. We do not document changes made before this date.

Subsequent printings of the CD will incorporate changes documented here. We will list print dates for the CD. If your copy of the Handbook CD does not contain a print date, then it is from the initial 6/2003 printing.

1. On page pri5363.htm, there was a typo in table 3.26. For the Box-Behnken design, the X3 for sequence 1 was changed from a 10 to a 15.

1. On page pri5322.htm, the best solution for Y2 was changed from 157.1 to 1571.05 (this was a typo).

1. Corrected the spelling of Springer-Verlag on page pmc7.htm.

2. Corrected a link for one of the SEMSTAT files.

1. Fixed a typo on page eda33q.htm.

1. On page prc7.htm, corrected the volume number for the Ljung and Box Biometrika paper.

1. On page prc24.htm, fixed the link to the binomial distribution (it was going to the beta distribution).

1. On page pmc435.htm, the formula in "step 1" for the yearly averages (A(p)) did not divide by the number of periods. Corrected the equation.

2. Corrected the spelling for George Kaupas (author of the Course Builder software) on the acknowledgements page and on the toolaids/cb/index.htm page.

3. On page eda35b.htm, reworded the section on the definition of kurtosis.

1. On page pmc3231.htm, fixed a typo in the first equation (the S(r) for the standardized case was missing the "- k").

1. On page prc263.htm, fixed a mistake in the Dataplot macro for obtaining the k-factor for one-sided tolerance intervals. Also rewrote the text for this section.

1. On page ppc333.htm, in the last paragraph, changed "10%" to "10 angstroms".

2. On page pmc625.htm, several of the Dataplot macros were in the wrong directory and so were not found when the user initiated the link. The macros were moved to the correct directory.

1. On page prc3.htm, there was no link to page prc35.htm (as there was to the other four sections). Added the appropriate link.

1. On page mpc3321.htm, the h matrix should have been divided by (1/6) (i.e., 0.5 instead of 3). Corrected this and the resulting change in the computation of the B vector.

2. On page prc263.htm, added a Dataplot macro for obtaining a more accurate approximation for the k-factor for one-sided tolerance intervals.

1. On page pri8.htm, corrected the reference for Miller's Technometrics paper.

1. On page eda35h.htm, fixed two typos.

1. On page apr312.htm, the link to lot acceptance sampling plans was incorrect. Fixed this link.

1. On page apr164.htm, modified the section that demonstrated the Dataplot commands for the lognormal distribution.

2. Since International SEMATECH no longer makes SEMSTAT available on its web pages, we will make the software available on the NIST site (note that NIST does not and will not support SEMSTAT). The page dataplot.htm was modified to include links to the SEMSTAT software.

   In addition, the links to SEMSTAT in pages apr452.htm, pmc449.htm, pmc44a.htm, and prc241.htm were modified to point to the NIST location rather than the International SEMATECH location.

1. On page pri331.htm, added a sentence clarifying i and j in Y(ij).

2. On page apr451.htm, modified the paragraph describing the use of the CHSPPF function in Dataplot and the CHIINV function in Excel for computing the confidence interval. This was an editorial change to provide better clarity.

1. On page eda3662.htm, we modified the description of the maximum likelihood estimators for the uniform distribution.

2. On page pmc332.htm, fixed a typo in an equation (phat(i) = D - i/n was changed to phat(i) = D(i)/n).

3. On page ppc46.htm, fixed a broken link to the process capability section.

4. On page eda33u.htm, fixed a typo for the Weibull plot description (changed ln-ln(1-p) to ln(-ln(1-p)).

5. On page pri33a.htm, the L36 design was incorrect. Corrected 5 columns in row 33 of this design.

1. Fixed a typo on page pmc22.htm (a "the the" was corrected to "the").

1. Fixed a broken link on page prc32.htm.

2. Fixed a broken link on page ppc32.htm.

3. On page eda35b.htm, rewrote the sentences describing left and right skewness.

4. On pages ppc332.htm and prc473.htm, fixed the "Back" and "Next" buttons at the bottom of the page.

5. On pages pmc16.htm, replaced the graph showing Cp for varying process widths (the 1.33 and 1.66 plots were reversed). Also changed the sentence

   A capable process is one where almost all the measurements fall inside the specification limits.

   to

   A process where almost all the measurements fall inside the specification limits is a capable process.

6. For the plots of probability density functions in section 3 of the EDA chapter, the y-axis label was changed from "Probability" to "Probability Density".

7. Changed the deflection value shown on the plot on page pmd61a.htm to be 1.239722 instead of 1.239723 to match the number used in the text.

8. Updated the printer-friendly files on the e-Handbook web site to the latest version of pdf files (May 1, 2006).

9. Updated the printer-friendly files on the e-Handbook web site to include pdf files for all eight chapters.

1. On page pri4.htm, the link to the chapter 3 ANOVA was incorrect. It was changed to go to the correct page.

2. The Back and Next links at the bottom of page prc473.htm were not working correctly. These were fixed.

3. On page apr221.htm, changed "linear in log(x)" to "linear in x".

1. Changed the link associated with the text "lack-of-fit test" on page pmd44.htm to point to the page pmd446.htm rather than the page pmd442.htm.

1. In Table 3.30 on page pri3364.htm, the alpha value was computed incorrectly. The value was changed from 1.681793 to 1.63299.

1. The link for "effects" in the Analysis of Variance definition in the glossary was fixed.

1. Links for "Design new experiment." and "Collect data." were added to the flowchart outlining the development of a process model in the Engineering Questions page. The link for "New model describes the data well?" was deleted since that is a simply a branch point.

2. Corrected several symbols to match the fractionation of the design with the correct resolution on page pri3344.htm.

3. On page pri3364.htm, removed the leading "+" sign from the "Block" column of table 3.29 to be consistent with the other tables on that page.

4. On page prc473.htm, changed the formulation of the degrees of freedom in the example for the Bonferroni simultaneous confidence interval from "4" to "2*2" to make clearer the "2*g" source of the "4".

1. The link on page pmc321.htm to Galit Shmeli's quality control web site was not working correctly. This was corrected.

2. On page eda35a.htm, it was not clear which version of the Levene test that Dataplot was performing. The Dataplot code was modified to include a line in the title of the output indicating which variant of the test is being performed. Added a clause on page eda35a.htm to indicate that the default in Dataplot is to perform the median based test.

3. On page apr172.htm, fixed the equations nhpp3.gif and nhpp4.gif.

4. On page eda35g.htm, corrected the test statistic for the Kolmogorov-Smirnov goodness of fit test. Also, added a note explaining the difference between the old and new formulas.

1. Page pmc32.htm contained a typo in the Fractional Factorials defintion. This actually encoded an equation, a line of text, and another image as a single image (the typo was in the line of text). This was modified by breaking the single image into two separate images for each of the equations and entering the text line as regular text (with the typo corrected).

2. Fixed broken links on pages ppc231.htm and ppc2311.htm.

3. Fixed typo on page prc13.htm.

4. On page prc14.htm, fixed the link to the standard normal table and fixed the bottom navigation links.

5. On page prc15.htm, fixed the bottom navigation links.

6. On page prc31.htm, fixed the bottom navigation links.

7. On page pmc541.htm, added links to the formulas for the mean and variance and added the formula for the covariance. This was in response to a user request to include the computational formulas in order to work out the example for the variance-covariance matrix.

8. On page prc242.htm, corrected the equation and the bottom of the page. Specifically, the value of delta in the numerator was changed from 0.15 to 0.10 (to reflect the text above) and the result was changed from 11 to 77.

9. On page eda353.htm, rewrote the paragraph that described the formulation of the null hypothesis for the case where we are testing that the difference between the two means is equal to a constant. This was done to provide better clarity.

1. Page prc16.htm contained an erroneous link for the median. This was corrected.

2. Joseph Voelkel of the Rochester Institute of Technology pointed out that our formulation for the Grubbs test was in fact two distinct tests (one for the minimum and one for the maximum). This means that the significance level needed to be divided by 2 to obtain the correct critical values.

   Page eda35h.htm was updated to make the distinction between these one-sided and the two-sided tests explicit and to provide the correct formula for the critical values in each case. The Dataplot software was modified to produce correct critical values and to optionally print the one-sided tests.

1. Page pmc231.htm contained a link to Galit Shmueli's web site for online quality control. Galit informed us that the site has been moved. The link was modified to point to the new site.

2. Fixed some links on page pmc55.htm. Specifically, the "Goodman" link and the links on the bottom navigation bar were corrected.

1. Updated dpbrows.htm to include instructions for configuring the Firefox browser to run Dataplot under Windows.

2. On page 6.5.4.2, the formula for the multivariate normal distribution was incorrect. The corrected formula was inserted.

3. Fixed typos on pages pmc.htm and pmc52.htm.

4. Clarified the formula for the autocovariance function on page eda331.htm. Added a reference (to the Chatfield Time Series text) in eda43.htm.

5. Fixed a typo in the title of page pmc4462.htm.

6. The numerical example on page mpc3321.htm was originally computed using single precision. The calculations were redone using double precision. This slightly modified the values for Y'(I-XQX')Y, s₁ and s_item3.

7. On page apr164.htm, changed LET SD = .5 to LET SIGMA = .5.

1. On page pmc435.htm, the equation for the initial values for the trend factor was missing some parenthesis. This caused some confusion in the meaning of the equation. A corrected version of the equation was inserted.

2. Updated the printer-friendly files on the e-Handbook web site to the latest version of pdf files (January 26, 2005).

1. On page eda3676.htm, the example gave the value for the 1% critical value rather than the 5% critical value. This was corrected.

2. Page pri3347.htm links to a number of files containing various designs. The file containing the 2^9-4 design had incorrect values in the X9 column. This column was corrected.

1. The link in the block plot page (eda333.htm) to the binomial distribution was going to the beta distribution. This was corrected.

2. The sample Dataplot output for the Grubbs test (eda35h.htm) was corrected (the critical values were generated for a two-sided test when they should have been computed for a one-sided test).

3. The Josephson junction cryothermometry case study (eda424.htm) was corrected to use all 5 columns of the data instead of just the first column. There were numerous changes in all pages of the case study.

4. The case study (eda429.htm) was modified to incorporate some additional background information provided by Ed Fuller. The only substantive change was to replace the power lognormal with the fatigue life as one of the candidate distributions.

1. NIST purchased their own web search engine . This replaces the FirstGov search engine that only allowed us to search the /www.itl.nist.gov/ web space rather than the restricted e-Handbook pages in /www.itl.nist.gov/div898/handbook/ web space. The new search software allows us to restrict the search to the /www.itl.nist.gov/div898/handbook/ web space.

2. On page mpc36.htm, the sentence
   The loadcell measurements (shown as *) are plotted on the y-axis against the corresponding values of known load shown on the y-axis.

   was changed to

   The loadcell measurements (shown as *) are plotted on the y-axis against the corresponding values of known load shown on the x-axis.

1. The example on page mpc3321.htm (section 3 of the MPC chapter) had an error in the A inverse matrix. This error propogated throughout the remaining computations on the page. Therefore significant edits were made to the example. In addition, some editorial corrections were made to the page.

2. Page pri3347.htm links to a number of files containing various designs. The file containing the 2^8-3 design was incorrect and was replaced with the correct design. The resolution listed in the file containing the 2^7-3 was incorrect and was replaced with the correct resolution. The link to the 2^8-1 design listed the wrong resolution (the resolution given in the file was correct) and was corrected.

1. The link on page eda3668.htm to Weibull maximum likelihood had a typo. This was corrected.

1. Page ppc231.htm (section 2 of the PPC chapter) corrected two links in the "testing" paragraph.

2. In section 3 of the MPC chapter, the gif files mpc352a.gif, mpc372a.gif, and mpc372b.gif seem to have corrupted for some unknown reason. This was causing some werid formatting in some of the probability distribution tables. Good copies of these gif files were restored.

1. Page mpc365.htm (section 3 of the MPC chapter) corrected the value of F(.05,8,20).

2. Corrected the formatting for the nu1 entries on the tables for the critical values of the F distribution on page eda3673.htm (section 3 of the EDA chapter).

3. The link to the SEMSTAT program was changed to its new site. The pages dataplot.htm and prc241.htm were modified.

4. Fixed a typo in page eda35e.htm (Anderson-Darling test).

1. Page pmd134.htm (section 1 of the PMD chapter) had a broken link with the anchor text "central composite design". The link was corrected to point to pri3361.htm.

2. Page pmd134.htm (section 1 of the PMD chapter) had a broken link with the anchor text "Chapter 5: Process Improvement". The link was corrected to point to pri.htm.

3. Page pmd134.htm (section 1 of the PMD chapter) had a broken link with the anchor text "Section 6". The link was corrected to point to pri53.htm. The anchor text was changed to "Section 5.5.3".

4. Page pmd141.htm (section 1 of the PMD chapter) had a link with anchor text "especially with respect to outliers" that pointed to the wrong page. The link was changed to point to pmd445.htm#nppi.

5. Page dep/dep144.htm (under section 1 of the PMD chapter) had a missing graphic. The correct graphics file was added to pmd/section1/plots/.

6. Page pmd442.htm (section 4 of the PMD chapter) had two typos in the first sentence in the second to last paragraph. "be adequately" was changed to "adequately" and "the the" to "the".

7. Page dep/dep443b.htm (under section 4 of the PMD chapter) had a missing graphic. The correct graphics file was added to pmd/section4/plots/.

8. Page pmd444.htm (section 4 of the PMD chapter) had a broken link with the anchor text "general discussion". The link was fixed by adding the correct section reference (section3). in the EDA chapter.

9. Page pmd444.htm (section 4 of the PMD chapter) had a broken link with the anchor text "Process Monitoring". The link was fixed by adding the correct chapter reference (pmc) to the link address.

10. Page pmd445.htm (section 4 of the PMD chapter) had a broken link with the anchor text "general discussion". The link was fixed by adding the correct section reference (section3). in the EDA chapter.

11. Page pmd445.htm (section 4 of the PMD chapter) had a broken link with the anchor text "gallery of graphical techniques". The link was fixed by adding the correct section reference (section3) in the EDA chapter.

12. Page pmd445.htm (section 4 of the PMD chapter) had a confusing sentence concluding the second paragraph.

    "Curvature or points that are far off the line indicate that the random errors are probably not normally distributed, or that the data has some outliers, in it respectively."

    was changed to

    "Distinct curvature or other signficant deviations from a straight line indicate that the random errors are probably not normally distributed. A few points that are far off the line suggest that the data has some outliers in it."

13. Page pmd452.htm (section 4 of the PMD chapter) had four missing equation files. The missing files, sigmaeqg1.gif, sigmaeqg2.gif, sigmaeqmu.gif, and sigmaeqx.gif were copied back to the appropriate equations directory section4/eqns.

14. Page howto.htm was not not interacting correctly with the navigation frame on the main pages of the e-Handbook. The links in howto.htm were changed so they opened new files in the primary browser window.

15. Page acknwldg.htm was not not interacting correctly with the navigation frame on the main pages of the e-Handbook. The links in howto.htm were changed so they opened new files in the primary browser window. Affiliation links were also corrected for some authors.

16. Page search.htm was not not interacting correctly with the navigation frame on the main pages of the e-Handbook. The link in navi.htm was changed so it opened search.htm in the primary browser window.

17. Fixed several inconsistent links in the navigation bars on pages linked to from the main e-Handbook page and under "Tools & Aids".

18. Added the page toolaids/cd.htm to provide information on the CD version of the e-Handbook. A link from the main page was added as well.

1. Page apr311.htm (section 3 of the APR chapter) had a few errors. Specifically, the table entries for r=20 and confidence levels 90% and 95% were incorrect. The Excel formula was modifed to use 1 - alpha rather than alpha.

2. Page eda3669.htm (section 3 of the EDA chapter) made a slight modification to the general pdf formula for better clarity.

3. The Course Builder scripts are now available for download. Added links to both installation documentation and usage documentation for the Course Builder.

   Note that the Course Builder is not active on the NIST public web pages. Rather you need to install the e-Handbook on a local server and install and run Course Builder on this local platform.

1. Updated the printer-friendly files on the e-Handbook web site to the latest version of pdf files (July 1, 2003).

2. Fixed typos in the PPC chapter (ppc134.htm and ppc332.htm).

3. The macro listed for the case study on page pmc625.htm (section 6 of the PMC chapter) contained an earlier version of the macro that was not aligned with the current version of the case study. The old macro was replaced with the new version.
4. The formula for the multivariate normal distribution on page pmc542.htm (section 5 of the PMC chapter) was incorrect. The correct formula was inserted.

1. In section 3 of the MPC chapter, page mpc365.htm contained links to pages mpc3653.htm and mpc3652.htm for the F-table and t-table. These pages are no longer part of the Handbook. The links were corrected to go to the F-table and t-tables in the EDA chapter.

2. In section 5 of the MPC chapter, page mpc55.htm had a typo in a link. This was corrected.

3. In section 5 of the PMC chapter, page pmc541.htm used both height and weight for one of the variable names. This was corrected to use height for all mentions of this variable.

4. The wrong content for page mpc542.htm (in section 5 of the PMC chapter) was inadvertently inserted. This was replaced with the correct content.

5. A typo was corrected in the Box-Cox sample data on page mpc52.htm (in section 5 of the PMC chapter).