ࡱ>   iRoot Entry F0xؼ@WordDocumentŨObjectPooln$tؼtؼSummaryInformation(  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~E; PICTfo3 m FObjInfo5 FCompObj6 F jObjInfo7 F      !"#$%&'()*+,-./0123456789<=>?@ABCDFZGHIJKLMNOPQRSTUVWXY[\]^`abcdefghklmnopqrstuvwxyz{|}~        !"#$%&'()*+,-./01234@789:;<=>?ABCDFGHIKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvyz{|}~Calculator - Random Number Functions 12 Calculator - Probability Functions 15 Analyze Menu - Distribution of Y: Continuous Variables 17 More Moments 17 Stem and Leaf Plot 18 Test Mean=value 18 Outlier Box Plot 19 Capability Analysis 19 Normal Quantile Plot 19 Analyze Menu - Distribution of Y: Nominal/Ordinal Variables 21 Test Probabilities for Nominal Distributions 21 Popup options for Frequency Table 21 Analyze Menu - Fit Y by X: Continuous by Continuous 22 Nonpar Density 22 Fit Transformed 22 Other Changes to the Fit Y by X Platform for Continuous Y and X 23 Analyze Menu - Fit Y by X: Continuous by Nominal 24 Normal Quantile Plot 24 The Matching Variables Command 24 Means Diamonds Have 95% Overlap marks 26 Other Changes to the Fit Y by X for Continuous Y and X 26 Analyze Menu - Fit Y by X: Nominal/X by Nominal Y 27 Cochran-Mantel-Haenszel Test 27 Analyze Menu - Fit Model 28 Standard Least Squares Parameter Estimates Details 28 Screening: Prediction Profile 28 Screening: Effect Screening Option for Coded Estimates 29 Screening: Contour Profiler 31 Analyze Menu - Fit Nonlinear 32 Analyze Menu - Correlation of Ys 33 Analyze Menu - Cluster 33 Graph Menu - Control Charts 33 Variability Analysis (Gage R&R Charts) 33 Gage R&R Charts 35 Gage R&R Variability Report 35   EMBED Word.Picture.6  Welcome to JMP JMP is statistical discovery software that can help you explore data, fit models, discover patterns, and discover points that dont fit patterns. As statistical discovery software, the emphasis in JMP is to interactively work on data to find out things. Using graphics, you are more likely to make discoveries. You are also more likely to understand the results. With interactivity, you are encouraged to dig deeper, for one analysis can lead to a refinement, one discovery can lead to another discovery; and you can experiment with statistics to improve your chances of discovering something important. With a progressive structure, you build context that maintains a live analysis, so you dont have to redo analyses, so that details come to attention at the right time. The job of software is to create a virtual workplace. The software has facilities and platforms where the tools are located and the work is performed. JMP provides the workplace that we think best for the job of analyzing data. With the right software workplace, researchers will celebrate computers and statistics, rather than avoid them. JMP aims to present a graph beside every statistic. You can and should always see the analysis in both ways, statistical text and graphics, without having to ask for it. The text and graphs stay together. JMP is controlled largely through point and click, mouse manipulation. If you click on a point in a plot, JMP identifies and highlights the point in the plot, also highlights the point in the data table, and highlights it everywhere else the point is represented. JMP has a progressive organization. You begin with a simple surface at the top, but as you analyze you see more and more depth. The analysis is alive, and as you dig deeper into the data, more and more options are offered according to the context of the analysis. In JMP, completeness is not measured by the feature count, but by the range of applications, and the orthogonality of the tools. In JMP, you get more feeling of being in control in spite of less awareness of the control surface. In JMP you get a feeling that statistics is an orderly discipline that makes sense, rather than an unorganized collection of methods. The statistical software package is often the point of entry into the practice of statistics. JMP endeavors to offer fulfillment rather than frustration, and empowerment rather than intimidation.  EMBED Word.Picture.6  System Hardware and Software Requirements Windows Operating System Microsoft Windows NT 3.51 or greater Microsoft Windows 95 MS-DOS 5.0 or later with Microsoft Windows 3.1 Machines Supported IBM-compatible personal computer with a 386 or higher microprocessor Math Coprocessor A math coprocessor is not required Distribution Media Diskettes Memory 8 MB minimum with 8 MB of swap file space 16 MB recommended with 10 MB of swap file space Note: The memory requirements are in addition to the amount of memory needed to run the operating system. Hard Disk Storage 6 MB minimum (without Win32s) 8 MB minimum (with Win32s) 8 MB recommended (without Win32s) 10 MB recommended (with Win32s) Macintosh Operating System System 7.0 or greater Machines Supported 68k Macintosh: JMP runs on any MacPlus or later Macintosh model with a 680x0 processor. PowerPC Macintosh: JMP runs on any PowerPC Macintosh processor Math Coprocessor A math coprocessor is not required. Distribution Media: Diskettes Memory 4 MB minimum 8 MB recommended Note: The memory requirements are in addition to the amount of memory needed to run the operating system. Hard Disk Storage 4 MB minimum 8 MB recommended Before you Install Before you install JMP be sure to make backup copies of your JMP diskettes. To protect your software, lock the JMP installation disks by sliding the locking tab on each disk so that the hole in the disk is open. Then make backup copies of the disks before you begin the installation process. If you need additional information on locking or copying disks, see the documentation that came with your Macintosh or Windows computer Register Your Product If you are a new user please take a moment to complete and mail the postage paid product registration card for your new software. Keep your portion of the card with your serial number in a safe easily accessible place. As a registered JMP owner you are eligible for: free technical support software updates JMPer Cable, JMPs quarterly newsletter a free JMP T-shirt or JMP mug  EMBED Word.Picture.6  Installing JMP The JMP package contains program disks that have installation instructions printed on the disks label. Windows Installation To install JMP under Windows insert Disk 1 and double-click on the INSTALL.EXE file. The Windows version of JMP runs best on Windows 95 or Windows NT. When run on Windows 3.1 a special system library called Win32s is required to allow for the 32-bit application. If the Win32s library (version 1.25 or later) has not been installed, you must run the Win32s setup provided in your JMP package before installing JMP software. Macintosh Installation To install JMP on a Macintosh, insert Disk 1 and double-click on the JMP Installer. Follow the instructions to complete a standard install. We recommend that you do a standard install the first time you install JMP.  EMBED Word.Picture.6  Customer Support Sales Support Sales support for JMP is available by calling SAS Institute at 919-677-8000 and ask for JMP sales. Hours of operation are 9:00 am to 5:00 pm Eastern Standard Time. Training and Education JMP Training and Education information is available by calling SAS Institute at 919 677-8000 x7312 during hours of operation, 9:00 am to 5:00 pm Eastern Standard Time. Technical Support Technical support is provided for as long as you license the software (annual license user) or for one year after purchasing the software provided you have returned your registration card (you are a registered user) and continues when you purchase upgrades. JMP technical support is divided into JMP statistical support and JMP system support (nonstatistical). If you encounter errors or have questions regarding the installation, you may contact Technical Support via the World Wide Web, telephone, fax, dial-up computer access, or e-mail. For technical support via the World Wide Web, use the following URL: http://www.sas.com/ts/ For technical support by phone, call (919) 677-8008 or (919) 677-8118 between 9 a.m. and 5 p.m., Eastern Standard time, on SAS Institute business days. The fax number for technical support is (919) 677-4444. Electronic mail support is available through the Electronic Mail Interface to Technical Support(EMITS). This facility allows you to track a technical support problem or add information to a previously reported problem via email. To obtain more information on EMITS, send electronic mail to SUPPORT@SAS.COM with the body of the message containing the command: help JMP Information on the World Wide Web All the information in this section is also available on the world wide web at the URL: http://www.JMPdiscovery.com/ This web page also gives you the ability to download a free JMP demo download current patches see a technical overview of JMP statistics see a description of all training courses available find out more about purchase and license options  EMBED Word.Picture.6  Overview of Enhancements in JMP Version 3.2 This overview lists changes and enhancements between JMP version 3.1 and version 3.2. Following sections describe most of these features in more detail and give examples. General This version of JMP for Windows contains an experimental real-time data capture feature. This feature is only available when JMP is running under Windows 95 and Windows NT 3.51 or higher. If you have an instrument connected to one of the COM ports of your computer and create a new column in your spreadsheet with its data source set to "Instrument", then JMP will attempt to read the data which is sent by your instrument to the COM port and enter it into the new column. In order for this feature to work properly, when you set the data source of the new column to "Instrument", you must correctly configure the COM port information in accordance with the properties of your particular instrument. This feature is experimental and is not guaranteed to work with all instruments. The Import Command can read FACS files. The FACS file import reads files from flow cytometery software packages that write their data files in the standard FCS format proposed by the International Society for Analytical Cytology. FCS versions 1,2, and 3 are supported. This includes most commercial Flow Cytometry systems from Becton Dickenson, Coulter Electronics, Ortho, Cytomation and others. The Open Command can read SAS data sets directly. SAS data sets are converted to JMP tables complete with column notes, table info, and conversion of SAS date/time values to JMP date/time values. You can now drag text files onto the JMP application and JMP automatically attempts to Import them as text. There is a new preference for gray backgrounds in analysis windows. Analysis Windows When you COMMAND-click on a selected point in a plot, the point is deselected. p-values show as <.0001 when they underflow the format. Lines in graphs are precisely clipped, so that they will not extend outside the frame when transferred to draw programs. More graphs support the log axis scale, including logistic and nonlinear. Data Tables - The Tables menu The data table window shows the number of selected rows and columns at the bottom left corner of the table. You can store data in short-integer format. This is a specialty feature that is used only when you need to conserve on storage space. To use this feature, you must first turn on the Enable short numeric choices option in preferences. Then you can select short integer formats in the Column Info dialog. If you use these storage formats, then JMP will only be able to store values as integers in the range corresponding to the length of the field. For example, the range for length 1 is 126 to 127. The Open dialog has a Show All Files check box that enables you to read JMP files that dont have the correct signature. This is needed when you need to open a data table from the Windows version of JMP. For the Macintosh version of JMP, the Import command can read a variety of file formats, including EXCEL, using XTND text translators. A translator to read EXCEL files is included with JMP. You can exclude columns from dialogs by excluding rows in the corresponding attributes table. The Summary Platform (Group/Summary command) has median, range and % total summary statistics. The dialog for Concatenate has a check box that lets you save the formulae and have them automatically evaluated in the output table. Note that the formulae are saved using the "first-seen" rule. The default setting of the check box is no evaluate.. The Join command is accessible by AppleScript (Macintosh only). Design of Experiments There is a Cancel Button in DOE Factor names and limits dialog. If a model is generated by an experiment dialog, the Fit Model dialog automatically shows the screening personality. D-Optimal Designs report has been reorganized so that it is more obvious that the efficiencies were from the Best Design, not the current one. Calculator When you COMMAND-click (ALT-click) on a nominal or ordinal variable in the calculator variable list, a popup menu appears listing all the variables values. The calculator has the new Random number functions: exponential, triangular(midpoint), Cauchy, gamma(alpha), poisson(lambda), binomial(n, p), geometric(p), negative binomial(n, p), and random number seed = conditional assignment. There are Probabilitiy Functions to compute values for the Normal Density, and to compute quantiles probabilities, and densities values for the standard Beta and Gamma distributions. When differentiating the calculator automatically creates temporary variables as needed to make differentiation process more efficient. Analyze Menu - Distribution of Y: Continuous Variables A popup command More Moments extends the Moments table by adding Sum, Variance, Skewness, Kurtosis, and CV. CV is coefficient of variation, which is the standard deviation divided by mean and multiplied by 100. A command Stem and Leaf in the popup menu at the top of the report does the stem-and-leaf plot invented by Tukey. The plot actively responds to clicking and brushing. The Normal Quantile plot has the Lillifors confidence bounds, reference lines, and a probability scale at the top. The Test mean=value command has a z-test feature, and signed-rank check option. The Capability Analysis allows you to enter your value for sigma, the process standard deviation. The outlier box plot has the means diamond display. Previously, there was no mean shown in the default Distribution graphs. The outlier box plot jitters points so that when they have the same value, you can see more than one point. Analyze Menu - Distribution of Y: Nominal or Ordinal Variables A Test Probabilities command tests that the probabilities are as specified. A popup menu in the Frequencies table lets you select the statistical information you want to include or exclude in the frequency table, including the new column for the standard error of the probability estimate. Fit Y by X - general If Fit Y by X cannot proceed due to a bad variable (such as all missing response values in a logistic regression), it tells you which variable was bad. Fit Y by X - Continuous by Continuous The Fitting menu is reorganized to include new fitting options: The Fit Transformed option fits a transformation on X or Y. Fit Line, Fit Polynomial, and Fit Transformed have a Plot Residual option. The equation of fit is shown for Fit line, Fit Polynomial, and Fit Transformed. Fit Nonpar Density has a feature to do mesh (surface) plots of the density. There is also an option to turn off the 5% contours, leaving only the 10% lines. The default kernel width has changed to .5 * stdDev * n**(1/6), which is half the old default. With a better default kernel, the slider controls to change the kernel width are now only revealed by checking the Kernel Control option in the menu. Fit Y by X - Continuous by Nominal A Normal Quantile plot shows both the differences in the means (vertical position) and the variances (slopes). Means diamonds have overlap marks drawn at sqrt(2)/2 out into the confidence diamond. When groups have equal sample sizes, these marks show if two means are significantly different by whether they overlap. The t-test report has more detail. The Jitter option adds random horizontal jitter so that you can see points that overlay on the same Y value. The horizontal adjustment of points varies from .375 to .625 with a 4*(Uniform-.5)**5 distribution. A Matching command in the Analysis popup menu lets you specify a matching variable to perform a matching model analysis. Optional lines connect the matching points. Fit Y by X - Nominal by Nominal A Cochran-Mante-Haenszel Test is offered. It prompts you to enter the name of a classification variable. Then it tests for association of the Y and X variables after adjusting for the classification variable. If there are more than 2 categories in the X or Y variable, then the tests will be different, depending on whether you look for scored levels, or keep the variable in categories. Analyze Menu - Fit Model Dialog A Close button has been added to close the dialog window. Analyze Menu - Fit Model : Standard Least Squares The Parameter Estimates table has a popup menu giving option to include confidence limits, standardized regression coefficients, and variance inflation factors for the parameter estimates. The Whole Model report has the option to request a Press statistic, which computes the sum of squares residual where the residual for each row is computed after dropping that row from the computations. Analyze Menu - Fit Model: Effect Screening Coding The effect screening reports have major revisions to evaluate the effects in a model in a way invariant to scaling. This report is important in testing effects as if they had been coded to be orthogonal. In addition when effects are not orthogonal, this option gives you a way to make them orthogonal so that they can be treated as a population of independent and identically distributed effects for normal quantiles plotting. The Transformed Parameter Estimates report appears open initially. The Transformed Parameter Estimates report shows both the original and transformed estimates. The report column Orthog Coded contains the orthogonalized estimates. If the design was orthogonal and balanced, then these estimates will be identical to the original estimates. If they are not, then each effects contribution is measured after it is made orthogonal to the effects before it. The is the column that is labeled Normalized Estimates Orthog t on the Y axis of the Normal Plot. If the model is saturated then pseudo-standard error is used to scale the values instead of the root mean square error. The report column Orthog t-Test contains the estimates divided by their standard error. These are equivalent to Type-1 Sequential tests. The significance level (p-value) is shown beside it. The old report column Normalized Estimate is now titled Orthog t-Test. The report column Prob>|t| is the significance level or p-value associated with the Orthog t-test values. The Scaled Estimate report column has been removed. The Pareto plot is now done with respect to the normalized estimates, which are more appropriate, especially when quadratic and crossed uncoded terms are present. An additional warning line appears for non-orthogonal designs stating Each Orthog Estimate is conditioned on the effects before it. Screening Prediction Profiler In the Prediction Profiler, new dialogs appear when you OPTION-click the graph to allow you to change the current value, the scale, and the number of grid points for continuous factors. Being able to change the number of grid points makes much more accurate desirability settings possible. In the Prediction Profiler, the popup menu has four items: 1) Confidence Intervals, which are initially on 2) Desirability Functions 3) Most Desirable in Grid, which searches all possible levels and sets the factors at the most desirable settings 4) Reset Grid, which brings up a dialog for each continuous factor. In the Save $ menu, Grid of Predicted saves predictions for all combinations in the grid, including the desirability. If you want to use the Prediction Profiler, but want different models for different responses, you can do a separate fit with these responses, saving the predicted value, and using the predicted value rather than the original response in the profiler. The Interaction Plots are interactive with respect to the current values of the profiler. A new Contour profiler offers interactive contour plots of the response surface that are use-ful for multiple-response optimization. Optionally, the contour profiler can show mesh plots. Screening Contour Profiler The check-mark popup menu on the screening platform has the Contour Profiler option that brings up an interactive contour profiling facility. This is useful for optimizing response surfaces graphically. Analyze Menu - Fit Nonlinear The nonlinear platform has two new save menu commands: Save Pred Confid and Save Indiv Confid for saving confidence intervals using the asymptotic linear approximation. Analyze Menu - Correlation of Ys There is an additional principal component command to use covariance matrices, or uncentered and unscaled. Analyze Menu - Cluster The rows in the data table and in the dendrogram are colored and marked. The color advances in the dendrogram tree for as long as the colors agree. The color/marker option is now a state that forces immediate recoloring as the diamond control is moved. Also, you can use the brush tool on the dendrogram. Graph Menu - Control Charts Control Charts can do variability charts with variance components. The Exclude row state is used when a control chart is first invoked. It is used in estimating sigma if it is not specified. Thus, all statistics which are based on the sigma (like the control limits) are affected. Subgroup statistics used in plotting the charts are not affected; that is, points based on Excluded rows will be displayed. Tools Menu The Annotate tool has enhanced features. After you create a test box for annotation (a yellow-stickie-type note), you can hold down the OPTION key (Mac) or ALT key (Windows) and click on the sticky to bring up a menu that lets you: change the color of the note change the type of border around the note specify no border This enables you to create notes that act as editable fields that can be used to change the axis labels of plots and charts.Annotate Tool Popup Menu  EMBED Word.Picture.6 Windows Menu If there is more than one data window open, the Close Data Windows command under the Window menu displays a dialog that gives you the option of saving changes to all modified tables (without prompt), discarding changes to all opened tables (without prompt), or taking a prompt for each modified table.  EMBED Word.Picture.6  Calculator - Random Number Functions  EMBED Word.Picture.6 The calculator has the following new random number functions: exponential, triangular(midpoint), Cauchy, gamma(alpha), poisson(lambda), binomial(n,p), geometric(p), negative binomial(n,p), and random number seed = conditional assignmentThe random number functions in JMP appear in formulas preceded by a ? to indicate randomness. Each time you click Evaluate these functions produce a new set of random numbers. The following is a description of all the random number functions: Uniform ?uniform generates random numbers uniformly between 0 and 1. This means that any number between 0 and 1 is as likely to be generated as any other. The result is an approximately even distribution. You can shift the distribution and change its range with constants. For example, 5 + ?uniform 20 generates uniform random numbers between 5 and 25. Normal ?normal generates random numbers that approximate a normal(0, 1) distribution. The normal distribution is bell shaped and symmetrical. You can modify the Normal function with constants to specify a normal distribution that has a different mean and standard deviation. For example, 5+?normal2 generates a normal distribution with a mean of 5 and standard deviation of 2. Exponential ?exponential generates a single parameter exponential distribution for the parameter lambda=1. You can modify the Exponential function to use a different lambda. For example, 0.1?exponential-0.1 generates an exponential distribution for lambda=0.1. The exponential distribution is often used to model simple failure time data, where lambda is the failure rate. Cauchy ?Cauchy generates a Cauchy distribution with location parameter 0 and scale parameter 1. The Cauchy distribution is bell shaped and symmetric but has heavier tails than the normal distribution. A Cauchy variate with location parameter alpha and scale parameter beta can be generated with the formula alpha+beta?cauchy. Gamma ?gamma(alpha  EMBED Word.Picture.6  ) gives a gamma distribution for the parameter, alpha, you enter as the function argument. The gamma distribution describes the time until the kth occurrence of an event. The gamma distribution can also have a scale parameter, beta. A gamma variate with shape parameter alpha and scale beta can be generated with the formula beta ?gamma(alpha). If 2alpha is an integer, a chi-square variate with 2alpha degrees of freedom is generated with the formula 2?gamma(alpha). Triangular ?triangular(mid  EMBED Word.Picture.6  ) generates a triangular distribution of numbers between 0 and 1, with the midpoint you enter as the function argument. You can add a constant to the function to shift the distribution, and multiply to change its span. Shuffle ?shuffle selects a row number at random from the current data table. Each row number is selected only once. When Shuffle is used as a subscript, it returns a value selected at random from the column that serves as its argument. Each value from the original column is assigned only once as Shuffles result. Poisson ?poisson(lambda  EMBED Word.Picture.6 ) generates a Poisson variate based on the value of the parameter, lambda, you enter as the function argument. Lambda is often a rate of events occurring per unit time or unit of area. Lambda is both the mean and the variance of the Poisson distribution. Binomial ?binomial(  EMBED Word.Picture.6  , probability  EMBED Word.Picture.6 ) returns random numbers from a binomial distribution with parameters you enter as function arguments. The first argument is n, the number of trials in a binomial experiment. The second argument is p, the probability that the event of interest occurs. When n is 1, the binomial function generates a distribution of Bernoulli trials. For example, n=1 and p=.5, gives the distribution of tossing a fair coin. The mean of the binomial distribution is np, and variance is np(1p). Geometric ?geometric(probability  EMBED Word.Picture.6  ) returns random numbers from the geometric distribution with the parameter you enter as the function argument. The parameter, p, is the probability that a specific event occurs at any one trial. The number of trials until a specific event occurs for the first time is described by the geometric distribution. The mean of the geometric distribution is 1/p, and the variance is (1p)/p2. Negative Binomial ?negBinomial(  EMBED Word.Picture.6  , probability  EMBED Word.Picture.6  ) generates a negative binomial distribution for the parameters you enter as function arguments. The first parameter is the number of successes of interest (r) and the second argument is the probability of success (p). The random variable of interest is the number of failures that precede the rth success. In contrast to the binomial variate where the number of trials is fixed and the number of successes is variable, the negative binomial variate is for a fixed number of successes and a random number of trials. The mean of the negative binomial distribution is (r(1p))/p and the variance is (r(1p))/p2. Random Number Seed This function lets you start a random number sequence with a seed you specify. To use the Random Number Seed function, assign it a value using the Assignment function found in the Conditions functions, and use the random number function you want as the results clause of the Assignment function. The example shown here uses the number 1234567 as the seed to generate a sequence of uniform random numbers. EMBED Word.Picture.6   EMBED Word.Picture.6  Calculator - Probability Functions  EMBED Word.Picture.6 Probability Functions now include the density function for the Normal distribution . Also, there are functions that calculate densities, quantiles and probabilities for the gamma and beta distributions: EMBED Word.Picture.6  The Normal Distribution function accepts a quantile argument from the range of values for the standard normal distribution with mean 0 and standard deviation 1 . It returns the probability that an observation from the standard normal distribution is less than or equal to the specified quantile. For example, the expression normDist(1.96) returns .975, the probability that an observation from the standard normal distribution is less than or equal to the 1.96th quantile. The Normal Distribution function is the inverse of the Normal Quantile function.  EMBED Word.Picture.6  The Normal Density function accepts a quantile argument from the range of values for the standard normal distribution. It returns the value of the standard normal probability density function (pdf) for the argument. For example, you can create a column of quantile values (X) and use normDensity(X) to generate density values. Then use GraphOverlay to plot the normal density by X, as shown to the right. EMBED Word.Picture.6  EMBED Word.Picture.6  The Normal Quantile (probit) function accepts a probability argument p, and returns the pth quantile from the standard normal distribution. For example, the expression normQuant(0.975) returns the 97.5% quantile form the standard normal distribution, which evaluates as 1.96. The Normal Quantile function is the inverse of the Normal Distribution function.  EMBED Word.Picture.6  The gamma distribution has two parameters, a>0 and b>0. The standard gamma distribution has b=1. The Gamma Distribution function is based on the standard gamma function, and accepts an argument with a quantile value and, optionally, a value for the shape parameter a, which as a default value of 1. It returns the probability that an observation from a standard gamma distribution is less than or equal to the specified X. gammaDist is the inverse of gammaQuant.  EMBED Word.Picture.6  The Gamma Density function (gamma pdf) accepts an argument whose value is a quantile. Optionally, you can specify a value for the shape parameter a, which has the default value 1. The figure to the right shows the shape of gamma probability density functions for various values of a. The standard gamma density function is strictly decreasing when a 1.  EMBED Word.Picture.6 When a > 1 the density function begins at zero when X is 0, increases to a maximum and then decreases.  EMBED Word.Picture.6  The Gamma Quantile function accepts a probability argument p, and returns the pth quantile from the standard gamma distribution with the shape parameter you specify. gammaQuant is the inverse of gammaDist.  EMBED Word.Picture.6  The beta Distribution has a positive density only for an X interval of finite length, unlike normal and gamma which have positive density over an infinite interval. The beta distribution has two parameters, a>0 and b>0, and constants A X B that define the interval for which the distribution has values. The Beta Distribution function accepts the response variable argument, whose range defines A and B. You can specify values for the shape and scale parameters, or use the default values of 1. The standard beta distribution occurs when A = 0 and B = 1.  EMBED Word.Picture.6  The standard beta distribution is sometimes useful for modeling the probablistic behaviour random variables such as proportions constrained to fall in the interval 0, 1. Examples of densities for several combinations of a and b are shown in the figure to the right. EMBED Word.Picture.6  EMBED Word.Picture.6  The Beta Quantile function accepts a probability argument p, and arguments or its shape and scale parameters. It returns the pth quantile from the standard beta distribution. betaQuant is the inverse of betaDist.  EMBED Word.Picture.6  Analyze Menu - Distribution of Y: Continuous Variables More Moments The More Moments popup command (Figure 1) in the Moments report adds, Sum, Variance, Skewness, Kurtosis, and CV (coefficient of variation, which is the standard deviation divided by mean and multiplied by 100) to the report. Figure 1 Moments Table with Optional Additional Moments  EMBED Word.Picture.6  Stem and Leaf Plot The Stem and Leaf popup command at the top of the histogram (Figure 2) does the stem and leaf plot invented by Tukey. The stem and leaf plot is a variation on the histogram. It was developed in the days when computers were rare. Each line of the plot has a Stem value that is the leading digits of a range of column values. The Leaf values are made from the next-in-line digits of the values. To reconstruct the data values from the plot, join the stem and leaf and use the scale factor. For example, on the first line of the stem-and-leaf plot for Weight shown to the right in Figure 2, you can read the value 172 (17 from the leaf and 2 from the stem). On the last line of the plot you can read 64 and 67 as values. The plot actively responds to clicking and brushing. When you click on leaf numbers they show in bold, and the corresponding rows in the data table and histogram bars are highlighted. Figure 2 Stem and Leaf Plot  EMBED Word.Picture.6  Test Mean=value The Test Mean=value popup menu command (Figure 3) has a dialog. that allows you to do a z-test if you specify the true standard deviation of the response, and to suppress or invoke the Wilcoxon signed-rank test Figure 3 Dialog for Test Mean=value option  EMBED Word.Picture.6  t-test and no Signed-rank  EMBED Word.Picture.6  z test using specified sigma, sign rank checked  EMBED Word.Picture.6 Outlier Box Plot Figure 4 shows two new options for the Outlier Box Plot: The Outlier Box Plot shows the means diamond. Previously, there was no mean shown in the default Distribution graphs. The Outlier Box Plot jitters points so that when they have the same value, you can see more than one point. Capability AnalysisFigure 4 Outlier Box Plot Options  EMBED Word.Picture.6 The Set Spec Limits popup menu option allows you to enter your own value for sigma, the process standard deviation. Normal Quantile Plot The Normal Quantile plot now shows the line of the mean and standard deviation and also the Lillifors bounds for normality. A probability scale appears at the top. Figures 5-8 explain normal quantile plots for four simulated distributions. Figure 5 Uniform Distribution EMBED Word.Picture.6 In the middle, the uniform distribution is steeper (less dense) than the normal. In the tails, the uniform is flatter (more dense) than the normal. In fact the tails are truncated at the end of the range, where the normal tails extend (infinitely).Figure 6 Normal Distribution EMBED Word.Picture.6 The normal distribution has a normal quantile plot that tends to follow a straight line. The points at the end have the highest variance and are most likely not to fall near the line. This is reflected by the flair in the confidence limits near the ends.Figure 7 Exponential Distribution EMBED Word.Picture.6 The exponential distribution is skewed, that is, one-sided. The top tail runs steeply past the normal line; it is spread out more than the normal. The bottom tail is shallow, and much denser than the normal. Figure 8 Double Exponential Distribution EMBED Word.Picture.6 The middle portion of the double exponential is denser (more shallow) than the normal. In the tails, it spreads out more (is steeper) than the normal.   EMBED Word.Picture.6  Analyze Menu - Distribution of Y: Nominal/Ordinal Test Probabilities for Nominal Distributions The popup command at the top of a histogram for nominal and ordinal variables has the Test Probabilities command. This command displays the dialogs shown in Figure 9 for the user to enter hypothesized probabilities. When you click Done, the Likelihood Ratio and Pearson chi-square tests are calculated for those probabilities. The entries are scaled so that the probabilities sum to one. Thus the easiest way to test that all the probabilities are equal is to enter a 1 in each field. If you want to test a subset of the probabilities, then you leave blank the levels that are not involved, and JMP substitutes estimated probabilities. Figure 9 Test Probabilities Dialog  EMBED Word.Picture.6  Popup options for Frequency Table The Frequencies table popup menu (Figure 10) lets you select which columns to display in the table: Count Probability, StdErr Prob, and Cum Probability. The new item is the standard error of the probability, computed as sqrt(p*(1 p)/n). The default table doesnt show the standard error. Figure 10 Frequencies Table  EMBED Word.Picture.6   EMBED Word.Picture.6  Analyze Menu - Fit Y by X: Continuous by Continuous Nonpar Density Nonpar Density has a new feature to do mesh (surface) plots of the density as shown in Figure 11. Also, There is also an option to turn off the 5% contours, leaving only the 10% lines. The default kernel width has changed to .5 * stdDev * n**(1/6), which is half the old default. With a better default kernel, the slider controls to change the kernel width are now only revealed by checking Kernel Control option in the menu. Figure 11 Mesh Plot with Nonparametric Density  EMBED Word.Picture.6  Fit Transformed The Fitting menu now has the command Fit Transformed. This displays a dialog with choices for both the Y and X transformation. Transformations include: log, square root, square, reciprocal, and exponential. The fitted line is plotted on the original scale as shown in Figure 12. The regression report is shown with the transformed variables, but an extra report shows measures of fit transformed in the original Y scale if there is a Y transformation. Figure 12 Fitting Transformed Variables  EMBED Word.Picture.6  EMBED Word.Picture.6  EMBED Word.Picture.6  EMBED Word.Picture.6  Other Changes to the Fit Y by X Platform for Continuous Y and X A Plot Residual command is now offered for Fit line, Fit Polynomial, and Fit Transformed. The equation of fit is shown for Fit line, Fit Polynomial, and Fit Transformed (see the analysis report in Figure 12).  EMBED Word.Picture.6  Analyze Menu - Fit Y by X: Continuous by Nominal Normal Quantile Plot The normal quantile plot shows difference in means by the vertical separation of lines, and differences in variance by the differences in slope. For example, in the Taguchi.jmp data, with HTime = 1 the variance of Shrink is smaller than with HTime = 1. With Gate, the distributions seem different, leading to different box plots, but the overall variance is about the same across the Gate groups. Figure 13 Normal Quantile and Box Plot Options for Oneway Analysis of Variance  EMBED Word.Picture.6  The Matching Variables Command The Matching Variable command in the Analysis popup menu addresses the case when the data in a grouped analysis come from matched (paired) data. For example, observations in different groups may come from the same subject. When this happens, the statistics from the grouped F and t tests are not right. A special case of this leads to the paired t-test. However, the paired t-tests in JMP apply only when the data is organized with the pairs in different variables, not in different observations. The Matching Variable command does two things: It fits an additive model that includes both the grouping variable (the X variable in the Fit Y by X analysis), and the matching variable you choose. It uses an iterative proportional fitting algorithm to do this. This algorithm makes a crucial difference if there are hundreds of subjects, because the equivalent linear model would be very slow and require huge memory resources. It draws lines between the points that match across the groups. If there are multiple observations with the same match id, lines are drawn from the mean of the group of observations. A Display popup menu item called Matching Lines can toggle these lines on and off, leaving just the analysis table. For example, consider the DOGS data table after stacking the LogHist0, LogHist1, LogHist3 and LogHist5 columns into the column called time, as shown in the data table in Figure 14. To see the results, chose Fit Y by X with Y as Y and time as X. Then select the Matching Analysis command from the Analysis popup menu beneath the plot and specify id as the matching variable. The match lines connect each subjects responses over time. Figure 14 Matching Variables Example  EMBED Word.Picture.6  EMBED Word.Picture.6 The analysis in Figure 15 shows the time and id effects with F tests. These are equivalent to the tests you get using the Fit Model platform, however you would have to run two models, one with the interaction term and one without. If there are only two levels, then the F test is equivalent to the paired t-test. Figure 15 Matching Variables Analysis Report  EMBED Word.Picture.6  Another use for the feature is to do parallel coordinate plots when the variables are scaled similarly. Means Diamonds Have 95% Overlap marks Overlap marks drawn in the means diamonds are  EMBED Word.Picture.6  CI above and below the group mean. For groups with equal sample sizes these marks show if the two group means are significantly different at the 95% confidence level. Figure 16 shows a Means Diamond when all Display popup menu options are in effect. Figure 16 Information Given by Means Diamonds  EMBED Word.Picture.6  Other Changes to the Fit Y by X for Continuous Y and X The t-Test Report has upper and lower 95% confidence limits.  EMBED Word.Picture.6  Analyze Menu - Fit Y by X: Nominal/X by Nominal Y Cochran-Mantel-Haenszel Test The contingency analysis platform has the Cochran-Mantel-Haenszel test for testing if there is a relationship between 2 categorical variables after blocking across a third classification. The Cochran-Mantel-Haenszel is a command on the popup menu at the top of the mosaic chart (Figure 17). When you select the command a dialog lets you select a grouping variable. The example in Figure 17 uses the HOTDOGS (hotdogs.jmp) data table from the sample data. The contingency table analysis for Taste by Type shows a marginally significant chi-square probability of about .07. Figure 17 Cochran-Mantel-Haenszel Command  EMBED Word.Picture.6  Figure 17 continued next page Figure 17 (continued) Cochran-Mantel-Haenszel Command  EMBED Word.Picture.6  However, if you stratify on fat to protein ratio (values 1 to 4), Taste by Type show no relationship at all (Figure 18). The adjusted correlation between them is .03, and the chi-square probability associated with the with the general association of categories is .28. Figure 18 Cochran-Mantel-Haenszel Tests Report  EMBED Word.Picture.6   EMBED Word.Picture.6  Analyze Menu - Fit Model Standard Least Squares Parameter Estimates Details The popup menu in the parameter estimates report (Figure 19) lets you include or suppress the standard error, t-ratio, significance p-value, 95% confidence limits, the standardized beta, and the variance inflation factor. Figure 19 Options to Show or Suppress Columns in the Parameter Estimates Table  EMBED Word.Picture.6   EMBED Word.Picture.6  Screening: Prediction Profile There are several new ways to control the Prediction Profiler shown in Figure 20: OPTION (or ALT)-click in the graph or horizontal axis area to get a dialog that lets you set the current factor value, the number of grid points, and the scale interval for that factor. OPTION (or ALT)-click in the left axis to change the Y response scale. The Confidence Intervals popup menu in the Prediction Profile lets you show or not show confidence intervals. The Most Desirable in Grid popup menu command in the Prediction Profile lets you search the grid for the most desirable factor setting. The dollar ($) popup menu lets you save a grid of predicted values in a data table. Figure 20 Options to Control the Prediction Profiler  EMBED Word.Picture.6  Screening: Effect Screening Option for Coded Estimates The effect screening reports (Figure 21) have major revisions to evaluate the effects in a model in a way invariant to scaling. The revisions are important in testing effects as if they had been coded to be orthogonal. In addition, when effects are not orthogonal, this option gives you a way to make them orthogonal so that they can be treated as a population of independent and identically distributed effects for normal quantiles plotting. The Transformed Parameter Estimates report appears open initially and shows both the original and transformed estimates. The report column Orthog Coded contains the orthogonalized estimates. If the design was orthogonal and balanced, then these estimates will be identical to the original estimates. If they are not, then each effects contribution is measured after it is made orthogonal to the effects before it. The old report column called Normalized Estimate is now titled Orthog t-Test. The column titled Prob>|t| is the significance level or p-value associated with the values in the Orthog t-Test column. The report column Orthog t-Test contains the estimates divided by their standard error. These are equivalent to Type-1 Sequential tests. The significance level (p-value) is shown beside it. This is the column that is plotted on the Normal Plot Y axis, labeled Normalized Estimates (Orthog t). The Scaled Estimate report column has been removed. The Pareto plot is now done with respect to the normalized estimates, which are more appropriate, especially when quadratic and crossed uncoded terms are present. An additional warning line appears for non-orthogonal designs stating Each Orthog Estimate is conditioned on the effects before it. Figure 21 Comparison of Old and New Effect Screening Reports old report  EMBED Word.Picture.6 new report  EMBED Word.Picture.6 old Pareto Plot  EMBED Word.Picture.6  New Pareto Plot  EMBED Word.Picture.6  Screening: Contour Profiler The check-mark popup menu on the screening platform has the Contour Profiler option that brings up an interactive contour profiling facility. This is useful for optimizing response surfaces graphically. Figure 22 shows an annotated example of the Contour Profiler for the TIRETREAD sample data. To see this example, run a response surface model with ABRASION, MODULUS, ELONG, and HARDNESS as response variables, and SILICA, SILANE, and SULFUR as factors. The following features are shown in Figure 22: There are slider controls and edit fields for both the X and Y variables. The Current X values generate the Current Y values. The Current X location is shown by the crosshair lines on the graph. The current Y values are shown by the small red diamonds in the slider control. Figure 22 The Contour Profiler  EMBED Word.Picture.6  The other lines on the graph are the contours for the responses, as set by the Y slider controls, or entering values in the Contour column. There is a separately colored contour for each response (4 in this example for the response variables). You can enter Low and High limits to the responses, which results in a shaded region. If you click and drag from the side zones of the Y sliders, this will set the limits, or you can enter in the Lo Limit or Up Limit columns. If you have more than 2 factors, then you use the check box on the right to switch the graph to other factors. OPTION (ALT) click on the slider control to change the scale of the slider (and the plot too if its an active X variable). For each contour, there is a dotted line in the direction of higher response values, so that you get a sense of direction. The Update Mode determines whether dragging the sliders produces a continuous update of the graph, or if it waits for the mouse to be released to update. On fast machines, Continuous mode is the best choice, since it calculates fast enough. Optionally, you can display Mesh Plots as in Figure 23 with the Contour Profiler. The mesh plots are displayed by default. The Surface Plot option in the popup menu in the Contour Profiler hides or displays the mesh plots.Figure 23 Optional Mesh Plots With the Contour Profiler EMBED Word.Picture.6  EMBED Word.Picture.6  Analyze Menu - Fit Nonlinear Two new save menu commands: Save Pred Confid and Save Indiv Confid let you save confidence intervals using the asymptotic linear approximation.  EMBED Word.Picture.6  Analyze Menu - Correlation of Ys There is an additional principal component command to use the covariance matrices or uncentered and unscaled in the computations instead of the correlation matrix.  EMBED Word.Picture.6  Analyze Menu - Cluster The rows in the data table and in the dendrogram are colored and marked. The color advances in the dendrogram tree for as long as the colors agree. The Color/Marker option is now a state that forces immediate recoloring as the diamond control is moved. Also, you can use the brush tool on the dendrogram.  EMBED Word.Picture.6  Graph Menu - Control Charts Variability Analysis (Gage R&R Charts) Variability (Gage R&R) analysis is a new feature in the Control Charts platform. In a Gage R&R analysis, a number of supposedly identical parts are taken from a production line. Each one is measured by a number of operators a number of times using different measuring instruments. You want to know the magnitudes of the variation due to operators, parts, and instruments. In the same way that a Shewhart control chart can identify processes which are going out of control over time, a gage R&R chart can help identify operators, instruments, or part sources that are systematically different in mean or variance. Gage refers to gages (gauges) or instruments and R&R refers to repeatability and reproducibility. The gages or instruments that take measurements in a manufacturing process are subject to variation. Too much variation in the measurement system can mask variation in the process (part-to-part variation). One type of measurement variation is caused by conditions inherent in gages. This variation, know as repeatabilty, results when one person measures the same characteristic or part several times with the same gage. Repeatability is the measurement-to-measurement variation. Another type of measurement variation, known as reproducibility, occurs when different individuals (operators) using the same gage take measurements on the same part. Reproducibility is the operator-to-operator variability. The Gage R&R analysis is a way to compare the part-to-part variation with operator-to-operator and measurement-to-measurement variation. The analysis uses a standard variance-components model. In JMP, a variability analysis can have 2 or 3 factors, and there is a choice of crossed or nested model. The result is a variability (multi-var) chart, an Analysis of Variance, Variance Components report, and computation of a discrimination ratio. For 2 factor crossed or nested designs JMP also displays a Gage R&R table. Note: This analysis requires balanced data. You can see a variability analysis for the abrasion example by using the Date and Shift variables in the abrasn2.jmp table to represent the roles of operator and part in the variability analysis. To do the analysis, choose Control Charts from the Graph menu. When the control chart dialog appears, first click the Variability button. This displays the variability analysis dialog shown in Figure 24. Complete the dialog as shown and click Chart to see a Gage R&R report for a 2-way crossed design (Figure 25). Figure 24 Control Chart and Variability Analysis Dialogs  EMBED Word.Picture.6  Gage R&R Charts The Variability Chart in Figure 25 shows the average abrasion measurement with max and min bars for each shift on each day. The check-mark popup menu for the Variability Chart gives options to connect the nested group means, and to show the overall group means with a dotted line. In this example you might be concerned that difference in average measurements between shifts A and B on 4/30/95 would make it impossible to detect existing variation in the manufacturing process. Figure 25 Variability Chart and Gage R&R Analysis  EMBED Word.Picture.6  Gage R&R Variability Report A Gage R&R report is also given. Computations in the Gage R&R table use variance estimates, s2, given in the Variance Component Estimates table and the Sigma Multiple (u) and Tolerance you specify (see Figure 24). The Measurement Unit Analysis in the table (Figure 26) lists these quantities: Repeatability (EV) - equipment variation, computed as u  EMBED Word.Picture.6  Reproducibility (AV) - operator to operator variation, computed as u  EMBED Word.Picture.6  interaction (IV) - interaction of operator and part computed as u  EMBED Word.Picture.6  Gage R&R (RR) - measure of repeatability and reproducibility, computed  EMBED Word.Picture.6  Part variation (PV) computed as u  EMBED Word.Picture.6  Total Variation (TV) - computed as  EMBED Word.Picture.6  ) In the Gage R&R report, % Tolerance lists the Measurement Unit Analysis for EV, AV, IV, RR, and PV divided by the tolerance you specify on the variablility chart dialog. The Analysis of Variance performs a set of F tests appropriate for a variance components model. The denominator term is constructed to have the correct expectation to test each effect. Low p-values indicate the variance is positive. The Discriminant Ratio characterizes the relative usefulness of a given measurement for a specific product. It compares the total variance of the measurement, M, with the variance of the measurement error, E. The Discriminant Ratio, D, is computed D = sqrt((2*M/E) 1) A rule of thumb is that when the Discriminant Ratio is less than 4 the measurement cannot detect product variation, so it would be best to work on improving the measurement process. A Discrimination Ratio greater than 4 adequately detects unacceptable product variation. Figure 26 Gage R&R Report  EMBED Word.Picture.6 Guidelines (Barrentine, 1991) for acceptable % RR are: < 10% excellent 11% - 20% adequate 21% - 30% marginally acceptable > 30% unacceptable. EMBED Word.Picture.6  EMBED Word.Picture.6  EMBED Word.Picture.6  Barrentine (1991), Concepts for R&R Studies, Milwaukee, WI: ASQC Quality Press.  EMBED Word.Picture.6  Index A  INDEX Analyze menu Cluster 11, 33 Correlation of Ys 11, 33 Distribution of Y 8, 17-21 Fit Model 9-11, 28-32 Fit Nonlinear 11, 32 Fit Y by X 8-9, 22-28 Annotate Tool (frame and background options) 12 beta distribution functions (calculator) 16 B  INDEX binomial random number function (calculator) 14 C  INDEX calculator efficiency 7 probability functions 7, 15 random number functions 7, 12-14 capability analysis (Distribution of Y command) 8 Cauchy random number function (calculator) 13 Close Data Windows command (Windows menu) 12 Cluster command (Analyze menu) 11, 33 Cochran-Mantel-Haenzel test (Fit Y by X command) 9, 27-28 coding convention for screening model 9-10, 31-32 coefficient of variation 8, 17 confidence limits for parameter estimates 9 contour profiler (screening model) 11, 31-32 Control Charts (Graph menu) 11, 33-36 Correlation of Ys command (Analyze menu) 11, 33 covariance matrix for principal components analysis 11 customer support 4 CV (coefficient of variation) 8, 17 D  INDEX Data table selected rows and columns 6 short numeric preference option 6 desirability functions (screening model) 10, 29 discrete probability tests (Distribution of Y) 8, 21 discriminant ratio (variability analysis) 36 Distribution of Y command 17-21 additional moments 8, 17 capability analysis 8 jittered points 8 Lillifors confidence bounds on normal quantile plot 8, 19-20 outlier box plot with jittered points 19 stem-and-leaf plot 8, 18 test discrete probabilities 8, 21 z-test 8, 18 DOE improvements 7 E  INDEX Excel files (Import command) 7 exponential random number function (calculator) 13 F  INDEX FACS file (Import command) 6 File menu Import command for Excel files 7 Import command for FACS file 6 Open command for SAS data sets 6 Fit Model command (Analyze menu) 9-11, 28-32 fit model dialog 28 parameter estimates table 28 screening model 9-11, 29-32 standard least squares 9, 28 studentized beta 28 variance inflation factor 28 Fit Nonlinear command (Analyze menu) 11, 32 Fit Y by X command 8-9, 22-28 Cochran-Mantel-Haenzel test 9, 27-28 equation of fit added to report 8, 23 jitter display options 9 matching analysis command 9, 24-26 mesh plot for nonparametric density 8, 22 normal quantile plot shows slopes 9, 24 overlap marks on means diamonds 9, 26 plot residuals for regression 8, 23 regression on transformed numeric variables 8, 22 Frequency table (Distribution of Y) 8, 21 G  INDEX gage R&R analysis (see variablility charts) gamma distribution functions (calculator) 16 gamma random number function (calculator) 13 geometric random number function (calculator) 14 Graph Menu (Control Charts variability analysis) 11, 33-36 gray background (preference option) 6 Group/Summary command statistical options 7 H  INDEX Hardware and Software Requirements 2 I  INDEX Import command Excel file (Mac only) 7 FACS file 6 installing JMP 4 J  INDEX jitter option Distribution of Y 8 Fit Y by X command 9 K  INDEX kurtosis (in moments table) 8, 17 L  INDEX Lillifors confidence bounds 8, 19-20 M  INDEX matching model analysis (Fit Y by X command) 9, 24-26 means diamonds (overlap marks) 9, 26 median (Group/Summary platform) 7 memory recommended 2-3 mesh plot Fit Y by X nonparametric density 8, 22 screening model 32 mesh plots (Fit Model platform screening model) 11 N  INDEX negative binomial random number function (calculator) 14 normal distribution functions (calculator) 15 normal quantile plot (Fit Y by X command) 9, 24 normal random number function (calculator) 13 normalized estimates in screening model 10, 30 O  INDEX Open command for SAS data sets 6 optimizing response surfaces (screening model) 11, 29-32 orthogonal t-test in screening model 10, 30 orthogonalized estimates in screening model 10, 30 P  INDEX parameter estimates table (standard least squares) 28 Pareto plots in screening model 10, 30 percent total (Group/Summary platform) 7 plot residual option (Fit Y by X regression) 8, 23 Poisson random number function (calculator) 13 prediction profiler (screening model) 10, 29 Preferences option gray background 6 short numeric choice enabled 6 press statistic 9 principal components analysis (Correlation of Ys command) 11 probability functions (calculator) 7, 15 R  INDEX random number functions (calculator) 7, 12-14 range (Group/Summary platform) 7 regression equation 8, 23 S  INDEX SAS data sets, use Open command 6 save commands for nonlinear fit 11, 32 screening model (Fit Model platform) coding conventions 9-10, 29-30 contour profiler 11, 31-32 desirability functions 10, 29 mesh plots 11, 32 normalized parameter estimates 10, 30 optimizing response surfaces 11, 29-32 orthogonal t-test 10, 30 orthogonalized parameter estimates 10, 30 Pareto plots 10, 30 prediction profiler 10, 29 saving grid of predicted values 11 short numerics (new preference option) 6 shuffle random number function (calculator) 13 skewness (in moments table) 8, 17 standard least squares (Fit Model platform) confidence limits for parameter estimates 9 parameter estimates table 9, 28 press statistic 9 standardized regression coefficients 9, 28 variance inflation factor 9 standardized regression coefficients 9, 28 stem-and-leaf plot 8, 18 Summary Platform statistical options 7 System Hardware and Software Requirements 2 T  INDEX technical support 4 Tools menu (Annotate Tool) 12 training and education 4 transformed numeric variables for regression 8, 22 triangular random number function (calculator) 13 V  INDEX variability analysis (Control Charts) 33-36 analysis of variance 36 control chart dialog 34 discriminant ratio 36 Gage R&R chart and table 35 variance components estimates 36 variability charts (Control charts command) 11 variance inflation factor 9, 28 W  INDEX Windows menu (Close Data Windows command) 12 World Wide Web for information and technical support 5 Z  INDEX z-test (Distribution of Y) 8, 18  PRINT \p PAGE " "  Whats New in JMP Version 3.2 Whats New in JMP Version 3.2  PRINT \p PAGE "% Copyright (c) 1991 by SAS Institute Inc. Cary, NC 27513 % Author: Walter E. Martin % Date: 07JAN91 % Description: Writes alignment marks and thumb tabs (on % odd pages). Change the thumb attributes % and margins for the desired effect. % gsave /tempdict 30 dict def /inch {72 mul} def /pagewidth 8.5 inch def /pageheight 11 inch def tempdict begin % the following parameters define the thumb attributes % /thumbnumber 1 def % relative thumb number /thumbtext ( ) def % text to be placed on thumb /thumbs 6 def % # of thumbs down the page /fontsize 9 def % size of font in points /Helvetica-Bold findfont % change the font if necessary fontsize scalefont % change the point size as required setfont % set the above as the current font /thumbgap 0.0 inch def % gap between thumbs /thumbwidth 0.5 inch def % width of the thumb from % the edge of the page % margin sizes /topmargin 1.0 inch def % distance from the top margin % to the edge of the page, /bottommargin 0.5 inch def % bottom of the page, /insidemargin 1.0 inch def % inside of the page, and /outsidemargin 1.0 inch def % outside of the page. /gutter 0.0 inch def % binding are if needed % added to the inside margin % draw the alignment marks at each corner of the page % stack contains: bottom right y % bottom right x % top left y % top left x /pagemarks { /bry exch def /brx exch def /tly exch def /tlx exch def tlx tly 3 add moveto % top left mark 0 8 rlineto tlx 3 sub tly moveto -8 0 rlineto tlx bry 3 sub moveto % bottom left mark 0 -8 rlineto tlx 3 sub bry moveto -8 0 rlineto brx tly 3 add moveto % top right mark 0 8 rlineto brx 3 add tly moveto 8 0 rlineto brx bry 3 sub moveto % bottom right mark 0 -8 rlineto brx 3 add bry moveto 8 0 rlineto } def % evenpage and odd page calculate the position of the % page alignment marks for mirror image pages. They % use Microsoft Word's wp$MARGIN variables to determine % text placement relative to the page margins. /evenpage { wp$left outsidemargin sub % top left x pageheight wp$top sub topmargin add % top left y pagewidth wp$right sub % insidemargin gutter add add % bottom right x wp$bottom bottommargin sub % bottom right y pagemarks } def /oddpage { wp$left insidemargin gutter add sub % top left x pageheight wp$top sub topmargin add % top left y pagewidth wp$right sub % outsidemargin add % bottom right x wp$bottom bottommargin sub % bottom right y pagemarks } def /thumb { /texttop pageheight wp$top sub def /textlen texttop wp$bottom sub def /pageedge pagewidth wp$right sub outsidemargin add def /thumblen textlen thumbs div def thumbnumber 1 sub thumbs mod % relative thumb pos. thumblen mul thumbgap 2 div add % top of thumb texttop exch sub % relative to texttop /thumbtop exch def % save as thumbtop newpath pageedge thumbwidth sub thumbtop moveto thumbwidth 3 add 0 rlineto 0 thumblen thumbgap sub neg rlineto thumbwidth 3 add neg 0 rlineto pageedge thumbwidth sub thumbtop lineto closepath 0 setgray fill gsave pageedge thumbwidth sub thumbtop moveto -90 rotate thumbtext stringwidth pop % get length of string thumblen thumbgap sub % and the length of the thumb exch sub 2 div % then determine X to center text Root Entry@ĮxؼWordDocumentŨObjectPool$tؼtؼSummaryInformation(E; chpljs32W odXLetter oEPRIV''''....Times New Roman Symbol &ArialMGaramond Narrow Times 1Courier"HelveticaNew YorkMLucida Casual MChicago#A(6l&8l&$q5j5$+7#Corporate Microcomputing Department5ܥhc e{Ũ  DŽŽŽŽŽ$Ž2222222!֤P&Pv6X722222N22  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~Lp2x FMicrosoft Word Document MSWordDocWord.Document.69q    DocumentSummaryInformation8_938266804k Ftؼftؼ_938266805g Fftؼftؼ_938266807c Fftؼftؼ  !#&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWY\]^_`abcdefghijklmnopqrstuvwxyz{|}~՜.+,0HPlt|  SAS Institute Inc.jj Oh+'0   H T `lt|_c22. Normal.doto$Corporate Microcomputing Department2Microsoft Word for Windows 95@G@\!ؼ@0hؼ5$qm  m!l#  l#  l# # # #_938266808 _ Fftؼ tؼ_938266809[ F tؼ tؼ_938266810 X F tؼ tؼ_938266812U F tؼ tؼ_938266813R F tؼ tؼ_938266814O F tؼ tؼ_938266815 L F tؼH#tؼ_938266817I FH#tؼH#tؼ_938266818F FH#tؼH#tؼ_938266819C FH#tؼH#tؼ_938266820@ FH#tؼH#tؼ_938266821= FH#tؼH#tؼ_938266823 : FH#tؼH#tؼ_9382668247 FH#tؼ%tؼ_9382668254 F%tؼ%tؼ_9382668261 F%tؼ%tؼ_938266828. F%tؼ%tؼ_938266829+ F%tؼ%tؼ_938266830( F%tؼ(tؼ_938266831% F(tؼ(tؼ_938266832 " F(tؼ(tؼ_938266834 F(tؼ(tؼ_938266835 F(tؼ(tؼ_938266836 F(tؼM*tؼ_938266837" FM*tؼM*tؼ_938266839 FM*tؼM*tؼ_938266840!# FM*tؼM*tؼ_938266841  FM*tؼM*tؼ_9382668424  FM*tؼ,tؼ_938266843 F,tؼ,tؼ_938266845%' F,tؼ,tؼ_938266846 F,tؼ,tؼ_938266848&* F,tؼ,tؼ_938266849 F//tؼ//tؼ_938266850)+ F//tؼ//tؼ_938266852 F//tؼ//tؼ_938266853(0 F//tؼ//tؼ_938266854 F//tؼy1tؼ_938266855-/ Fy1tؼy1tؼ_938266857 Fy1tؼy1tؼ_938266858.2 Fy1tؼy1tؼ_938266859 F3tؼ3tؼ_93826686013 F3tؼ3tؼ_938266862 F3tؼ3tؼ_938266863,D F3tؼ[6tؼ_938266864 F[6tؼ[6tؼ_93826686657 F[6tؼ[6tؼ_938266867 F[6tؼp8tؼ_9382668686: Fp8tؼp8tؼ_938266870 Fp8tؼp8tؼ_9382668719; Fp8tؼp8tؼ_938266872 Fp8tؼp;tؼ_9382668738@ Fp;tؼp;tؼ_938266875 Fp;tؼp;tؼ_938266876=? Fp;tؼ``=tؼ_938266877 F``=tؼ``=tؼ_938266879>B F``=tؼ``=tؼ_938266880 F``=tؼ``=tؼ_938266881AC F``=tؼ`?tؼ_938266882 F`?tؼ`?tؼ_938266883<L F`?tؼ`?tؼ_938266885 F`?tؼ`?tؼ_938266886EG F`BBtؼ`BBtؼ_938266887 F`BBtؼ`BBtؼ_938266889FJ F`BBtؼ`BBtؼ_938266890t F`BBtؼPDtؼ(_938266891IK FPDtؼPDtؼ_938266892 FPDtؼPDtؼ_938266894 HT FPFtؼPFtؼ _938266895'()*+,-./0 FPFtؼPFtؼ>?@_938266896GHIJKLMNOPMO FPFtؼPFtؼ^_`_938266898ghijklmnop FPFtؼPnItؼ~_938266899mmaryInformationNR FPnItؼPnItؼ_938266900 FPnItؼ@Ktؼ_938266901QS F@Ktؼ@Ktؼ_938266903} F@Ktؼ@Ktؼ_938266904 PXz F@Ktؼ@Ktؼ_938266905w F@)Ntؼ@)Ntؼ_938266906UWt F@)Ntؼ@)Ntؼ_938266908q F@)Ntؼ@)Ntؼ_938266909VZn F0sPtؼ0sPtؼ_938266910k F0sPtؼ0sPtؼ_938266911Y\h F0sPtؼ0sPtؼ_938266913e F0Rtؼ0Rtؼ_938266914[]b F0Rtؼ0Rtؼ_938266915_ F0Rtؼ0RtؼPIC66815  FLPICT6817 ^` F k  %  x!!!!","X"k"""""""B#V#u#&&&&&&**'*(*)*****).?.@.00111220212ܻܙuD7vKI U]cIUcU]c]uD7KIuD7vKI uD\IuD7KIuD7vKII uDIU uD|IuD7KIuD7vKI;12223266-64677L9S9;;=!=/=== >>?+?,?y??j@n@o@@@@@@@uAvAB%B-B0BBB;CC?CLCMCbCcCeCfCCCCCRDSDDDEE&E2EEEEFFFG GGGZG[GGGUVWX}~ef}~߀ŁӁ^rÂĂłƂ˂ق$zvz]cU]cuD7KIuD7vKI uDAIV]cuD7KIuD7vKI uDAIuD7KIuD7vKI uD,IuD7KIuD7vKI uDI uD)IuD7KIuD7vKII.$]^uvwxyz{ۄ܄>N݅CDKLS\tu}!"ȇɇۇǧǣuD7KIuD7vKI uDZI]c]]cuD7KIuD7vKI uDIV]cuD7KIuD7vKI uDxYIUuD7KIuD7vKII uDpBI6_`gh BCUVʊԊՊ֊ȋɋ);"#:;<=ƾ窢璊uD7KIuD7vKI uDtIuD7KIuD7vKI uDsI]VuD7KIuD7vKI uDpIuD7KIuD7vKII uD[I]c]c uDZI6= !HI`abcdef}~1:MWXYpqrsȏ؏ӐԐ@H򮦷򌄕uD7KIuD7vKI uDIUU]cVcuD7KIuD7vKI uDIV]cuD7KIuD7vKI uDȤIU]cuD7KIuD7vKII uDtuI]c4KP)/FN”ÔؔԕՕ #$%&XYpqrs×ܗݗuD7vKI uDVIuD7KIuD7vKI uD8MIuD7KIuD7vKI uDFIuD7KIuD7vKI uDIU]cuD7KIuD7vKII uD(IU]cV/ '2s{ƚǚޚߚ 5678ԝ՝࿷য়wouD7KIuD7vKI uDIuD7KIuD7vKI uDxIuD7KIuD7vKI uDIuD7KIuD7vKI uDpIuD7KIuD7vKII uDhIUU]c uDVIuD7KI+!"6:̠ՠ QZơȡQZno¢ڢۢCLuᴬ᜔cuD7KIuD7vKI uDJIuD7KIuD7vKI uD<3IV]cc]cuD7KIuD7vKII uDpIUU]cVcUVc uDI9Ȥɤ) Ħ456789PQRSxuD7KIuD7vKI uDIuD7KIuD7vKI uDoIuD7KIuD7vKI uDIuD7KIuD7vKI uDIcU]c]cuD7KIuD7vKII uDKIU/ç˧̧ͧڧۧ,-2BGHKL_hlmƩѩթ\gĪŪƪǪ ܬBIaoίӯ 仳uD7KIuD7vKI uD8I]cUVcuD7KIuD7vKII uDI]cUc]UcU]cF :>^gʰܰ fo±ñձޱ-.EFGH !"̴մ)Nǿ诧藏uD7KIuD7vKI uDIuD7KIuD7vKI uDLIuD7KIuD7vKI uD\IuD7KIuD7vKII uDhIU]U]cU]c6NOfghi./dmɷӷܷ6;?C˸̸͸θϸ'(?@ABPYuD7KIuD7vKI uDpIuD7KIuD7vKI uD?I]ccUVcUVcuD7KIuD7vKI uD?IuD7KIuD7vKII uD5I57@34KLMNOPghijkѼڼ߼Zp4>jkuD7vKI uDIU]ccuD7KIuD7vKI uDIuD7KIuD7vKI uDIUuD7KIuD7vKI uDIuD7KIuD7vKII uD`|I2ۿ<P^k$s}*+,-=>?VWXYDNƾ൭जcuD 7KIuD 7vKIuD 7KIuD 7vKIuD 7KIuD 7vKIuD7KIuD7vKII uDIU]cU uDIuD7KI<!'-3dmCK !(/81234567NOPQŽ뭥uD7KIuD7vKI uDXIuD7KIuD7vKI uDIU]cuD 7KIuD 7vKII uDI]]cUcAEF]^_`6;ELMS]b"#XadoﶮuD7vKI uDIU]c]cUVuD7KIuD7vKI uD(IuD7KIuD7vKI uD8IuD7KIuD7vKII uDHI]U]c9uv#NPuvwx89:;ĺĺĺuD7]cKIuD7]cvKIuD7]cKIuD7]cvKI ]cIuD]cI]c]U]U]ceuD7KIuD7vKII uD(IU uDIuD7KI4;RSTUbd !"9:;<WXYbmʐԄԙU]cuD7KIuD7vKIU]uD7]cKIuD7]cvKI]c]uD7KIuD7vKII uDIU]cuD]cIuD7]cKIuD7]cvKI ]cI3 $%&'YZ abyz{|yynyuDU]cI uDIuD#7KIuD#7vKI uD IuD"7KIuD"7vKI uD IuD!7KIuD!7vKI uDIuD7KIuD7vKI uDIc U]uD7KIuD7vKII uD`I+z{}~ \]_`hi#$&'/0RSUV^_defhiqrDEGHPQIuDU]cI uDI]cdfgop !#$,-+,./78!"$%'(8-12457TVsvwz{uuDuDUVUV uDVI uDIuDU]cIG-./ OX>p (   ) U _ k  ) Q   A   %         f                        ( & 0!(/' [r7d(NCs7k=]y )I                                             -In!00?`iql x!!!!"X"""B#                     '       '      ,(,,(,#!!P#$B#u###5%K%V&o&&&&&&_'t')5)*+*=*L** +++---).@../00011K1f111  '   '        '  '    '  '  '         '     !h#!(,'11242a2 33'67|8809A999I:::!;==>?h?d@@@@uABBBCSDDEEFGZGGHdeWXtӐғĔԔԕ   $$$        '           '  r  $''L'$'#'' Pt"P'"''" 'Xtu7ܗmsƚ    $'$$$$'  2  2   '    <  <   P*8p$' Pt$@( !ll !ll'@ P$!"9:ҝӝ"O̠ /QnܢȤ <  <   < <   < <   '      '       '     !''H$'L#$P*8p ť8TUVkΨ\Ȫجfıű ' '           ' '    '     '     X X P( ''!$#!! P p P p$$''ű-IٳNjk/L ϸ'CPҺ3OkݼT    ~  ' '   g ' '    '     6          '   t     '(#'#!'$'(T޾4jxEs.ZxDo      '          77 j             !'( P $$'P$!!!56Ro=Ea} ;N0bd=@@            '            '    V  '    !l'('$!!'## P@ "V? (_p   ?  t      @.T T T T T T   " . P ' P  P((($!' !l !l!l`a} #SP~ =\ CVz}5 \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ L-#.*5b*Cer9Xy$Am<d Dq \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ -- 0\_#&RU&0WjAo1dh+Z\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ --DG5To/C^'Ss7cf #Y\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ --Yq +.e&'r-/5_%(`S4\ \ \ \ \ \ \ \ \ \ \ \     98-,4vC{:;r'7GW_ 'O]t( ^ -^      L         X    C O V X ` c      K   =[-&6o9}34UVxyz{  \ 98GK0@0NormalP U]a c2`2 Heading 1p  ]a c0,@, Heading 2xc$,@, Heading 3 U],@, Heading 4 (U]c$@$ Heading 5]c"A@"Default Paragraph Font(O(b,book,indentedU(@(TOC 4   (V,@,TOC 3 UVc(@(TOC 2( U2@2TOC 1 P"@( U] @Index 3* @*Index 2 LU]c* @*Index 1 LU]c(@q Line Number" @"Footer !*@*Header V]c4@4 Footnote Textp]c )@ Page NumberUVc4o4 heading 1,h1p  ]a c0*O* heading 2,h2 (Uc$$O$ heading 3,h3U$O$ heading 4,h4(U&O& heading 5,h5 ]c(O(bl,blist!(4O"4lh,Helv list,listHelv item"]c(O2( h2line,hline#LOf,figure$OaRgb%]c6Ob6lhb,HelvB list,TimesB list&(U]cOrl,list'PO!lg(U]cOQgb14)c O d*UDODbl8,bl-8after,bist 8 after+ hX c "O"e,(O(il,index letter-(U*O*refs,references .(Oline/( (O( list2 item 0] cOlist21P0O"0 endnote text2U]cto2t WP DefaultsP3Jp@ P !$`'0*-/2p5@8;=@ ]a c$OqB$ l8,8after4c (OR(l15U]c>Ob>lii,list item italic6( UV]c0Or0lb,timesB list 7(U]c.O. even header 8V]c O odd header9Vc.O. i2,Indented2 :]c*O* +bl,+blist ;]c0O0bindent<LP]c*O*-b,-book =P]c.Oq.+l,+list>]c$O$l2 ?U]c2O2lc,Chi list,Chi @] c$O$-l AU]c4O1"4Contents HeaderB  Uc2O22 blist2,b2C0P]c.O. include,inc D]c2O2+b,+bookE]c(Ob(Header1 FhU]a{{5                                  ! " # $ % & ' ( ) * + , -"./01 N qx#-*/I7?7F[NV]di=su{=|sԚlOl.[R;dea f{ :m   L sy`_a -!J"?#8$%&)',()<*+,-V.P/f0111223349\EiJ12J]jQsx$= N; IB#1VX136NPE]_w:RT!9; $& ay{z} \_h#&/RU^dhqDGPcfo #,+.7{ :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::\!0028 sxrz=T####''((***)+-+0+;+<+>+..|33Y4b4w4444444465:566E9I9==|?? @@!@(@@@@@BBfBnB CCVC[CCCCCDDDDBEIE;F@FI IIIJJVKZK$L*LILOLMMMMOO"S+SSS TT=UCUUUVVVWAWGWWWXXXX Z%Zp]s]]]____&_+_,_2___` `L`V````aKaUayccddgg)g0gCgFgiglgllllmmKmQmmmmmn,nnnNoUooooorrrruuuuww8y;yayhyyy"}+}}}~~~~PXǀπ$>GaiȄۄ%(OWUXYaʇӇ 1:MUɌΌ rwɑӑݑ’̒u|œ̓=DÔʔ t|ޕBJSXU\ EJ ܟԠڠȥ٥s~ƦnvJO/Fv $9>[r qvdm"0^dKQ9? 28":Dw9C/7xUZ4>PV5A -CK/;_i;H 1?/5|e#Corporate Microcomputing Department>\\JMPNT\source\Install\Jmp3\WinOS\MASTERS\JMP\OTHER\JMP322.DOC@chpljs32Ne06:winspoolchpljs32chpljs32W odXLetter0EPRIV''''thumbwidth 2 div % move from the edge of the page fontsize 3 div sub % half of the thumbwidth and half rmoveto % the text point size, centered 1 setgray thumbtext show grestore "  Whats New in JMP Version 3.2 Whats New in JMP Version 3.2 Whats New in JMP Version 3.2 Whats New in JMP Version 3.2 "8/= @ @ "8/= @ @ 8/= @ @ (h8/= @ @ t ,Uk^     & ( R T ] ^ i j ~    ' ( O P | ~ YZpq56bc$'JcI uDI]cVU ]ceU]cHYJM?Bpr46gj:<Z\ux &(EHjm ijk uD|IceuD7KIuD7vKII uD|I uDIVcQObjInfo8 F PIC66819  F"LPICT6820 ac FObjInfo1 FPIC66828  FXLPICT6829 df F% ObjInfo0 F$PIC66831  FLPICT6832 gi F[ ObjInfo4 FZPIC66835  FLPICT6836 jl Fp ObjInfo7 FPIC66839  FLPICT6840 mo FObjInfo1 FPIC66848  FLPICT6849 pr FvObjInfo0 FPIC66852  FLPICT6858 su FObjInfo9 FPIC66860  FLPICT6862 vx FvObjInfo  PIC$%&'()*+,-./0 56789 L@PICTDEFGHIJKLMNOP y{UVWXYz`ObjInfofghijklmnopuvwxyPIC66823  FLPICT6824 |~ FvObjInfo5 FPIC66826  FLPICT6842  F_BObjInfo3 FPIC66845  FLPICT6846  FjObjInfo3 FPIC66854  F&LPICT6855  F"ObjInfo7 F!      !"#$%&'()*+,-./0123456789:<=>?@ABCDFZGHIJKLMNOPQRSTUVWXY[\]^`abcdefghklmnopqrstuvwxyz{|}~PIC66868  F-LPICT6870  F)ObjInfo1 F(PIC66872  F4LPICT6873  F0ObjInfo5 F/PIC66876  F;LPICT6877  F7ObjInfo9 F6PIC66880  F>LPICT6881  F|(ObjInfo2 F=PIC66883  FALPICT6885  FObjInfo6 F@PIC66887  FDLPICT6909  FOObjInfo0 FCPIC66911  FGLPICT6913  F[8ObjInfo4 FFPIC66915  F\LPICT6815  FJdObjInfo7 FIPIC66828  FnLPICT6829  F_ObjInfo0 F^PIC66831  FqLPICT6832  Fmb%ObjInfo4 FpPIC66835  FLPICT6836  FtObjInfo7 FsPIC66839  FLPICT6840  FV ObjInfo1 FPIC LPICT ObjInfoPIC LPICTntry " ObjInfoentPICtPool PLPICTryInformation "~ ObjInfo4  F! PIC66895'()*+,-./0  FSL@PICT6896GHIJKLMNOP  F0`ObjInfo8ghijklmnop FRPIC66904  FZLPICT6905  FVObjInfo6 FUPIC66908  FLPICT6858  F] ObjInfo9 F\PIC66860  FLPICT6862  FObjInfo  PIC$%&'()*+,-./0 56789L@PICTDEFGHIJKLMNOP UVWXYF%`ObjInfofghijklmnopuvwxyPIC6842  FLPICTfo3  FObjInfo5 FPIC6846  FLPICTfo3  FBObjInfo4 FPIC6855  FLPICTfo7  F"ObjInfo8 FPIC6870  FLPICTfo1  FObjInfo2 FPICnfo9  FLPICT6880  F0ObjInfo1 FPICnfo2  FLPICT6883  FObjInfo5 FPICnfo6  FLPICT6887  F*ObjInfo9 FPICnfo0  FLPICT6911  FObjInfo3 FPIC66828  FLPICT6829  F]ObjInfo0 FPIC66831  FLPICT6832  FObjInfo4 FPIC66835  FLPICT6836  F~ObjInfo7 FPIC66839  FLPICT6840  FObjInfo1 FPIC66883  FLPICT6885  FObjInfo6 FPIC66887  FLPICT6889  F,ObjInfo0t FPIC66891  FLPICT6892  F)ObjInfoPIC LPICTfo *ObjInfoPICnfo4   FL PICT6895'()*+,-./0  F-@ObjInfo6GHIJKLMNOP F`PICnfo8ghijklmnop  FLPICT6904  F6 ObjInfo5 FPICnfo6  FLPICT6908  F4 ObjInfo8 FPICnfo9  FLPICT6860  F|ObjInfo2 FPIC6842  FLPICTfo3  FJYObjInfo5 FPIC6846  FLPICTfo3  Fxr2ObjInfo4 FPIC6855  F!LPICTfo7  FvObjInfo8 F PIC6870  F(LPICTfo1    F$ObjInfo2 F#PIC6883  F.LPICTfo5   F+ObjInfo6 F*PIC6887  F1LPICTfo9  F4/ObjInfo0 F0PIC6911  F7LPICTfo3  F4ObjInfo8 F3PIC6829  F=LPICTfo0  F:ObjInfo1 F9PICnfo7  FKLPICT6839  F@ObjInfo0 F?PICnfo1  FNLPICT6858  F ObjInfo9 FMPIC66860  FTLPICT6862   FQObjInfo3 FPPIC66864  FZLPICT6866 !# FWObjInfo7 FVPIC66873  F`LPICT6875 $& F]ObjInfo6 F\PIC66877  FcLPICT6879 ') FObjInfo0 FbPIC66881  FiLPICT6882 *, FfObjInfo3 FePIC6885  FoLPICTfo6 -/ FlObjInfo7 FkPICnfo rLPICT 02ObjInfoqPICnfo yLPICTfo4  35 Fu ObjInfo5'()*+,-./0 Ft@PICnfo6GHIJKLMNOP  FL`PICTfo8ghijklmnop 68 F|ObjInfo4  F{PICnfo5  FLPICTfo6 9; F`ObjInfo8 FPIC6842  FLPICTfo3 <> F`ObjInfo5 FPIC6846  FLPICTfo3 ?A F`ObjInfo4 FPIC6855  FLPICTfo7 BD F`ObjInfo8 FPIC6870  FLPICTfo1 EG F`ObjInfo2 FPICfo9  FLPICTfo0 HJ F`ObjInfo1 FPICfo3  FLPICTfo8 KM F`ObjInfo9 FPICfo0  FLPICTfo1 NP F`ObjInfo7 FPIC66839  FLPICT6840 QS F0ObjInfo1 FPIC66848  FLPICT6849 TV FObjInfo0 FPIC66852  FLPICT6853 WY FTObjInfo4 FPIC66855  FLMETA6857 Z\ FHPICTfo7 ] FObjInfo9 FPICnfo0  FLMETAfo1 ^` FHPICT6858 a FObjInfo9 FPIC66860  FLMETA6862 bd FHPICT6873 e FObjInfo5 FPICnfo6  FLMETA6877 fh FHPICT6879 i FObjInfo0 FPIC66881  F LMETA6882 jl FHPICTfo3 m FObjInfo5 FPICTfo6 FObjInfo7 FLo2x@ @ddPro ddPro0 ,  Helvetica .+Discrimination Ratio@ @0"} ddPro( &1 (!Source ddPro( #2* Date ddPro( />* Shift ddPro( 8&(!; Discrim Ratio ddPro( #I2+ 1.1979003 ddPro( /I>* 1.0434612@ @0@"9zLh TB [r r[ddPro ddPro0 ,  Helvetica .+Variance Component Estimates ddPro( %A (" Component ddPro( $1* Date ddPro( 0=* Shift ddPro( <I7* Date*Shift ddPro( HU* Error ddPro( A%("D Var Comp Est ddPro( &N3+  1.300000 ddPro( 2N?* 0.012500 ddPro( >NK* 2.583333 ddPro( JLW(TO 11.916667 ddPro(n %(" % of Total ddPro([ $1+  8.22 ddPro([ 0=*  0.08 ddPro(] <I(F 16.34 ddPro(] HU*  75.36 ddPro(. %(" Cum. Total ddPro( $1+  1.30 ddPro( 0=*  1.31 ddPro( <I*  3.90 ddPro( HU(R 15.81 ddPro( %>("Cum. % ddPro( $1=+  8.2 ddPro( 0=<(:! 8.3 ddPro( <I<(F 24.6 ddPro( HU=(R 100.0 ddPro( @%p("CSqrt(VC) ddPro( $J1l+  1.140 ddPro( 0J=l*  0.112 ddPro( <KIm+  1.607 ddPro( HKUm*  3.452@ [r0[r"Rh@0#L2 `F XS SXddPro ddPro0 ,  Helvetica .+Analysis of Variance@ XS0"} ddPro( %, ("Source ddPro( $1* Date ddPro( 0=* Shift ddPro( <I7* Date*Shift ddPro( HU* Error ddPro( H%^("KDF ddPro( $H1^*  4 ddPro( 0H=^*  1 ddPro( <HI^*  4 ddPro( HFU^(RI 30 ddPro( |%("SS ddPro( $g1(.j 130.600 ddPro( 0i=+  22.500 ddPro( <iI*  89.000 ddPro( HgU(Rj 357.500 ddPro(c %(" Mean Square ddPro(B $1+!  32.650 ddPro(B 0=*  22.500 ddPro(B <I*  22.250 ddPro(B HU*  11.917 ddPro(  %("F Ratio ddPro( $1(. 1.46742 ddPro( 0=*  1.01124 ddPro( <I*  1.86713 ddPro( !%K("$Prob>F ddPro( $"1K+ 0.35962 ddPro( 0"=K* 0.37150 ddPro( <"IK* 0.14216@ XS0XS"FFL. |ND p< <ddPro ddPro0 I,  Helvetica .+Gage R&R@ <0O"D ddPro( & (#Measurement Unit Analysis ddPro( %2]* Repeatability (EV) ddPro( 1>e* Reproducibility (AV) ddPro( =J* Operator * Part Variation (IV) ddPro( IVW* Gage R&R (RR) ddPro( UbZ* Part Variation (PV) ddPro( an]* Total Variation (TV) ddPro( mzM* Sigma Multiple ddPro(} y4* Tolerance ddPro(T %2(/ 17.778071 ddPro(R 1>+ 5.871903 ddPro(R =J* 8.277467 ddPro(T IV(S 20.470845 ddPro(R Ub+ 0.575788 ddPro(T an(k 20.478941 ddPro(R mz+ 5.150000 ddPro(}R y* 50.000000 ddPro( &*(# % Tolerance ddPro( %3%+ 35.5561 ddPro( 1?%* 11.7438 ddPro( =K%* 16.5549 ddPro( IW%* 40.9417 ddPro( Uc&*  1.1516 ddPro(q <(?Gage R&R Report assumes that column 'Date' represents Operator, ddPro(e * #and column 'Shift' represents Part.@ <0:"L+  Y?> >ddPro ddPro( ,  Helvetica .+ RR ddPro$  ! @(2 ddPro( $8 +PV ddPro$ 4 > @(72 >"< "# ###############  >h      " ddPro8 *,Times ( +L4@v ddPro ddPro$ , Symbol .+s ddPro  ,Times @+PV ddPro  ( 2 " "################  \    L4@Z ZddPro ddPro$  ,Times .@+2 ddPro$ 3 >@)2 ddPro$ ,  Helvetica@( EV ddPro8 * @)+ ddPro$ $8 @) AV ddPro8 8H ++ ddPro$ CS @( FIV ddPro$ N Y@(Q2 Z"X@ "# ###############  Z\     @" ########Lf 4@v ddPro ddPro$  , Symbol .+s ddPro  ,Times @+IV ddPro  ( 2 " "################  \    L*4@     #$%'*+,.123589:<?BEHKLMNOPQRSTUVWXYZ[]`abcdefghijklmoruvwxyz{|}~z ddPro ddPro$ , Symbol .@+ s ddPro  ,Times @+AV ddPro  ( 2 " "################  \    L*4@v ddPro ddPro$ , Symbol .@+ s ddPro  ,Times +EV ddPro  ( 2 " "################  \    Lq4@LJ(* LjHF)((B'   'ddPro@ ' "4# ddPro0eBm 2,  Helvetica .+130@ ' "4#@"4# ddPro0Bm 2(135@ ' "4#@"4# ddPro0Bm ~2(140@ ' "l4#@"S4# ddPro0Bm M\2(X145@ ' ":4#@"!4# ddPro0+Bm +2('150 ddPro0Zm HW+3A ddPro0Zm r)*B ddPro0Zm )4A ddPro0Zm )*B ddPro0ZSm )4A ddPro0Z)m .=)*B ddPro0Zm bq)4A ddPro0Z˖m )*B ddPro0Zm )4A ddPro0Zmm )*B ddPro0Mm O ~(R4/29/95 ddPro0Mm  )^4/30/95 ddPro0MIm  8)a5/1/95 ddPro0Mm l )^5/2/95 ddPro0Mm  )^5/3/95 ddPro0:^m  S(Shift within Date ddPro0CTm i(Variability Chart@ ' 0v"d@0' # ddPro0  (Abrasion ' f f f     / @  @    @?`!`@0$"* UUUU MM  "fLd">"="f="i"h"fh"fwP""""("K"\"*"*2"""z"z<"%"$"p$"p3d"xX"W"RW"Rfd"z""R"RP"""f"fP"""\"\<@"N,#@"+#@" ,#@"yg-#@"*#@0!5#@"6\"6"?"H"Q"Z"c"l"u"~""#@"q\"q"q"q"q"q"q"q"q"q"q"q#@"\"""""" ")"2";"D"M#@"{N\"{N"{W"{`"{i"{r"{{"{"{"{"{"{#@"^""""""""""" #g   HH"9d=8=I=F=@=:=1=5=2=1===== ==  =  5= =G=C=<=2=(=%=%========!=%=*=-=3=X=== Z= ==  = !=!!= $!P=   =  f=G=5=5=-=-=*=*=%=!=!=!=!=!=!=!=!=!=!=!=!=0=b====  = == ==J=0=!=!=!=!=!=$=.=.=*=*== p== ===== = 8=>=:=4=$=$=$===,=*=2=2=`==0= j=  O= G=P= r= X=  = =R=,=,="="="="=(=A=j=\=Z=\===L= = }=  q=  v=    =    P= d=d=\=`=p=:=(="="="======!!!!!!n ^   #)!OQ3S| ~ +   $+++$.48D9J9\9z99 9 m 9%#   9- 9      9     9L,9   9 9IP)JPPTJ^ jw ~ |]DBBBHh 1)     rBBBB<?Es vs|}{ vkFBB^o{o{a99  99  M9 9#? 9 9   9 9 9V9T9H9;1$.$$8HSt  y _ ` f^ 04*'!($$*0DPju~`0004*'!###&,NVZ#    @        G QM)&&&&&&&&)/////C}   ! $  Ya~ g      O    i   !   ?    LR;;;IK      M ;///////+(**))))))!!!!!!!!!!!!!!$$$*9UMII        OOGJRe`!      !vk"       "   $$$m  m!l#  l#  l# # # #Ln2xm  m!l#  l#  l# # # #Ln2xm  m!l#  l#  l# # # #Ln2xm  m!l#  l#  l# # # #Ln2xL..WW(|NN NNddPro NN 1{HH 7{4DC><:<'27 !("(Ž(9>(8"DQ(>"$N'"1/ |1 , @@@M @L@?@@F@hC@`@E@@@@ @B @A@?@ @;?@`8@:@?5 @@ @8@6@?@@@H!@'?@@|_i@ ?@>@,@?L)@[ @Ȉ+ @& @< @  @( @$@' @? @'Ā @ϓĉ@ @;>D= @? CI"? O'ȟ: ]" :_"D '= $H; KDH?x$ 7H5 ȑ 2~ @ I"$D xDH2)'0 ?}'D? ΓH <<>D* D'' D"$" <'E?D) |"E$ D'DDHE<|" 'ğɈF D?E/$>u=$O" I>'Q"/; "DB :)D>N9O``;:> 3 0 3.?`p0?/>G E pGB qA8  AD  A@P>l qA@P  A@! AD qy8S/ ?n =A?1N'HH 7'N A9PcE$P  UEPIEPAEP AE$P  A9Þq |  @@ @  @ @@@@h@`@ @@@ @ @@@ @?@`@@? @@ @@@@@@@@@@@@@@@@@@?@@@9 @ @@ @ą @DȀ @/ @ @ @}_"B C? O?$\ DHy' 'H0 H :I  "" $  I?@ |"DD $DH;<" ^D OH <$H \D i=$" IOH"'"/ |H?"DB D`:IND O       ?` p?   p qA8 AD A@P qA@P  A@ AD qy8????+ ??1N{mHH 7{mN zd Be BU rU0 BM BM {D |  @@ @  @ @@@@h@`@ @@@ @ @@@ @?@`@@? @@ @ mp*S`݊d6])@}鵰?wll [ wtp ۝ c ?ͱҀ zmn` ә nv =) KX _󹝸ǀ Gkg A{ܰ @ۧ3 @? @ͶS @zm @;I @n9` @?9 @Z @ݜ98 @ @7tRh @۷9H @v9 @ֈ@n9@@;@h@oݘ<#/@pD#(@DR@ =罀 y F9 '؀ >b?` :!yDՀ G"'!D> B"" /B">| 4DDDӃ D_, _p '``        ?` p?   p qA8 AD A@P qA@P  A@ AD qy8????+ ??@ 0EH#@ C#@ G#L;<!"LA6l LA6       !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNPQRSTUVWXYZ\]^_`abcdefhijklnopqrstuvwxyz{|}~ ddPro qHHjdm sssllllllll ddPro$ We#,  Helvetica .+`X ddPro$ W*e9)Y ddPro$ WBeb)Term ddPro$ bBpg* SILICA ddPro$ mB{j* SILANE ddPro$ xBl* SULFUR ddPro$ We(`Current ddPro$! cq(l 1.3738843 ddPro$! n|* 58.467407 ddPro$ w+ 2.3 ddPro$ WeW(` Grid Density 1c!pP 9@ 8"`Q ddPro$ d!rF+ 20 x 20 `@pP ddPro$ WVe(`Y Update Mode  1cZp@ 8"` ddPro$ dZr+ Continuous `p@"#@ffffff"#@"#@fff"# ddPro$t *c(-Response ddPro$i *`* ABRASION ddPro$^ *]* MODULUS ddPro$S *Q* ELONG ddPro$H *b* HARDNESS ddPro$t (Contour ddPro$i + 147 ddPro$^ ( 1392 ddPro$S" ( 436.22642 ddPro$H + 66.59434 ddPro$t B(Current ddPro$i B( 158.54834 ddPro$^ B* 1483.9451 ddPro$S B* 355.08418 ddPro$H B* 71.828211 ddPro$t Gv(JLo Limit ddPro$i Rt+ ddPro$^ Rt* ddPro$S L|(O 350 ddPro$H O{+ 64 ddPro$t z(}Up Limit ddPro$h| + ddPro$]| * ddPro$S z(} 500 ddPro$H |+ 72@ 0C""eRX;zdHHX;z]a:"& ""& ""& ""& " ddPro( GVu (QContour Profiler@ 0EUx"Qb ddPro(/ :0( Slide these or edit ddPro(* these then the cross ddPro(* hairs move, and the ddPro(* current predicted ddPro(* values change. ddPro( V8(YSlide this off-axis ddPro(* factor and the whole ddPro(* surface changes.P "P########  PIPIJKLMNOP####"JP "k########  Pdkdefghijk####P "}########  Tv}vwxyz{|}####"8>P "Y########  DSXSTUVWX####"RP "}########  Hw}wxyz{|}####"=9 ddPro( 5G( These choose ddPro(* the axes of the ddPro(* contour plot. "5 (P "d*########  H]&c-],-^(-_&(`&'a'(+,b()c)+####oPHHoPc "(,("  "(,("  #%$%$%#  y' (* +* (y' P "########  @#### ddPro$2 9X +<66.33 ddPro$ "0E(+ SILANE ddPro$o 9X+f33.67  1V-o33 """"""#### """"   offffff "### ###  ,""o " ### ###  ,  o "### ###  $  o "### ###   o "### ###   o33 "### ###  ,o "### ###  ,o333 "####  $o33 "B####   oYYY "7### ###  47B7:=@Boffffff ",# ## ###  4,7,/257o "#### ## ###  ,","$),1"1 1 1YYY13311o333 "##### ## #   o "### ###  <o33 "X####   XYoYYY "M### ###  DMXMOQRTVXo "C### ## ###  4BMBDGJMffffff17BYYY1,7ffffff1",1"1 1 1YYY1133133311oYYY "b### ###  Dbmbdfgikmo "[### ## ###  4XbX[^`b1MX1BM17B1,7ffffff1",1"1 1 1YYY1133133311oYYY "x####  4x| x y z  {  |  o "m# ## ###  Dmx m oqrtvx 1bm ffffff1Xb YYY1MX ffffff1BM 17B YYY1,7 ffffff1", 1" 1  1  1 YYY1 1 331 3331 1 oYYY "### #  Lo "~ ### ## ###  4x x ~   ffffff1m xYYY1b mffffff1X b1M XYYY1B M17 B1, 7ffffff1" ,1 "1  1  1 YYY1 1 331 3331 1 oYYY "# ## ###  \""  "offffff "### ## ###   "1x"YYY1mx"ffffff1bm"YYY1Xb"1MX"1BM"17B"1,7"ffffff1","1""1 "1 "1"YYY1"331"1"3331"1"ffffff1"-1"-1x"-YYY1m"x-1b"m-1X"b-1M"X-1B"M-17"B-1,"7-ffffff1"",-1""-1 "-1" -1"-YYY1"-331"-1"-3331"-1"-off "_####  4Y_^_\^Z\YZY_o "_### ## #  <_jhjfhcfac_a_jo ###### ## #   juo "u####  4musuqsoqmomu1uo "u### ## #  <u~|~y|wyuwu1o ######## #  $o "####  $11o "### ###  4}}~11o #### ## #  4zz{|}11o #### ## #  ,xxyz11o "##### #   y11o "### ## #   x11o #### ## #  ,xxyz11o #### ###  <zz{|}~1o "####### #  $o "####  <~~1o "### ###  do "### #  \YYY1U_1U_1xU_1mUx_1bUm_1XUb_1MUX_1BUM_3317UB_1,U7_1"U,_1U"_1 U_1U _331U_3331U_1U_1U_1U_1U_YYY1_j1_j1x_j1m_xj1b_mj1X_bj1M_Xj1B_Mj17_Bj331,_7j1"_,j1_"j1 _j1_ j331_j3331_j1_j1_j1_j1_jYYY1ju1ju1xju1mjxu1bjmu1Xjbu1MjXu1BjMu17jBu1,j7u331"j,u1j"u1 ju1j u331ju3331juo "j### #### ###   juo "j### ## ###  $jujutujt1ju1juYYY1u1u1xu1mux1bum1Xub1MuX1BuM17uB1+t8o ",u### #### #   "u,o ""u##### #  4u"u||}}~~"uo "u##### #  , u| uzz{{|u|o " u##### #  ,u zuxxyyz uzo33 "u##### #  ,uxuvvwwxuxo333 "u### #   uvo "u### #   o "u##### #  $uwuwvwuvYYY111x1mx1bm1Xb1MX1BMo "C########  47C79;=Co "7##### #  4,7,/257o ",### #  ,#,#&),111x1mx1bm1Xbo "X##### ###  <MXMOQSUXo "M### #  <BMBDGIKM111x1mxo "m##########  <bmbceghmo "b### #  DXbXZ[]_`b11o "##########  ,xxyzo "x##### #  Lmxmnprsuvxo33 "m####   kmYYY1o "########  Lo "####  4{{}o "##### #  To33 "####   o "LU### ###  \BUL_BU_D^_E]^F\]G[\HZ[IYZJXYKWXLUW17UB_1,U7_1"U,_1U"_1 U_1U _1U_3331U_1U_1U_1U_1U_off #####   B_Cao "B_### #####  T7_Bj7_j;ij<hi=gh>fg?ef@deAcdB_c1,_7j1"_,j1_"j1 _j1_ j1_j3331_j1_j1_j1_j1_jo ":j####  ,7j:n7jn8mn9lm:jlo "7j### #####  <,j7u,ju3tu4st5qs6pq7jp1"j,u1j"u1 ju1j u1ju3331ju1ju1ju1ju1juo "2u####  D,u2|,u|-{|.z{/yz0xy1wx2uwo "-v### #####  ,"v-"v+,-v3331u"1 u1u 1u3331u1u1u1u1uo ")### #  L")"#$%&'()1"1 1 133311111o #### #####  T" !"1 1 133311111o "####  ,o "### #####  <  1 133311111o "### #  D   o " ### ## ###  $   133311111o " ### ###  L   133311111o  #####̌̌   """"$ "#o333 """""### #####  411111o "### ###  D11111o "### #####  41111o "### ###  <1111off #####   o333 "### ## ###  $111o "### ###  4111o #### ###  $111o #### ###    1 1 1 o #### ## #  <     1 1 1 o "####  4""!!"o "### ## ###  ,"""1"1"1"o "'####  <(-(-())**++,,-o ""### ## ###  D"-"-"##$$%%&&''-1"-1"-1"-1"-@̌̌ "#@"#@̌̌"#@"#@̌̌"#@"#@̌̌"#@"#@̌̌"#@"#@̌̌"#@"#@" #@̌̌" #@"#@̌̌"#@"#@"#@̌̌"#@"#@"#@̌̌""#@"#@̌̌"" #@""#@̌̌",#@"*#@̌̌",#@"'#@̌̌"/ #@".#@̌̌"7#@"4#@̌̌"7#@"2#@̌̌"9 #@"6#@̌̌"; #@"9#@̌̌"= #@";#@̌̌"?" #@"<'#@ffffff" #@" #@ffffff" #@"#@ffffff"#@"#@ffffff"#@"#@ffffff"#@"#@ffffff"#@"#@ffffff"#@"#@ffffff"#@"#@ffffff"#@"#@ffffff"#@"#@ffffff""#@"!#@ffffff""#@" #@")#@",#@"&#@ffffff"- #@",#@ffffff"2 #@"1#@ffffff"7 #@"4#@YYY"7#@"3#@ffffff"9 #@"6#@ffffff": #@"7#@ffffff":" #@"7'#@"#@"#@"#@"#@"#@"#@"#@"#@"#@"#@"#@"#@33"#@",#@".#@",#@"'#@""#@"#@"#@"#@" #@"#@"#@"#@"B #@"F#@"B#@"<#@"7#@"3#@"X #@"\#@"X#@33"R#@"M#@"L#@"m#@"o#@"m#@"g#@"a#@"#@"#@"}#@"x#@"u#@"#@"#@"#@"#@"#@"#@s3s3s3"U #@"yZ#@s3s3s3"xa #@"re#@s3s3s3"rj #@"it#@s3s3s3"lu #@"fz#@s3s3s3"f #@"^#@s3s3s3"b #@"[#@s3s3s3"\ #@"W#@s3s3s3"X #@"S#@s3s3s3"U #@"P#@s3s3s3"R #@"N#@s3s3s3"P #@"L#@s3s3s3"O #@"L#@s3s3s3"P #@"N#@s3s3s3"R #@"Q#@s3s3s3"W #@"W#@"T#@s3s3s3"b#@"b#@s3s3s3"b#@"\#@s3s3s3"m #@"r #@s3s3s3"m#@"j#@"#@s3s3s3"#@"#@s3s3s3"#@"}#@"t #@s3s3s3"#@"#@ #@ V,#@ T.#@!.#@!T#@!T# ddPro$Z Rv+0.3835 ddPro$ZO )\SILICA ddPro$Z  0)^2.0165@  3#@!3#@!#@!# 1_ ddPro(" \ (Slide these or edit these ddPro(* to change the ddPro(* corresponding contour line.ddPro 1l ddPro($ i(l Set these to ddPro(* shade the ddPro(* area outside ddPro(* the limits.ddProP  "s########  Pszswwzyzstxywxvwuvtu####"vP "########  H####"P "y########  Lu|uvwxyz{|####"w'P "########  L####"TP "C########YYY  PCJCDEFGHIJ####"F P ":)########  L:)A0:)*;*,<,/=/0>.0?,.@+,A)+####">+ ddPro( T9(^<Contour for Modulus ddPro(* Contour for Abrasion ddPro(* Contour for Hardness ddPro(* Contour for ElongP "########  P####" "]/ ddPro ddPro( ,  Helvetica .+Pareto Plot of Estimates@ 0"~ ddPro$  (% +Term ddPro$ & 3%* temp ddPro$ 1 >** gl ratio ddPro$ < I?* temp*temp ddPro$ G TD* gl ratio*temp ddPro$ R _I* gl ratio*gl ratio ddPro$ ] j* ht ddPro$ h u#* ht*ht ddPro$ s 1* ht*temp ddPro$y ~ 6* ht*gl ratio ddPro$ P(($SOrthog Estimate ddPro$ &s3+# -18.896428 ddPro$ 1s>* -17.161973 ddPro$ <tI+ 12.213839 ddPro$ GsT(Pv -11.360751 ddPro$ Rw_+ 9.502501 ddPro$ ]vj(fy -7.759403 ddPro$ hvu* -3.576335 ddPro$ sv* -2.711088 ddPro$y ~v* -2.581989ddPro  1' ddPro$- (($.2ddPro@  UUUU"&c #ddPro 1' ddPro$ ()*.4ddPro@  UUUU"&c #ddPro 1('0 ddPro$ %(3)+.6ddPro@  UUUU"&)c #ddPro 1R'Z ddPro$ O(])*.8ddPro@  UUUU"&Sc #333 UUUU1&1 8@"&.#@# #@##@##333 UUUU11< 8@"1*#@# #@##@##@"++ #333 UUUU1<G 8@"<#@# #@##@##@"6 #333 UUUU1GR 8@"G#@# #@##@##@"A! #333 UUUU1R] 8@"R#@# #@##@##@"L= #333 UUUU1]h 8@"]#@# #@##@##@"WT #333 UUUU1hs 8@"h#@# #@##@##@"bh #333 UUUU1s~ 8@"s#@# #@##@##@"mp #333 UUUU1~ 8@"~#@# #@##@##@"xw #@ ~#@#o#@!#@##@ #@!#@!#@!#@0#8 ddPro ddPro( ,  Helvetica .+Pareto Plot of Scaled Estimates@ 0" ddPro$  )& +Term ddPro$ ' 4&* temp ddPro$ 2 ?+* gl ratio ddPro$ = J@* temp*temp ddPro$ H UE* gl ratio*temp ddPro$ S `J* gl ratio*gl ratio ddPro$ ^ k* ht ddPro$ i v$* ht*ht ddPro$ t 2* ht*temp ddPro$x  7* ht*gl ratio ddPro$ Q)(%TScaled Estimate ddPro$ 't4+# -19.559661 ddPro$ 2t?* -17.764330 ddPro$ =uJ+ 13.060561 ddPro$ HtU(Qw -11.759495 ddPro$ Sx`+ 9.574875 ddPro$ ^wk(gz -8.031745 ddPro$ iwv* -3.722367 ddPro$ tw* -2.806243 ddPro$x w* -2.672612ddPro  1( ddPro$, )(%.2ddPro@  UUUU"'c # ddPro$ ) )*.4@  UUUU"'c # ddPro$ &)4)+.6@  UUUU"'*c # ddPro$ P)^)*.8@  UUUU"'Tc # UUUU1'2 8@"'.#@# #@##@## UUUU12= 8@"2*#@# #@##@##@",* # UUUU1=H 8@"=#@# #@##@##@"7 # UUUU1HS 8@"H#@# #@##@##@"B# # UUUU1S^ 8@"S#@# #@##@##@"M? # UUUU1^i 8@"^#@# #@##@##@"XU # UUUU1it 8@"i#@# #@##@##@"ci # UUUU1t 8@"t#@# #@##@##@"nq # UUUU1 8@"#@# #@##@##@"yx #@ #@#o#@!#@##@ #@#x#@!#@##@0#d5R R5 5R,  Helvetica .( Transformed Parameter Estimates0  ( Term* Intercept* temp* gl ratio* ht* temp*temp* gl ratio*temp* gl ratio*gl ratio* ht*temp* ht*gl ratio* ht*ht(UOriginal+ -4.33333(S -25.87500* -23.50000* -10.62500+ 25.29167(S -22.00000+ 18.54167+ -5.25000* -5.00000* -7.20833( Orthog Coded+ 15.20000( -18.89643* -17.16197+ -7.75940( 12.21384( -11.36075+ 9.50250( -2.71109* -2.58199* -3.57633( Orthog t-Test+#  1.8998( -2.3618*  -2.1450*  -0.9698+  1.5266( -1.4200+  1.1877(  -0.3389*  -0.3227*  -0.4470(!Prob>|t|+ 0.1159* 0.0646* 0.0848* 0.3767* 0.1874* 0.2149* 0.2883* 0.7485* 0.7600* 0.67360'O(1@ 8# ddPro(e ;@( Confidence Intervals ddPro(@*Desirability Functions ddPro(@*Most Desirable in Grid ddPro(@* Reset GridddPro(@)=Coc  UUUU A "7 1r@ 8# ddPro(b v@(Save Residuals ddPro(@*Save Predicted ddPro(@*Grid of Predicted  A@ H"i ddPro0`  ($ ddPro0> t}+p".@d dPro 333333 q*J0T-J*T-Q0T-J"-lddPro #E ddPro( &' (!Term ddPro( #24* Intercept ddPro( />* Run ddPro( ;J#* temp ddPro( G&~(!JEstimate ddPro( #C2~(-F 47.279855 ddPro( /F>~+ -4.009982 ddPro( ;FJ~* -22.86751 ddPro(| &(! Std Error ddPro(x #2+ 17.01647 ddPro(x />* 1.874967 ddPro(x ;J* 11.09245 ddPro(H &(!t Ratio ddPro(> #2+  2.78 ddPro(? />(9 -2.14 ddPro(? ;J*  -2.06 ddPro( '("Prob>|t| ddPro( $3+ 0.0167 ddPro( 0?* 0.0537 ddPro( <K* 0.0616 ddPro( &V(! Lower 95% ddPro( #2V+ 10.204105 ddPro( /">V+ -8.09519 ddPro( ;JV(E! -47.03593 ddPro( U'("X Upper 95% ddPro( $^3+ 84.355605 ddPro( 0^?* 0.0752264 ddPro( <^K* 1.3009043 ddPro(e &("Std Beta ddPro(X %39+  0 ddPro(X 1?lh*  -0.4586 ddPro([ =Ks(G-0.44205 ddPro( '(#VIF ddPro(% $2(. 0 ddPro(+ 0>(: 1.0163339 ddPro(+ <J* 1.0163339 N@DFGGGGGFD@@0N"FLG ( VY YddPro ddPro( ',  Helvetica .+$Prob>|t| ddPro(  &3+ 0.0167 ddPro(  2?* 0.0537 ddPro(  >K* 0.0616@ Y0N"H ddPro0  (Parameter Estimates ddPro( ' ($Term ddPro( &32* Intercept ddPro( 2?* Run ddPro( >K"* temp ddPro( E'x($HEstimate ddPro( &A3t(0D 47.279855 ddPro( 2D?u+ -4.009982 ddPro( >DKu* -22.86751 ddPro(~ '($ Std Error ddPro(z &3+ 17.01647 ddPro(z 2?* 1.874967 ddPro(z >K* 11.09245@ Y0"H@DFGGGGGFD@ ddPro(B -(*t Ratio ddPro(8 ,9+  2.78 ddPro(9 8E(B -2.14 ddPro(9 DQ*  -2.06 Y 1Y @ 8"N ddPro(O X (Std Err ddPro(*t Ratio ddPro(* Prob>|t| ddPro(*95% Confid LimitsddPro()S ddPro((G Stdized Beta ddPro(*VIF Y*3-<BHH3-<;"+      *3-<BHH3-<;      *3-<BHH3-<;      Lm'D Xm  m!l#  l#  l# # # #Ln2x "dK KdddPro ddPro0      #$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOQTWXY[^_`abcdefghijklmnopqrstuvwxyz{|}~ ,  Helvetica .+Cochran-Mantel-Haenszel Tests@ dK0" ddPro( %@ (!Stratified by ddPro( >%t)> Protein/Fat ddPro( #18(-CMH Test ddPro( /=g* Correlation of Scores ddPro( ;I* Row Score by Col Categories ddPro( GU* Col Score by Row Categories ddPro( Sa* General Assoc. of Categories ddPro(j #1(- Chi-Square ddPro(T /=+ 0.0324 ddPro(T ;I* 0.4204 ddPro(T GU* 1.4812 ddPro(T Sa* 5.0559 ddPro( #1(-DF ddPro( /=+  1 ddPro( ;I*  2 ddPro( GU*  2 ddPro( Sa*  4 ddPro( #1>(- Prob>Chisq ddPro( /=E+ 0.8572 ddPro( ;IE* 0.8104 ddPro( GUE* 0.4768 ddPro( SaE* 0.2816@ dK0dK"]BL-  ~c cddPro ddPro0 +,  Helvetica .+Tests@ c00"( ddPro( */ +Source ddPro( (6** Model ddPro( 4B"* Error ddPro( @N-* C Total ddPro( LZ@* Total Count ddPro( b*}(&eDF ddPro( )U7}(3X 4 ddPro( 5QC|(?T 48 ddPro( AQO|*  52 ddPro( MQ[|*  54 ddPro( |*(&-LogLikelihood ddPro(h (6+ 4.169230 ddPro(l 4B(> 37.283964 ddPro(l @N* 41.453194 ddPro($ *(& RSquare (U) ddPro(  )7+ 0.1006 ddPro( bp"(l Test ddPro( n|U* Likelihood Ratio ddPro(| z1* Pearson ddPro( `^n(ja ChiSquare ddPro( lvz+  8.338 ddPro(~ xv*  8.337 ddPro([ `n(j Prob>ChiSq ddPro(> lz+  0.0799 ddPro(~> x*  0.0800 ddPro(k W(JWarning: 20% of cells have expected count less than 5, Chi-squares suspect@ c0c"TL0~0 L1d4 m  m!l#  l#  l0|h h|ddPro ddPro0  "P,  Helvetica .( Crosstabs@ |h0 "U"K ddPro(@   @(UTaste |hC `C `C ` @ @ ddPro(a &3(0Type ddPro( 6C?(@Count@ |h 55/#@ AA/# ddPro( AN?+ Bland@ |h MM/# ddPro( MZI* Medium@ |h YY/# ddPro( Yf\* Scrumptious@ |h ee/#@ qq/#@"5Y<# 17ZA 8 ddPro( 6fC(@iBeef@ |h"5<# 17A 8 ddPro(a 6C)6Meat@ |h"5<# 17A 8 ddPro(0 5B(?Poultry@ |h"5<#@"51<# ddPro(z AN(K3 ddPro(D AN)66 ddPro( AN)61 ddPro( M}Z(W16 ddPro(D MZ)<8 ddPro( MZ)015 ddPro(z Yf(c1 ddPro(D Yf)63 ddPro( Yf)61 ddPro( AN1(K"10 ddPro( MZ1* 39 ddPro( Y%f2+ 5 ddPro( e}r(o20 ddPro(J er)617 ddPro( er)617 ddPro( er1)654 |hX h@DFGGGGGFD@@0 y7"o,@"5<# ddPro0 d ( Taste By Type@ |h0l"_ ddPro(@   @(Taste |hu u u  @ @@"9# ddPro( (5+%[0@ |h"9# ddPro(N 3(0.25@ |h"~9# ddPro(} y4("0.5@ |h"N9# ddPro( IV3(S0.75@ |h"9# ddPro( ('5($+1 |h """"1 < 8LLL wwww1< 8@ UUUU" #@ UUUU#  #@ UUUU# #@ UUUU# # """"1, 8@ UUUU",> #@ UUUU! #@ UUUU# #@ UUUU!, #YYY UUUU1 * 8@ UUUU" > #@ UUUU#  #@ UUUU# #@ UUUU# #@ 9 #@! #@!9#@!9#@"# ddPro( Rq+*Beef@ |h"# ddPro(e )FMeat@ |h"# ddPro() )<Poultry ddPro(m  (Type@ |h" # ddPro(0 2(Bland@ |h"1 # ddPro( p}<(zMedium@ |h" # ddPro( #0O(- Scrumptious |h 1 OX 8333333 wwww1OW 8@ UUUU"O #@ UUUU#! #@ UUUU# #@ UUUU# #333 UUUU12OW 8@ UUUU"2O #@ UUUU!W #@ UUUU# #@ UUUU!2O #YYY UUUU1 O0W 8@ UUUU" O #@ UUUU# #@ UUUU# #@ UUUU# #@"L#@!Z#@##@!L#@ a#@! a#@! #@!#o@DFGGGGGFD@@0|h#F NHH NtU   *3-<BHH3-<;)x2      *3-<BHH3-<;x$      *3-<BHH3-<;x       ddPro(w N ( Mosaic Plot ddPro(* Crosstabs ddPro(*Tests ddPro(*Correspondence AnalysisddPro()| ddPro((KCochran-mantel-Haenzel... |h """"1P 8 """"1O@ 8 """"""1O 8# # # #Ln2x R RddPro@ R"*bH#@##@"$#@"B#@#$#@";{#@"|#Q(.@X#@"5#@"5#@"# "#### R"# "5#### R"5# ddPro$ $3:^,  Helvetica .+6- standard ddPro$* error P "#5### R"5#P ";#5### R";5# ddPro$H !7(*overlap ddPro$* marks@ R0Q"4 ddPro$ ;($95% ddPro$* confidence ddPro$* intervalP "B#o### R"Bo#P "#m### R"m# ddPro$ :0(# standard ddPro$* deviation ddPro$* of groupP "F|#### R"F|#P " |### # R" |# ddPro 1DQ ddPro0 CR (N.ddPro ddPro$ >{J (G~...... ddPro$ {( ~......L&L h   ddProP "##### #  "## # ddPro$  ,  Helvetica .+ 2  "# #  " # ddPro$  )2LDL.+hLb {\Lv% %F<R R<ddPro ddPro$ ?,  Helvetica .+ Matching Fit@ <R0K"< ddPro$ !'(Source ddPro$ ,>* Whole Model ddPro$ *7* time ddPro$ 5B* id ddPro$ @MY* Error (no Interaction) ddPro$ KX2* Interaction ddPro$ Vc`* Error (with Interaction) ddPro$x !(SS ddPro$ k,((n 64.30518 ddPro$ *k7* 12.76281 ddPro$ 5kB* 51.51195 ddPro$ @kM* 21.04439 ddPro$ KkX* 21.04439 ddPro$ Vnc+ 0.00000 ddPro$R !(DF ddPro$W ,(( 18 ddPro$T *7+  3 ddPro$W 5B(> 15 ddPro$W @M*  44 ddPro$W KX*  44 ddPro$T Vc+  0 ddPro$  !(MS ddPro$9 ,(( 3.572510 ddPro$9 *7* 4.254270 ddPro$9 5B* 3.434130 ddPro$9 @M* 0.478282 ddPro$9 KX* 0.478282 ddPro$$ Vc+ ddPro$ !(F Ratio ddPro$ ,$+  7.469 ddPro$ *7$*  8.895 ddPro$ 5B$*  7.180 ddPro$ @ M(+  ddPro$ K X(*  ddPro$ V c(*  ddPro$ %!J((Prob>F ddPro$ *,O+  0.0000 ddPro$ **7O*  0.0001 ddPro$ 5*BO*  0.0000 ddPro$ @9MR+  ddPro$ K9XR*  ddPro$ V9cR*  ddPro$ erD(n Intercept ddPro$ p}H+ -1.982111@ <R"aP#@##@##@## ddPro$ e^rx(natime ddPro$ p^}* LogHist0 ddPro$| {^* LogHist1 ddPro$q ^* LogHist3 ddPro$f ^* LogHist5 ddPro$n er(nEstimate ddPro$k p}+ -0.730528 ddPro$|m {( 0.4714242 ddPro$qm * 0.1948956 ddPro$fm * 0.0642083@ <R"aSy#@#?#@##@## ddPro$& er(nid ddPro$& p}* 1 ddPro$|& {* 2 ddPro$q& * 3 ddPro$f& * 4 ddPro$[& * 5 ddPro$P& * 6 ddPro$E& * 7 ddPro$:& * 8 ddPro$/& * 9 ddPro$$& * 10 ddPro$& * 11 ddPro$& * 12 ddPro$& * 13 ddPro$&  * 14 ddPro$&  * 15 ddPro$& "* 16 ddPro$ er(nEstimate ddPro$  p} + -0.432046 ddPro$|  { * -1.655259 ddPro$q  !+ 0.861142 ddPro$f  !* 0.779357 ddPro$[   ( -0.197105 ddPro$P   * -0.206528 ddPro$E   * -0.715687 ddPro$:   * -0.799805 ddPro$/  !+ 0.314635 ddPro$$  !* 0.911285 ddPro$  !* 1.231842 ddPro$  !* 1.897813 ddPro$   ( -0.428940 ddPro$    * -0.428940 ddPro$    * -0.163722 ddPro$  " * -0.968041@ <R"a_#@!$+#@##@!a# ddPro$ (54(1  5 ddPro$ (*5\) Iterations@ <R0<R"1Y ddPro ddPro(9Ā 9,  Helvetica .+ Y By Time@ 0O"6 ddPro$  +^Y g pg pg p00@":# ddPro$O 1+C-4.000@ ":#@":# ddPro$i 1(-3.000@ ":#@"z:# ddPro$ t1(}-2.000@ "m:#@"a:# ddPro$ [h1(d-1.000@ "T:#@"G:# ddPro$ AN1(J0.000@ "::#@"-:# ddPro$ '41(01.000@ " :# 1!; 8@ UUUU z:z #1NP81qxsz818181NP81xz818181NP81?xAz81Z\81ln81uNwP81VxXz81bd81ln81NP81xz81|~81z|81~NP81xz8181NP81xz818181NP81xz818181NP81TxVz81eg81np81NP81FxHz81OQ81TV81NP81LxNz81EG81MO81NP81*x,z815781BD81NP81xz818181NP81xz818181|N~P81xz81~81~81NP81xz81818@"N*#@#)#@#*#@"N*#@#)#@#*#@"N*#@#)#@#*#@"uN*#@#) #@#* #@"N*#@#)#@#*#@"~N*#@#)#@#*#@"N*#@#)#@#*#@"N*#@#)#@#*#@"N*#@#)#@#* #@"N*#@#) #@#*#@"N*#@#)#@#*#@"N*#@#) #@#* #@"N*#@#)#@#*#@"N*#@#)#@#*#@"|N*#@#)#@#*#@"N*#@#)#@#*#@ : #@!#@!:#@! :#@"d# ddPro$D 7`+#LogHist0@ "# ddPro$D b)+LogHist1@ "# ddPro$Ds )(LogHist3@ "# ddPro$DI )*LogHist5 ddPro$6 {(~Time@  #@!#@!#@!#@0#     ) !"#$%&'(+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~B  HH )^=^^^^-^^=^^^^A^=^^^^M^^=^^^^W^^=^^^^^^ =^^^^^^ =^^^^^=^^^^^^ =^^^^^^=^^^^^^=^^^^^=^^^^^^=^^^^^^=^^^^^^=^^^^^^=^^^^^=^^^^^^=^^^^c^^=^^^^a^^=^^^^-^=^^^^/^^=^^^^/^^=^^^^\^^=^^^^e^^=^^^^^=^^^^^^= ^^^ ^^^= ^^^^^^= ^^^^^=^^^^^^=^^^^{ ^^=^^^^u^^=^^^^c^^=^^^^-^=^^^^-^^=^^^^$==^=^=^=^='=^^^^'=^^^^=^^^^=^^^^= ^^^^=)^^^ ^= ^^^  ^= ^^^ ^= ^^^  ^=^^^^9=^^^^5=^^^^'=^^^^'=^^^^'=^^^^^^^'=^^^^'=^^^^=^^^^=^^^^= ^^ ^ ^=)^^ ^ ^= ^^ ^  ^= ^^ ^ ^= ^^^  ^=^^^^9=^^^^5=^^^^'=^^^^'=^^^^'=^^^^^^^'=^^^^'=^^^^=^^^^=^^^^= ^^^ ^=)^^^ ^= ^^^  ^= ^^^ ^= ^^^  ^=^^^^9=^^^^5=^^^^'=^^^^'=^^^^'=^^^^^^^'=^^^^'=^^^^=^^^^=^^ ^^= ^^^ ^=)^^^ ^= ^^^  ^= ^^^ ^= ^^^  ^=^^^^9=^^^^5=^^^^'=^^^^'=^^^^'=^^^^^^^'=^^^^'=^^^^=^^^^=^^^^= ^^^^=)^^^ ^= ^^^  ^= ^^^ ^= ^^^  ^=^^^^9=^^^^5=^^^^'=^^^^'=^^^^'=^^^^^^^'=^^^^'=^^^^=^^^^=^^^^= ^^^ ^=)^^^ ^= ^^^  ^= ^^^ ^= ^^^  ^=^^^^9=^^^^5=^^^^'=^^^^'=^^^^'=^^^^^^^'=^^^^LkIM)xm  m!l#  l#  l# # # " ddPro ddPro(@,  Helvetica .@(PShrink 9a9a9a @@@@ @"3# ddPro$l $1 +"D0@ "s3# ddPro$ p$|1(y'1@ "Y3# ddPro$ V$b1(_'2@ ">3# ddPro$ ;$G1(D'3@ "$3# ddPro$ !$-1(*'4@ " 3# ddPro$ $1('5  1 5q "?"ZE"T?"Z?"@?"TE"@?#"d?";]"R\">]"Rc">]"x]#"R]&"] UUUU"R2= 1UBWD1XBZD1RBTD91`b1M`Ob1H`Jb1`b1`b1=`?b1`b1E`Gb1ZB\D1]B_D1`BbD1ZB\D1@`Bb1=`?b1@`Bb91XBZD1 B"D1uBwD1=B?D1%B'D1]B_D1BD1UBWD1Z`\b1]`_b91``bb@" 3=#@!p#@##@! 3#@"Q# ddPro$b ;K+-1@ "p# ddPro$b [h) 1 ddPro(H B_ (EGate  1 q 1 r  8"# ##############";######## ####"###1UW81XZ81RT81RT8181MO81HJ818181=?8181EG81Z\81]_81`b81Z\81@B81=?81@B81@B81XZ81 "81uw81=?81%'81]_8181UW81Z\81]_81]_81`b8"p"f@ p  #@! #@!p#@! p#@"p# ddPro$b iy (l-3@ "# ddPro$bz )-2@ "# ddPro$ba )-1@ "# ddPro$bE )0@ "# ddPro$b, )1@ "# ddPro$b )2@ " # ddPro$b )3 ddPro(H n (q Normal Quantile@   #@!#@! #@! # ddPro( (RShrink >`>`>` @@@@ @")#@"r)#@"X)#@"=)#@"#)#@")# ddPro$m & +A0 ddPro$ o{&(x1 ddPro$ Ua&(^2 ddPro$ :F&(C3 ddPro$ ,&()4 ddPro$ &(5  1 +g "<5"Z5"A5"V5#"A5"Z;"V5"]5"S"NS":S"NY":S"|S#"NS."S UUUU"Q(= 1T8V:1W8Y:1Q8S:91VX1LVNX1GVIX1VX1VX1<V>X1VX1DVFX1Y8[:1\8^:1_8a:1Y8[:1?8A:1<8>:1?8A:91WVYX1V!X1tVvX1<V>X1$V&X1\V^X1VX1TVVX1Y8[:1\8^:91_8a:@")=#@!f#@##@!)#@"G# ddPro$c 1A+-1 "e ddPro$c Q^) 1 ddPro(I 5X (8HTime  1 g1 h ":###############"# ########### ###1TV1WY1QS1QS11LN1GI111<>11DF1Y[1\^1_a1Y[1?A1<>1?A1?A1WY1!1tv1<>1$&1\^11TV1Y[1\^1\^1_a"N@ f#@!#@!f#@## ddPro$c _o (b-3@ "# ddPro$c y)-2@ "# ddPro$ck )-1@ "# ddPro$cO )0@ "# ddPro$c6 )1@ "# ddPro$c )2@ "# ddPro$c )3 ddPro(I d (g Normal Quantile@  #@!#@!#@!#ddPro 1 ddPro0?  @(.ddPro#Ln2xL.)"h`DUL U0r0R RddPro ddPro04@ ,  Helvetica .+Transformed Fit Log to Sqrt@ R0" ddPro(#@  % (!,Log(weight) = 0.11353 + 0.57171 Sqrt(height) ddPro0@ )9f +Summary of Fit@ R0'9r"4c@0$<u# ddPro0@ CS*Analysis of Variance@ R0AS"N ddPro(@ Uc. (_ Source ddPro(@ ao)* Model ddPro(@ m{!* Error ddPro(@ y,* C Total ddPro(@ UHc\(_KDF ddPro(@ aIo`+  1 ddPro(@ mE{^(wH 38 ddPro(@ yE^*  39 ddPro(@ Ubc(_eSum of Squares ddPro(@ ao+! 0.9448466 ddPro(@ m{* 0.7820163 ddPro(@ y* 1.7268629 ddPro(@H Uc (_ Mean Square ddPro(@. ao + 0.944847 ddPro(@. m{ * 0.020579 ddPro(@ UcE(_F Ratio ddPro(@ aoC(k 45.9123 ddPro(@ m{B+ Prob>F ddPro(@ yF+  <.0001@ R0>L"C ddPro0@  ( Parameter Estimates@ R0"@0# ddPro0@ *Fit Measured on Original Scale@ R0" ddPro(}@ m ( Sum of Squared Error ddPro(q@ z* Root Mean Square Error ddPro(e@ 5* R-square ddPro(Y@ [* Sum of Residuals ddPro(}@ ( 9349.3005 ddPro(q@ * 15.685478 ddPro(e@ * 0.5136652 ddPro(Y@ + 42.56216@ R0"@0R# ddPro ddPro0 k,  Helvetica .+weight By height@ 0x"h ddPro(@  @(eweight RqRqRq (((   @"2# ddPro(d ++760@ "2# ddPro(x ~+(80@ "p2# ddPro( jx*(t100@ "[2# ddPro( Uc*(_120@ "G2# ddPro( AO*(K140@ "32# ddPro( -;*(7160@ "2# ddPro( '*(#180 1 3 81uuww81Y[81WY81CE81@B81|~81TV819;81d|f~81ik81^`81rt81km81umwo81uw8181y{81FH818181wy81qs81]_81xz81df81qs81ce81xz81df81TV81eg81km81ln81jl81df81a|c~81TV81`b81NP81(*8@"9##@##@#$#@##@##@##@##@##@##@##@##@##@ 2#@#y#@!2#@##@"2# ddPro(X (8+50@ "E#@"W# ddPro(X M])%55@ "j#@"}# ddPro(X s)&60@ "#@"# ddPro(Xe )%65@ "#@"# ddPro(X@ )%70@ "# ddPro(J q(theight@  #@!#@!#@!#@DFGGGGGFD@ ddPro(2 5(Fitting :J@DFGGGGGFD@@"N# ddPro(2 _)KTransformed Fit Log to Sqrt@  #@##@!#@##@0#LA31!BALpv ! t t t{HH tt&&&&&&&&&&&& ' ' '                          "   "   " 9   5A2&&&&&&&&&&&&&&&&&&&nt                                  &&&&&&&&&&&&&&&&&&&&&&&&>                                                                  r      b>&&&&&&&&b                                          !      D&&&&&&&2                  pzn2&&&&&&&>      m      q2&&&&&&&JN""""""            $            N      k,&&&&&&&>~------f------'      T! ! ! ! ! ! ~! ! ! ! ! !       q2&&&&&&&&&222AMbtt      +mO   [      a      y   zzzJ222&&&&&&&&&&&5 ddPro P?HH ?%141111141% #&&&&&&&25))) ddPro( ,  Helvetica .+! Show Points ddPro(* Fit Mean ddPro(* Fit Line ddPro(*Fit Polynomial ddPro(*Fit Transformed... ddPro(* Fit Spline ddPro(*Density Ellipses ddPro(*Nonpar Density ddPro(*Paired t Test ddPro(* ddPro(*Grouping Variable... wwww"x*3-<BHH3-<;       @AL HH ALXc@AL HH ALKV@AL HH AL-8] ddPro@ 0#@0# ddPro$ ,  Helvetica .+Nonparametric Bivariate Density ddPro$   (Variable ddPro$  (,($ Sepal length ddPro$  &3)@(/ Petal length ddPro$ # X(& Kernel Std ddPro$ **V+ 0.181358 ddPro$ &*3V* 0.297226 ddPro$ X w([Slider ddPro$ T)(%W0.181358 ddPro$ &U3@+ 0.297226@ 0#"/~1"@0'.#1(. @ ddPro$\ (Apply 3iHH3iJ #"""""""""""/3GTVrx{                | XVzvU t t Cd_5NN #,* @05"@0#@ '#@!#@"'#@"'#@"'#@"{'#@"o'#@"e'#@"Z'#@"P'#@"E'#@":'#@"0'#@"$'#@"'#@ ''# ddPro$@ @(b Petal length F yF yF y? ??  $$$?@"'# ddPro$] %+A1.0 ddPro$s %(2.0 ddPro$ n{%(w3.0 ddPro$ Yf%(b4.0 ddPro$ CP%(L5.0 ddPro$ .;%(76.0 ddPro$ %%(!7.0@  '#@## ddPro$K 0+ 4.0@ "=# ddPro$K 3F)4.5@ "S# ddPro$K H[)5.0@ "h# ddPro$K ]p)5.5@ "~# ddPro$K s)6.0@ "# ddPro$Kt )6.5@ "# ddPro$K^ )7.0@ "# ddPro$KI )7.5@ "# ddPro$K4 )8.0 ddPro$A X@([ Sepal length ddPro$ ( Petal length By Sepal length@ 0"@ #@!#@ #@ #@"#@!# ddPro$1 b@+\Quantile Density Contours ""@@@"#ffffff",̌̌"5fff">"Gٙٙٙ"P222"Y ddPro$2 @(1 ddPro$2 !@) 2 ddPro$2 *@) 3 ddPro$2 (3@) 4 ddPro$2 1<@) 5 ddPro$2 :E@) 6 ddPro$2 CN@) 7 ddPro$2 LW@) 8 ddPro$2 U`@) 9@ "#dG^HHG^ *!!!!/9/;7CEJ\[ _ W e ^jgdej f ken n o qc o y ojyleZ g O N X a [ ae[ ]_ M Q E><32)%!-      /;&>F6Ob V\ b l n u s[ o <R O&./ L@T$ m  m!l#  l#  l# # # #Ln2xLw: !( LjJu0* XS~ ddPro ddPro0 T,  Helvetica .+  Frequencies@ 0]"O ddPro( '( ($Level ddPro$ &2 * 11 ddPro$ 1=* 12 ddPro$ <H* 13 ddPro$ GS* 14 ddPro$ R^* 15 ddPro$ ]i* 16 ddPro$ ht!* Total ddPro( 0'U ($3Count ddPro$ &D2] +  2 ddPro$ 1D=]*  5 ddPro$ <DH]*  3 ddPro$ GDS]*  4 ddPro$ RD^]*  4 ddPro$ ]Di]*  1 ddPro$ hAt\(qD 19 ddPro( [' ($^ Probability ddPro$ &t2 + 0.10526 ddPro$ 1t=* 0.26316 ddPro$ <tH* 0.15789 ddPro$ GtS* 0.21053 ddPro$ Rt^* 0.21053 ddPro$ ]ti* 0.05263 ddPro(^ ' ($ StdErr Prob ddPro$? &2 + 0.07041 ddPro$? 1=* 0.10102 ddPro$? <H* 0.08365 ddPro$? GS* 0.09353 ddPro$? R^* 0.09353 ddPro$? ]i* 0.05123 ddPro( ' ($Cum Prob ddPro$ &2& + 0.10526 ddPro$ 1=&* 0.36842 ddPro$ <H&* 0.52632 ddPro$ GS&* 0.73684 ddPro$ R^&* 0.94737 ddPro$ ]i&* 1.00000 ddPro$ s3(| 6 ddPro$ s&K) Levels@ 0("|F`p@DFGGGGGFD@"_HH"_>?            ddPro( U? (XCount ddPro(* Probability ddPro(* StdErr Prob ddPro(*Cum Prob *3-<BHH3-<; DR      *3-<BHH3-<;DT      *3-<BHH3-<;1C:Q      2B@DFGGGGGFD@ ddPro ddPro0 {,  Helvetica .+Test Probabilities@ 0~"m ddPro( '* +Level ddPro$ $2 * 11 ddPro$ /=* 12 ddPro$ :H* 13 ddPro$ ES* 14 ddPro$ P^* 15 ddPro$ [i* 16 ddPro( /'r ("2 Estim Prob ddPro$ $I2r + 0.10526 ddPro$ /I=r* 0.26316 ddPro$ :IHr* 0.15789 ddPro$ EISr* 0.21053 ddPro$ PI^r* 0.21053 ddPro$ [Iir* 0.05263 ddPro( t' ("w Hypoth Prob ddPro$T $2 +5  ddPro$f /=(81.00000 ddPro$f :H* 1.00000 ddPro$f ES* 1.00000 ddPro$T P^+  ddPro$T [i*  ddPro$ ft(o,Click then Enter Hypothesized Probabilities. ddPro$ q* 0Omitted values will use estimated probabilities. ddPro${ |* =The remaining values will be scaled to sum to 1 minus omitted@  @6" ddPro0j 0 + Done@ @<f"- ddPro0j ?_)4Help@ 0"\ ddPro0 @(Test Probabilities@ 0"| ddPro( )9 @+Level ddPro$ &4% @* 11 ddPro$ 1?%@* 12 ddPro$ <J%@* 13 ddPro$ GU%@* 14 ddPro$ R`%@* 15 ddPro$ ]k%@* 16 ddPro( >) @($A Estim Prob ddPro$ &X4 @+ 0.10526 ddPro$ 1X?@* 0.26316 ddPro$ <XJ@* 0.15789 ddPro$ GXU@* 0.21053 ddPro$ RX`@* 0.21053 ddPro$ ]Xk@* 0.05263 ddPro(z ) @($ Hypoth Prob ddPro$W &4 @+# 0.10526 ddPro$W 1?@* 0.21053 ddPro$W <J@* 0.21053 ddPro$W GU@* 0.21053 ddPro$W R`@* 0.21053 ddPro$W ]k@* 0.05263 ddPro( l{7 @(vTest ddPro$ x_ @* Likelihood Ratio ddPro$t A@* Pearson ddPro( ld{ @(vg Chi-square ddPro$~ x @+  0.5053 ddPro$t~ @*  0.5000 ddPro(H l{ @(vDF ddPro$H x @*  2 ddPro$tH @*  2 ddPro(/ l{ @(v Prob>Chisq ddPro$  x @+$ 0.7767 ddPro$t  @* 0.7788@ 0"m  m!l#  l#  l# # # #Ln2xL+!$,= =ddPro = 1S1S2@ ":#@!S#@##@!# 1 T28!4HH4!!4 S2-0 0 00+0 0 00-0 0 0<-0 0 0<.0 0 0< .0 0 0< 9  68!?|886D"B"C"D"BD"6A"xA CA A" @A <6A"A AA A" @A p6A#AxA@A# @A 6A& AAAFAf F#6A A@A:A@:6A A@AA  @6A$A@AAD!B!<73D"DB$" G""LB 1b1p738C?;p8C8  <x  <x <<<<x   <<x<<   8    8p    ??<p    < p  ? ? ??< x @` ` `x  x        << ?`!  @` ?sB  p? < s?  '#~w|o x? ?        ?; # <       ? ?    ? ? ?8 <   w n <8 x ? ?À G  0 xp    ? ?  x`  }@< >   8   ;?  ? ? ` C    ?  x 0 < <x  xp< x  ` `x> >     x  8  ?p ?` ? ? 8<p<   <p < p    < x  x   ?x?<<  8     8   < p  <p  <x<<< <     @ S2#@!2#@!S#@!S#@"S# ddPro(% K\,  Helvetica .+N-4@ ="o# ddPro(% gx)-3@ ="# ddPro(%z )-2@ ="# ddPro(%^ )-1@ ="# ddPro(%A )0@ ="# ddPro(%& )1@ ="# ddPro(%  )2@ ="# ddPro(% )3@ ="2# ddPro(% +9)4 ddPro( P(S Normal Quantile@ = 7#@!7#@!#@!#@"# ddPro(6 (-7@ ="#@"# ddPro(D (-6@ ="#@"# ddPro(S (-5@ ="#@"# ddPro(a (-4@ ="#@"# ddPro(p (-3@ ="#@"~# ddPro(~ x(-2@ ="v#@"o# ddPro( ix(s-1@ ="h#@"a# ddPro( [j(e 0@ ="Z#@"R# ddPro( L[(V 1@ ="K#@"D# ddPro( >M(H 2@ ="=#@"5# ddPro( />(9 3@ =".#@"'# ddPro( !0(+ 4@ =" #@"# ddPro( !( 5@ ="#@" # ddPro( ( 6 =n47HH40Kcccccccc  c c cc cc cc cc cc cc cc c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c cc cc c cc c cc c cc c cc c cc c cccccc cc cc cc cc cc cc cc cc c ddPro0 6= (9.L+X!LE,!      !"#$%&'()*+,-./01234@789:;<=>?ABCDFGHIKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvyz{|}~)> >ddPro@ >"# ddPro(7 ,  Helvetica .+ 0@ >"#@"# ddPro(R ( 1@ >"#@"# ddPro(l ( 2@ >"#@"u# ddPro( o ~(y 3@ >"g#@"Z# ddPro( T c(^ 4@ >"M#@"?# ddPro( 9 H(C 5@ >"2#@"%# ddPro(  .() 6@ >"#@" # ddPro(  ( 7@ >":#@!T#@##@!# 1T31 U38!HH4!! T3-0 0 00-0 0 00+0 0 0<-0 0 0<-0 0 0<+0 0 0<38!>|885D"B"@"D"BD"6A"xA BA A" @A 4A"A CA A" @A 4A#A{A@A# @A 6A& AAAFAf F#6A A@A:A@:4A A@AA @3A$A@AAD!B!6.D"DB$" D""LB 1b16.8C?:p8C8      < <    < < < <     ? ? < <          ??< <9r     8 p    8 p    8 p      8 p    8 p  8 p     <8<p<<   8p??? ?  8p   ?8?p$$$ 4  8p @  p 0@8@p     8 p   /|< <8 |p    x <8 pOÀ   ?<8p  0 p  <   0 8p 0 ?  `  @ ; s   8 ; s  Á? ? 8  q?  A? 8 0 8 q A???8 0 8  s c 0Np? ? :? w?  # wp n` :w ?? # wp n`  : w   # wp n`  : w   #? wp? n`  : w ? ? # wp n`  : w   # wp n`  :> w?   # wp n`  : w ? ? # wpno 8? s   c ?  0 ? ?   8 08 q@ T3#@!3#@!T#@!T#@"T# ddPro(& L]+C-4@ >"p# ddPro(& hy)-3@ >"# ddPro(&y )-2@ >"# ddPro(&] )-1@ >"# ddPro(&@ )0@ >"# ddPro(&% )1@ >"# ddPro(&  )2@ >"# ddPro(& )3@ >"3# ddPro(& ,:)4 ddPro( Q(T Normal Quantile@ > 8#@!8#@!#@!#nPHHPJ  c  cc  cc  cc  cc  cc  cc  cc  cc  cc  cc  cc  cc  cc  c c c c c c c c c c c c c  c  ccc  ccc  ccc  ccc  ccc  ccc  ccc  ccc  ccc  ccc  ccc  ccc c c c c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c c c c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c c c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c c c cc cc cc cc cc cc cc cc cc cc cc cc cc c c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c cc c c ddPro0 7> +.*A AddPro@ A"#@"# ddPro(C ,  Helvetica .+ -3@ A"#@"# ddPro(^ ( -2@ A"#@"# ddPro(y } ( -1@ A"u#@"h# ddPro( b q(l0@ A"[#@"M# ddPro( G V(Q1@ A"@#@"2# ddPro( , ;(62@ A"%#@"# ddPro(  (3@ A" # 1W@ ":#@!W#@##@!# 1W61 X68!gHH4!!g W6-0 0 00+0 0 00+0 0 00-0 0 00-0 0 00+0 0 0038!>|885D"B"@"D"BD"6A"xA CA A" @}#4A"A CA A" @}'4A#AzA@A# @}.6A& AAAFAf F?6A A@A:A@:>4A A@AA @p3A$A@AAD!B6.D"DB$" D""LB 1c6.8C?8p8C8   8 p  À   <8 <p < <  8 p   < <8 < 858!>|8?85D"B"@"D"r>D"4A"xA @A Eb@ @A 4A"A BA O @A 6A#A{A@M @A 6A& AAAFE?F#4A A@A:A@:4A A@AA@3A$A@AQ O!B!6.D"DB$" D":OO@1b16.8C?8p8 8 1s a   ќ   @ `  L@  ; 7 ;r 1`3 ;  ݘ π  @   |@| z    ?    _@  |@| ~      O@   ~@>> ?     G  \@ @:@ ?  c 9     p 8 6 v f  9 3      ? 8? 6? v f   9 3     ? 8 6 v f  113    O  8O 0? p?  c   ! qc #     o o8o 0pa?  ?a?a<c<     c o 8o  0pa  ? a?ac   c 8c 0c 8p/ ` ÿ???       `' 8? @0p? `?    8  0       80 sc  Ð ` ` `>@  ??    8??? ?  0 p ;   8 0p@ `@ Z9#@!9#@!Z#@!Z#@"Z# ddPro(% Rb+L-4@ C"v# ddPro(% n~)-3@ C"# ddPro(%s )-2@ C"# ddPro(%W )-1@ C"# ddPro(%: )0@ C"# ddPro(% )1@ C"# ddPro(% )2@ C"# ddPro(% #)3@ C"9# ddPro(% 2?)4 ddPro( W(Z Normal Quantile@ C =#@!=#@!#@!#nBHHB P  c c c c c c c c c c c c c c c c c c!! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c!! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c!!! c c c c c c c c c c c c c c c c c!! c c c c c c c c c c c c c c c c c c!! c c c c c c c c c c c c c c c c c c!!! c c c c c c c c c c c c c c c c c c ! c c c c c c c c c c c c c c c c c ! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c!!! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! c! ! ddPro0 =C @+.L,V!<L#H"dts 4s sddPro ddPro0  ddPro ddPro0= 0?,  Helvetica . +;. ddPro0 :(Normal@ 0<"1@"-#@"-# ddPro(! * +-3@ "-#@"-# ddPro(= *(-2@ "-#@"-# ddPro(Z *(-1@ "-#@"-# ddPro(v *(0@ "x-#@"j-# ddPro( ds*(n1@ "\-#@"M-# ddPro( GV*(Q2@ "?-#@"1-# ddPro( +:*(53@ "#-# 1-1$. UUUU1-.@ 8# UUUU1-3@ 8# UUUU1-;@ 8# UUUU1-J@ 8# UUUU1-[@ 8# UUUU1-`@ 8# UUUU1x-h@ 8# UUUU1j-y`@ 8# UUUU1\-kL@ 8# UUUU1M-]7@ 8# UUUU1?-N2@ 8# UUUU11-@.@ 8# UUUU1#-2/@ 8#@"-^#@!#@##@!-# 11$ 8@"t#@"#@"#@"t&#@"t&#@ UUUU"t #@ UUUU"2 #@"#@##@##@"#@##@# #@# #18169810381/281-08@".#@!#@##@!#@ #@!#@!#@!#?O@DFGGGGGFD@@0#ddPro 1-; ddPro0P ,< +jittered pointsddProddPro 1 ddPro0uO +T means diamondddPro w,  Helvetica .+Test Mean=value@ s0z"i ddPro$ $[ *Hypothesized Value ddPro$ !/J* Actual Estimate ddPro$ ,:M* Using Std Dev of ddPro$ l$(o 60 ddPro$ !_/(*b62.3368 ddPro$ ,o:+  5 ddPro$ CQ@(LTest Statistic ddPro$ N\/* Prob > |z| ddPro$ Yg+* Prob > z ddPro$ dr+* Prob < z ddPro( 7EFn (AHz Test ddPro$ CKQn +  2.037 ddPro$ NK\n*  0.042 ddPro$ YKgn*  0.021 ddPro$ dKrn*  0.979 ddPro( 7pF (As Signed-Rank ddPro$e CQ +(  46.500 ddPro$b N\+  0.062 ddPro$b Yg*  0.031 ddPro$b dr*  0.969@ s0s"mLN|h hddPro ddPro0     "%&'),-/2568;<>ABCDEFGHIJLORSUXY[^_adghjmnpsvwxz}~ w,  Helvetica .+Test Mean=value@ h0z"i ddPro$ $[ *Hypothesized Value ddPro$ !/J* Actual Estimate ddPro$ l$(o 60 ddPro$ !_/(*b62.3368 ddPro$ 8F@(ATest Statistic ddPro$ CQ-* Prob > |t| ddPro$ N\)* Prob > t ddPro$ Yg)* Prob < t ddPro( ,F;m (6It Test ddPro$ 8JFm +  1.987 ddPro$ CJQm*  0.062 ddPro$ NJ\m*  0.031 ddPro$ YJgm*  0.969@ h0h"bjLV L4c$$(,8L3I"ttY     HH6   .4L   t nn |[UO        :F&)==dj          ?A            "I9 !! !   OD        02      u$*II      *,x     a"1) Q  = ))8! )#   5Z,%%,,e" hj   r#1;33Ia g r    d33773#  2ry yddPro yGW@DFGGGGGFD@ ddPro( *,  Helvetica .+Weight@ y0."%@"-# ddPro( (+60@ y"-# ddPro(3 ((80@ y"-# ddPro(T '(100@ y"-# ddPro(u '(120@ y"c-# ddPro( `m'(j140@ y"B-# ddPro( ?L'(I160@ y"!-# ddPro( +'((180 y 1- 8 1". 8 "/HHM;"/!-,. << +9, ,$ >-9<  <9.92 ",,+<<<<*,-3 ,$( .:..2'?<<>9U`-,U`U`-U`. U`U`9U`+2U`9U`.U`U` U`.>U`>"U`.<U` U`9 UX.9 UX UX UX UX$. UX< UX  UX UX UX, UX*, UX( UX<9 UX<:UX9 UX, U2 U(. U,<U- U.. U/( U U U9 U U. U< U6U< U.> U.U`..U`.>U`.U`-.U`..U`U`>U`-U`.U`. U`:U`. U`+U`U`.U`9@"-^#@!#@##@!-# 1 8 1" 8"HHM;"!``                 @@@@@@ @P@ABBDHPP`0@`0PPHHDBAA@@P@ @ @@@@@@@@@@            @".#@!#@##@!#@ #@!#@!#@!# ddPro(/ ( Stem and Leaf@ y0" ddPro(& #0+ Stem ddPro( #0)"Leaf ddPro( /<(917 ddPro( /<)2 ddPro( ;H(E16 ddPro( GT* 15 ddPro( S`* 14 ddPro( S`)2 ddPro( S` )5 ddPro( _l(i13 ddPro( _l)4 ddPro( kx(u12 ddPro( kx)3 ddPro( kx )8 ddPro( kx)8 ddPro( k x)8 ddPro( w(11 ddPro( w)1 ddPro( w )2 ddPro( w)2 ddPro( w )2 ddPro( w)2 ddPro( w$)3 ddPro( w*)5 ddPro( w#0)6 ddPro( w)6)9 ddPro(s (10 ddPro(s )4 ddPro(s  )5 ddPro(s )5 ddPro(s  )6 ddPro(s )7 ddPro(g (9 ddPro(g )1 ddPro(g  )2 ddPro(g )2 ddPro(g  )3 ddPro(g )5 ddPro(g $)5 ddPro(g *)8 ddPro(g #0)9 ddPro(g )6)9 ddPro([ (8 ddPro([ )1 ddPro([  )4 ddPro([ )4 ddPro([  )5 ddPro(O (7 ddPro(O )4 ddPro(O  )9 ddPro(O )9 ddPro(C (6 ddPro(C )4 ddPro(C  )7 ddPro( #C0f(-FCount ddPro( /C<P* 1 ddPro( SC`P*$2 ddPro( _ClP* 1 ddPro( kCxP* 4 ddPro( wCP* 9 ddPro(s CP* 5 ddPro(g CP* 9 ddPro([ CP* 4 ddPro(O CP* 3 ddPro(C CP* 2@ y o#@!o#@!#@!# ddPro(-( N(Multiply Stem.Leaf by 10@ y0y"I1A@DFGGGGGFD@@0#[HH[:F                   ddPro( =9(@ More Moments ddPro(* Test Mean = value ddPro(* Test Dist is NormalddPro()[ ddPro((5@Stem ddPro()and LeafL/~m  m!l#  l#  l# # # #vY YddPro ddPro( 5,  Helvetica .+Moments@ Y08"0 ddPro( %%+Mean ddPro( $11* Std Dev ddPro( 0=P* Std Error Mean ddPro( <I\* Upper 95% Mean ddPro( HUY* Lower 95% Mean ddPro( Ta* N ddPro( `mG* Sum Weights Y;K@DFGGGGGFD@@0p"jB ddPro(P (Moments@ Y0" ddPro(O &+Mean ddPro(O %2* Std Dev ddPro(O 1>* Std Error Mean ddPro(O =J* Upper 95% Mean ddPro(O IV* Lower 95% Mean ddPro(O Ub* N ddPro(O an* Sum Weights ddPro(O mz* Sum ddPro(}O y* Variance ddPro(qO * Skewness ddPro(eO * Kurtosis ddPro(YO * CV ddPro( &@(# 10.73333 ddPro( %2B+ 4.79176 ddPro( 1>B* 0.87485 ddPro( =J@(G 12.52259 ddPro( IVB+ 8.94408 ddPro( Ub@(_ 30.00000 ddPro( an@* 30.00000 ddPro( m z>(w 322.00000 ddPro(} y@+ 22.96092 ddPro(q B+ 0.52494 ddPro(e A( -0.63195 ddPro(Y @( 44.64368@ Y0J"; ddPro0P PY +A . Y%C^HH%CWE"                     ddPro1L ddPro( I `(L More MomentsddPro ddPro( e%+ 10.73333 ddPro( $i1+ 4.79176 ddPro( 0i=* 0.87485 ddPro( <eI(Fh 12.52259 ddPro( HiU+ 8.94408 ddPro( Tea(^h 30.00000 ddPro( `em* 30.00000Ln2x  {{{ 03 =,Times .+=, )shape )1), )scale )1( betaQuant)/()Z)Lm Lr"   {{{ 0: D,Times .+D, )shape )1), )scale )1( betaDensity)6()Z)Ld /4 ddPro HH" @# # # # # @ # # # # @ # #! #! #! @! #! #! #! #! #@! ! @#! #! #! #@! ! @#! #! #! #@! ! @#! #! #! #@! O NЇBARCDEF! N Nu"/ /! ! Yp X!4&  !& J^rP& #@& & @#& #& @& & @#& #& #& @& & @#& #& #'@''@#'#'#'@'"'#@''@'#'#'#@'* @* #* #* #@* * @* #* #* #@* * r@* Ы0a( 0s!(%IX(?)J. localoca. renleft . I-. lP. aringo. germandbls. logicalnot. guilsingllef. si circ. uperior/3M5>@O\yOlz>zB=3?wa3</: ` ` 7:AT8no01234567892:SUWKXZ\]<E77g`ListBoxBorde2Z7+[7 2\7@Bdrr,377 MMMM yyyy17uu.wD7@@@/@0@,@-22500105\`3{}5@G} }@5@J@2@3549 4mardimerc96xdd7@@U@S@.@1@X d6)7 !"#$%&'7ayFridaySatu92`hp5 y yl@@*@@&,.7bcdefghijklmno5n.sam.dim.janv5\`d@@)plopLpl& +nopqrstuvwxyz[+nmarmijue)*PX`hp# r y y yl@@*, >?@abcdefghijk, aySundayMon* * @&@'@(@@+@)@#&nopqrstuvwxyz[FrSarr@""# ddPro$@ ,  Helvetica .+0@ ""# ddPro$^ (1@ ""# ddPro$| {(2@ "a"# ddPro$ \j(e3@ "C"# ddPro$ >L(G4@ "%"# ddPro$ .()5@ ""# ddPro$ ( 6@  "#@!#@!"#@!"#@""# ddPro$5 )+ 0@ "7# ddPro$5 /?).1@ "L# ddPro$5 DT).2@ "`# ddPro$5 Xh).3@ "u# ddPro$5 m}).4@ "# ddPro$5{ ).5@ "# ddPro$5f ).6@ "# ddPro$5Q ).7@ "# ddPro$5= ).8@ "# ddPro$5( ).9@ "# ddPro$5 )1 ddPro$ o(r betaquant 1/10 ddPro$% H(KGraphs of Standard Beta pdfs ddPro( G+bM, Symbol (Q.addPro() = .5 ddPro((].bddPro() = .5 ddPro(l j+aaddPro() = 5 ddPro((bddPro() = 2 ddPro(2 Hc(RaddPro() = 2 ddPro((^bddPro() = .5 @  #@!#@!#@!#  {{{ 0) 3,Times .+3, )shape )1), )scale )1(betaDist)%()Z)L <  | | ddPro ddPro, C,Times .* gammaQuant{{{ |0D N ddPro, K U)N, ddPro, P r)shape ddPro, m {)1 ddPro, = I( @( ddPro, s |)6)L L."z`t  ddPro  }HH })                                                                                     #####$!!##%!!!!!!#@")# ddPro$k &,  Helvetica .+0.0@ ")# ddPro$x &(0.1@ "v)# ddPro$ q&(z0.2@ "i)# ddPro$ dr&(m0.3@ "[)# ddPro$ Vd&(_0.4@ "N)# ddPro$ IW&(R0.5@ "A)# ddPro$ <J&(E0.6@ "3)# ddPro$ .<&(70.7@ "&)# ddPro$ !/&(*0.8@ ")# ddPro$ !&(0.9@ " )# ddPro$ &(1.0@ ")# ddPro$` #0+0@ "Y# ddPro$` S`)04@ "# ddPro$`z )08@ "# ddPro$`L ).12@ "# ddPro$` )016@  #@!#@!#@!# ddPro$ 31@b(<4 shape = 1 ddPro$ hBus+5 shape = 3 ddPro$~ yf+$ shape = 5 1J1K ddPro$Q Q(Tgamma density quantile@  ) #@!#@!)#@! )#   {{{ 0I S,Times .+S, )shape )1( gammaDensity)E()6)L  s s {{{ s09 D,Times .+D , )shape )1( gammaDist)5():)L G G {{{ G09 C,Times .* normQuant)5())L L N N {{{ N0@ J,Times .* normDensit ddPro ddPro$@,  Helvetica .(NnormDens . i. i. i   ?@"4# ddPro$o 1+%C0.00@ "4#@"|4#@"t4#@"l4# ddPro$ ht1(q0.10@ "d4#@"\4#@"T4#@"K4# ddPro$ GS1(P0.20@ "C4#@";4#@"34#@"+4# ddPro$ '31(00.30@ "#4#@"4#@"4#@" 4# ddPro$ 1(0.40 1 5 8@"4#@##@"#@## 2HH 2 2 @##@ 4 #@!#@!4#@ 3 3#@"4# ddPro$d ,<+-3@ "?#@"J# ddPro$d BR)-2@ "T#@"_# ddPro$d Wg)-1@ "j#@"u# ddPro$d o|)0@ "#@"# ddPro$dy )1@ "#@"# ddPro$dc )2@ "#@"# ddPro$dN )3 ddPro$Q my(pX@  #@!#@!#@!#y)<())L  ? ? {{{ ?00 ;,Times .* normDist),())LL *p m  m!l#  l#  l# # #  ddPro # HH#F1g9F1F1g9F1F1g9F1F1"g9g9F1F1"g9g9F1F1"g9g9F1F1#g9g9g9F1F1"g9g9F1F1"g9g9F1F1"g9g9F1F1g9F1F1g9F1F1F1F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1 F1"g9F1F1"g9F1F1"g9F1F1(g9g9F1F1(g9g9F1F1(g9g9F1F1)g9g9g9F1F1(g9g9F1F1(g9g9F1F1(g9g9F1F1"g9F1F1"g9F1F1F1F1g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9F1"g9F1F1"g9F1F1"g9F1F1"g9F1F1"g9F1F1"g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1F1F1F1g9F1F1g9F1F1g9F1F1!g9g9F1F1!g9g9F1F1!g9g9F1F1"g9g9g9F1F1!g9g9F1F1!g9g9F1F1!g9g9F1F1g9F1F1g9F1F1F1F1  ddPro( },  Helvetica .+ Normal Distribution ddPro(* Normal Density ddPro(* Normal Quantile (probit) ddPro(* Chi-Square Distribution ddPro(* Chi-SquareQuantile ddPro(* Student's t Distribution ddPro(* Student's t Quantile ddPro(* F Distribution ddPro(* F Quantile ddPro( sh(v ddPro(* Gamma Distribution ddPro(* Gamma Density ddPro(* Gamma Quantile ddPro(* Beta Distribution ddPro(* Beta Density ddPro(* Beta Quantile#Ln2x(x x( (x#',Times .+seed, Symbol) +{( ^=( [i)1( Rif ( "1234567)#,+ otherwise("seed),(#results )?uniformL` ` ddProccc 0Lxx` ddProccc 0Lxx` ddProccc 0Lxx` ddProccc 0Lxx` ddProccc 0Lxx` ddProccc 0Lxx` ddProccc 0Lxx` ddProccc 0LxxL] \ m  m!l#0 ddPro f|HHfuet F1g9F1F1g9F1F1g9F1F1 g9g9F1F1 g9g9F1F1 g9g9F1F1!g9g9g9F1F1 g9g9F1F1 g9g9F1F1 g9g9F1F1g9F1F1g9F1F1F1F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1F1F1g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9F1F1g9F1F1g9F1F1g9F1F1 g9g9F1F1 g9g9F1F1 g9g9F1F1!g9g9g9F1F1 g9g9F1F1 g9g9F1F1 g9g9F1F1g9F1F1g9F1F1F1F1  ddPro( cR,  Helvetica .+ Uniform ddPro(* Normal ddPro(* Exponential ddPro(* Cauchy ddPro(* Gamma ddPro(* Triamgular ddPro(* Shuffle ddPro(* Poisson f}HHfvv F1g9F1F1g9F1F1g9F1F1 g9g9F1F1 g9g9F1F1 g9g9F1F1!g9g9g9F1F1 g9g9F1F1 g9g9F1F1 g9g9F1F1g9F1F1g9F1F1F1F1g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9g9F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1g9F1F1F1F1F1g9F1F1g9F1F1g9F1F1 g9g9F1F1 g9g9F1F1 g9g9F1F1!g9g9g9F1F1 g9g9F1F1 g9g9F1F1 g9g9F1F1g9F1F1g9F1F1F1F1  ddPro( wO('z Binomial ddPro(* Geometric ddPro(* Negative Binomial ddPro$ * Random Number Seed  l#  l# # # #Ln2xTa addPro a 1a@ 8#*3-<BHH3-<;S\      *3-<BHH3-<;!*       ddPro( 3],  Helvetica .@+= Shadow Frame ddPro(@*Rectangle Frame ddPro(@*No FrameddPro(@). @ a UUUU 00 # ddPro( ;( Yellow Background ddPro(*Matching Background ddPro(*White Background L^ m  m!l#  l#  l# # # #p2    n .  & -l'l'l' & 'Lp2xm  m!l#  l#  l# # # #p2    n .  & -l'l'l' & 'Lp2xm  m!l#  l#  l# # # #p2    n .  & -l'l'l' & 'Lp2xm  m!l#  l#  l# # # #p2    n .  & -l'l'l' & 'Lp2xm  m!l#   l#  l# # # #bin210173tp2    n .  & -l'l'l' & 'Lp2xNNN2 22pxؼD@>2NeNWhats New in JMP Version 3.2 February 1997 This document gives a detailed description with examples of all changes and enhancements that have occurred after JMP version 3.1. It assumes you have access to the manuals that came with the full JMP software     package. The following three manuals are included with the full JMP package: The JMP Introductory Guide is a collection of tutorials designed to help you learn JMP strategies. The JMP tutorials range from single-step procedures to complex analyses. You can read the tutorials for reference, or work through them step by- step. Each tutorial uses a file from the sample data folder. By following these examples, you can quickly become familiar with JMP menus, graphical displays, options, and report windows. The JMP Users Guide has complete documentation of all JMP menus, an explanation of data table manipulation, and a description of the calculator. There are chapters that show how to do common tasks such as manipulating files, transforming data table columns, and cutting and pasting JMP data, statistical text reports, and graphical displays The JMP Statistics and Graphics Guide documents the statistical platforms, discusses statistical methods, and describes all report windows and options. JMP Version 3.2 Copyright 1997 by SAS Institute Inc., Cary, NC, USA All rights reserved. Printed in the United States of America. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, or otherwise, without prior written permission of the publisher, SAS Institute Inc. Information in this document is subject to change without notice. The software described in this document is furnished under the license agreement printed on the envelope that contains the software diskettes. The software may be used or copied only in accordance with the terms of the agreement. It is against the law to copy the software on any medium except as specifically allowed in the license agreement. First printing, November 1996 JMP, JMP Serve, and SAS, are registered trademarks of SAS Institute Inc. All trademarks above are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. indicates USA registration. Contents  TOC Welcome to JMP 1 System Hardware and Software Requirements 2 Windows 2 Macintosh 2 Before you Install 3 Register Your Product 3 Installing JMP 4 Windows Installation 4 Macintosh Installation 4 Customer Support 4 Sales Support 4 Training and Education 4 Technical Support 4 JMP Information on the World Wide Web 5 Overview of Enhancements in JMP Version 3.2 6 General 6 Analysis Windows 6 Data Tables - The Tables menu 6 Design of Experiments 7 Calculator 7 Analyze Menu - Distribution of Y: Continuous Variables 8 Analyze Menu - Distribution of Y: Nominal or Ordinal Variables 8 Fit Y by X - general 8 Fit Y by X - Continuous by Continuous 8 Fit Y by X - Continuous by Nominal 9 Fit Y by X - Nominal by Nominal 9 Analyze Menu - Fit Model Dialog 9 Analyze Menu - Fit Model : Standard Least Squares 9 Analyze Menu - Fit Model: Effect Screening 9 Analyze Menu - Fit Nonlinear 11 Analyze Menu - Correlation of Ys 11 Analyze Menu - Cluster 11 Graph Menu - Control Charts 11 Tools Menu 12 Windows Menu 12 1?/5|e#Corporate Microcomputing Department>\\JMPNT\source\Install\Jmp3\WinOS\MASTERS\JMP\OTHER\JMP322.DOC@chpljs32Ne06:winspoolchpljs32chpljs32W odXLetter0EPRIV''''