…assign the Name variable as Label, and use any combination of suitable variables for any of the Analyze and Graph menus.
• CAR POLL (crosstabs)
…categorical variables for generating crosstabs with Fit Y by X
• Cereal (general)
…collection of nutritional information on cereals. Explore to find most nutritious cereals (Distribution of Y’s) ; see if manufacturers are different using simple analysis of variance (Fit Y by X); try to discriminate among manufactures using the Fit Model command with multiple numeric Y variables.
• CLIPS1 (control chart)
…example for cusum Control Chart; experiment with other types of control charts.
• COATING (control Chart)
…example for Mean, Standard Deviation, and Range Control Charts; experiment with other types of control charts.
• COWBOY HAT (Spin)
…simple example of a spinning plot; ‘x ’and ‘y’ are coordinates an ‘z’ is a functional response; experiment with Fit Y by X -plot y by x and see the cowboy hat contour colored by the z variable.
• DOGS (repeated measures)
…used in long AppleScript demo (see AppleScript folder); repeated measurements are taken on dogs in 2 treatement groups-use Fit Model and specify the logs of the measures as Y variables and drug as the effect to do a repeated measures analysis.
• FAILURE3 (Pareto)
…experiment with the Pareto Chart platform-‘failure’ is the Y variable: for a single pareto chart, use only the first 7 rows (either take a subset of the rows or exclude the remaining rows from the analysis); use the first 14 rows and a single grouping variable, or all rows and two grouping variables.
• FITNESS (multiple regr)
…use Fit Model with ‘Oxy’ (oxygen uptake) as Y and any of the other variables as effects; try both Standard Least Squares and Stepwise as fitting personalities.
• GOLF BALLS (multivar)
…use Fit Model with ‘Distance’ and ‘Durability’ as Y variables and ‘Brand’ as a single effect. Choose the Multivariate fitting personality; experiment with saving discriminant scores and look at how the multivariate analysis of variance model classified the brands.
• INGOTS (logistic)
…use Fit Y by X with ‘ready’ as Y and ‘heat’ as X to see a logistic probability plot; experiment with inverse prediction (check menu option); use Fit Model if you want a model that includes both ‘heat’ and ‘soak’ as effects.
• LITTLE POND (contour)
…use the Contour Plot command to see pond depths at x, y coordinates; experiment with contour options such as fill, label and so on.
• ODOR CONTROL (surface)
…Use Fit Model and specify ‘Odor’ is Y and the other variables as effects; select response surface from the effect macro popup menu; open all reveal tables to see the response surface solutions. Request grid and see surface movie.
• POGO JUMPS (ternary plot)
…use the Ternary Plot command in the Graph menu and specify ‘Yat’ ‘Yee’ and ‘Sam’ as Y variables; after you see the ternary plot you can select Fill areas and specify ‘Total’ as the response; use the crosshair tool to read the coordinates of the points; be sure to experiment with the magnifying glass tool.
• POPCORN (factorial)
…use Fit Model with ‘yield’ as response, the other variables (except trial) as factors and select Factorial from the effects macro popup menu. Run again with Factorial to Degree (2) and compare. Note the Profile plot of interaction between ‘POPCORN’ and ‘BATCH.’
• RATS-DMBA (survival)
…used in long AppleScript demo (see AppleScript folder); Use Survival command with Days as time, Censor as censor, Group as group; or use Fit Model for proportional hazard model. Days is Y, Censor is censor, and Group-int is the grouping variable (Effect in Model)
• REACTOR (screening)
…used in long AppleScript demo (see AppleScript folder); use Fit Model and specify two-level full factorial, with Screening personality
• SOLUBILITY (outliers)
…use Correlation of Y’s with all the solvent variables as Y variables; display the scatterplot matrix to look for bivariate outliers (extreme points); use the Spinning Plot command in the Graph menu and choose as the solvents as variables; use the hand tool to rotate the plot and experiment with different combinations of three solvents. Label and color the outliers.
• SPRING (plots and charts)
…use the Bar/Pie chart and the Overlay Plot commands; use ‘date’ as the X variable for Bar/Pie and ‘day’ as X for Overlay (X must be a categorical variable for Bar/Pie and numeric for Overlay); select ‘Humid:1PM’ and ‘Humid:4:00’ PM as Y variables; play with whole window options (check mark popup) and individual chart options (popup menus showing at the bottom of the chart.
• TIRETREAD (screening)
…use Fit Model to do a multivariate screening analysis. Select ‘ABRASION’ ‘MODULUS’ ELONG’ AND HARDNESS’ as Y variables, and the remaining variables as factors. Choose the Screening personality from the fitting popup menu. Experiment with the prediction profile and its desirability profile, and the other screening options.
• US POP (nlin)
…use the Nonlinear Fit command to see a simple nonlinear fit of exponential distribution. Compare nonlinear fit with plain linear fit or polynomial fit using the Fit Y by X command.
Note-US POP is a stationary file because it contains formulas with parameters that have starting values. When you open a stationary file, the original file is left intact.
• VA cancer (survival)
…use the Survival command to see life table survival estimates. select Time as the response, Censor as censor and either type or treatment as a grouping variable.
…use Fit Model to see proportional hazard survival estimates. You can have more than one effect in the model.
… use Nonlinear Fit to fit a parametric time failure model with any one of the loss functions provided in the table.
Note-VA cancer is a stationary file because it contains formulas with parameters that have starting values. When you open a stationary file, the original file is left intact.
• XYZ Stock Avges (Plots)
…use Overlay plot command to show Date on the x-axis and High, Low, Close as Y variables. Select the range option and turn off the connect points option for all Y variables.
…use Overlay plot with Date as X and both XYZ (stock prices) and moving Avg. as Y variables to compare them. The moving Avg column has a formula that shows you how to compute moving averages.
