The Graphical Interface
From Minitab Like Sygraph, this package is a hybrid of menu- and programming-driven software a drawback for users that have to switch back and forth between these two worlds. UNIQUE GRAPHICS FEATURES: Minitab has two graphical features that were not found in other software packages. First, Minitab has an option to display the confidence intervals on sigma when conducting homogeneity of variance tests. It also has a heterogeneity of regression graphics display that accompanies its analysis of covariance procedures. STRENGTHS: Minitab has very good univariate descriptive graphics and very good diagnostic features associated with the standard statistical procedures, such regression analyses, analysis of variance (ANOVA), time series, logistic regression and correspondence analysis. This package also has especially nice displays of the effects of factorial designs and a heavy accent on graphical display of results from quality-control analyses. WEAKNESSES: The three-dimensional graphics in Minitab are limited and the graphical possibilities are not very creative. Also, the manual was filled with little errors in the text. The graphics displayed in the manual are hard to see and not particularly appealing to the eye. Furthermore, the style of presentation of the commands in the programming mode (not menu mode) is a strain to understand, unless the user has previous experience with Minitab or has the desire to really sit down and learn the programming commands. Minitab also presents an option for 'condensed plots' in the section entitled Exploratory Data Analysis. These seemed interesting, but it was not possible to understand what these plots were as there was no accompanying visual display or description. Minitab is not designed for those scientists interested in multivariate procedures or imaginative multivariate representations of data. LIKELY USERS: This package appears to be good for those scientists that do straightfoward graphical analysis of quality-control, time series, ANOVA, regression and factorial experimental designs.
|
|||