MINITAB(ID:723/min018)

Interactive statistical programming 

Interactive solution of small statistical problems.

"MINITAB is an interactive statistical computing system [...] developed by the Department of Statistics at Pennsylvania State University. MINITAB is written in FORTRAN and originated as a student-oriented adaptation of the National Bureau of Standards OMNITAB system. It is designed to allow students in introductory statistics courses to communicate with the computer through commands similar to English sentences."
Related languages
OMNITAB II => MINITAB   Influence
MINITAB => MINITAB II   Evolution of
References:
  • Ryan,T.A. et al, "MINITAB Student Handbook", Duxbury Press 1976. view details
  • Farber, Elizabeth "Guide To Minitab" NY 79 view details
  • Velleman, Paul F. and David C. Hoaglin. "Applications, Basics, and Computing of Exploratory Data Analysis" Duxbury Press, 1981 view details
  • Croarkin, M. Carroll "Statistics and Measurements" J. Res. Natl. Inst. Stand. Technol 106(1) January?February 2001 pp279?292 view details Extract: Statistical Computing
    Statistical Computing
    The ubiquitous use of statistics at NIST has come about for many reasons, one of which is certainly the development of state-of-the-art statistical computing tools within SED. In the early 1960s, Joseph Hilsenrath of the Thermodynamics Section, Heat and Power Division, conceived the idea of a spreadsheet program for scientific calculations. Together with Joseph Cameron and the support of several NBS sections, this idea led to a program called Omnitab [5]. Omnitab is an interpretive computing system with a command structure in English that performs scientific calculations on data in columns in a worksheet.
    When Cameron became Chief of SEL, he formed a team, headed by David Hogben, to complete the development of Omnitab as a sophisticated statistical package. By 1966, it was already strong in data manipulation, regression analysis with related diagnostic graphics and tests, one and two-way analysis of variance, special functions, and matrix operations. It quickly became the standard tool for statistical calculations at NIST. It was so innovative at the time that when Brian Joiner left SEL in the 1960s to teach at Pennsylvania State University, he took a copy of Omnitab with him for his students. A few years later, Joiner formed a company that revised the code and offered it for sale as the commercial package, Minitab.
    Omnitab is strong on analytical procedures but not on graphics output. In 1969, when James Filliben brought his perspective on exploratory data analysis (EDA) to NBS, he immediately saw the need for software with strong graphics capability, and he set about developing code to support his consulting activities that incorporated the best features of EDA. There was never a steering committee for this project as there was for Omnitab, but from the breadth of problems and data encountered in the NBS laboratories, a diverse and versatile package, called Dataplot [14], was conceived. The package is a workhorse for graphical and statistical analysis at NIST and is a repository for datasets from important NIST experiments. Because it is a free and down-loadable resource maintained by the Information Technology Laboratory, Dataplot has recently been interfaced with an on-line statistics handbook that is under development within the Statistical Engineering Division and SEMATECH. From the handbook pages, the reader can run examples of statistical approaches presented in case studies in the handbook.

    Resources
    • MINITAB - Interactive Statistics
      external link
    • Minitab: What is it? by Bob Hayden
      Minitab is software for statistical number crunching that has been around for many years. It had its roots in the Omnitab package developed at the National Bureau of Standards around 1970.  Omnitab was an attempt to create a program that was more user-friendly than the other mainframe packages of its day. This ease of use inspired teachers at Penn State to adapt Omnitab for an introductory statistics course there, thus creating Minitab in 1972. Early versions of Minitab were designed to run on many different platforms, and this, combined with its ease of use, led to its rapidly becoming the de facto standard statistical package for use in teaching statistics. This early use in education influenced the development of Minitab. For example, it incorporated simulation tools that helped students grasp things like sampling distributions.


      Originally Minitab did fewer statistical procedures than other mainframe packages. The next stage in its development was the addition of novel features that were not included in other packages. Perhaps the first important example was the inclusion of the exploratory data analysis tools developed by John Tukey and his colleagues. These tools were communicated to a wider audience by Paul Velleman and David Hoaglin in their book _ABCs of EDA_. The book not only explained the new techniques, such as boxplots and stem and leaf plots, but included computer code for implementing them. This code was incorporated into Minitab, making it one of the first packages to provide these tools. Later Minitab added extensive quality control tools. In this era Minitab was a statistical Swiss Army knife, providing quick analysis and exploration of moderate-sized datasets. It became popular in industry, although some academics still looked down on it as a package for beginners.


      In recent years, Minitab has added many of the traditional statistical techniques included in larger stats packages, and concentrated more on sales to business and industry. The last three versions have run only in Microsoft Windows, although earlier versions are still available for other platforms.


      What is it NOT?


      Minitab is a tool for data analysis with an emphasis on ease of use. Although it incorporates some features to aid instruction, it is not really a tutorial package. You can use it to learn statistics by doing statistics, but it does not teach you statistics by itself the way, say, ActivStats does

      external link