Statistics: Software packages for teaching

From MathWiki

Software packages for teaching

Package or Language Pros Cons
Excel
  • widely available
  • support available
  • menu interfaces for statistics
  • very limited range in statistics
  • little developer commitment to statistics
  • many functions outdated
  • easy to get what Excel does but hard to get what you want

Minitab

  • Excellent package for teaching: relatively simple and easy to use
  • Good materials available for students
  • Good quality but functions are limited to intro level statistics

SPSS

  • Simple analyses are easy to, advanced analyses are likely to be impossible
  • Widely used by Psychologists, Sociologists, etc.
  • Easy but limited graphical interface
  • commercial use not cheap
  • limited advanced functions
  • new methods are added in about 10 years
  • need to buy extra modules to do many things

Stata

  • used widely by econometricians
  • excellent functions well developed for economics
  • has large active research community
  • fast even on large data sets
  • graphics easy but not exciting
  • interface uses code

SAS

  • widely used by large corporations for data management as well as analysis
  • top quality highly engineered procedures
  • implements new methodologies within 5 years
  • the ‘standard’ in many areas: e.g. pharmaceutical research
  • Founder and CEO, James Goodnight is worth $4.1G
  • Very expensive
  • Interface uses code
Excel
  • widely available
  • support available
  • menu interfaces for statistics
  • very limited range in statistics
  • little developer commitment to statistics
  • many functions outdated
  • easy to get what Excel does but hard to get what you want

Minitab

  • Excellent package for teaching: relatively simple and easy to use
  • Good materials available for students
  • Good quality but advanced functionality is limited in scope

SPSS

  • Simple analyses are easy, advanced analyses are often impossible
  • Simple analyses are easy, advanced analyses are likely to be impossible
  • Widely used by Psychologists, Sociologists, etc.
  • Easy but limited graphical interface
  • commercial use not cheap
  • limited advanced functions
  • new methods are added in about 10 years
  • need to buy extra modules to do many things

Stata

  • used widely by econometricians
  • excellent functions well developed for economics
  • has large active research community
  • fast even on large data sets
  • graphics easy but not exciting
  • interface uses code

R

  • Simple analyses can be challenging but everything's possible
  • much new statistical research prototyped and disseminated within months
  • excellent exploratory graphics
  • very good presentation graphics
  • has large active user community including many active applied statisticians. The number of user-contributed packages has grown from 500 to 1,000 in the last two years.
  • extremely powerful if you’re into programming
  • easy to link with C/Fortran routines
  • free, gets some core funding from EU
  • open source with committed group of core developers – likely to continue on its trajectory for a while
  • interface uses programming code -- challenging to learn