There is no 'best' statistical tool. It depends on what you need and how you want to use a program. It also depends on the needs for data-management. I use R, SPSS, STATA and SAS, but also some specific programs like LISREL or MSP5. An advantage of R is that many people contribute to it and it has a lot of methods and statistics other programs do not have. In R you can do GLM, GLMM, network-analyses, text-analyses, multilevel analyses etc. Some of these are not available in for instance a standard version of SPSS. I use an intergration of SPSS and R to optimize the things I can do. Data-management is easier in SPSS then in R. Especially when you need to aggregate data. A 'good' statistician knows several programs and knows when to use which to optimize his/her performance.
Eviews, STATA and SPSS are all good softwares, It depends on the data and kind of results you are looking for as there are no much difference in these statistical packages however there are several unique features available in one software and missing in other. For instance there is no much difference in estimates of regression coefficients in Eviews and Stata but in Stata it gives 95% confidence interval for each coefficient. Similarly AOV tables are missing in Eviews but available in other packages. Hope it satisfies your query.
STATA is best but it depends on your mind: I mean STATA is mostly used by codes & commands and very user friendly but can also be used through menu items. same is for E-views but its use is not popular through codes and commands. SPSS is also good but as the friends stated above, it depends on the nature and kind of Data as well. One thing more R is best among all but for that you must be knowing little about programming and has very wider scope than all of the above software's. It covers about all short comings of the above software's.
The differences between programs are not as large as has been suggested. All programs have menu's, but those for R you have to install (Rcmdr... programs), but when you want to do something special you have to create some kind of syntax the program can run (would not call that programming). In all programs you can also do some programming. SPSS has a programming language (SAXbasics) that looks a lot like VBA, but is not used by many people. It really comes down to what you prefer. But for some very specific types of analysis you might not have much choice and have to use a specific program. In many cases you will learn to prefer one program above other as you will work with it most of the time. I have worked with SPSS since the mid '80 and have gained some experience in using it productively. But sometimes I still miss the program GLIM4.
- Badges
- Science topic