I used SPSS and Eviews for statistical analysis in my research articles. But I found variety of analysis and graphs in many articles in reputed journals, which may not be created by the earlier mentioned softwares, or I may not learned to create it.
As far as I know, the editors of reputed journals look for 'what's new' in the article that they receive. Specifically, they look at the research question in the article, the methodology adopted to answer the question and the extent of robustness checks done.
As far as I know, the authors are required to inform the editors which software they used.
Please do read this interesting presentation on publishing.
There is no specific software which finds favours with editors of reputed journals. Your choice of software depends on your research objectives & most of the reputed research software available are reasonably good for a variety of analysis. Output is also likely to be similar with minor differences in presentation. User Interface may also differ slightly.
The acceptance of your paper does in no way depend upon your graphs, although they enhance readability. A properly developed research focus,the clarity of your research questions, appropriateness of analysis tools & robustness of results alongwith suitability for the journal (subjectwise or sometimes regionwise) for that article will ensure acceptance of your paper. I've published in & am associated with a few international journal groups as reviewer & editorial board member & am speaking from my own experience, not hearsay.
All Softwares give the same results for the same dataset in the specific analysis, so there is no preferred statistical software to be used for the editor.
Like my senior colleagues have already said, the clarity and dependability of your research results are not determined by the software you use, rather on data analysis techniques that you have employed and other aspects of the work. However, it is not wrong to learn how to use different data analysis packages. It makes you more versatile in angina your data. So if you have opportunity, learn how to use STATA or Epi-data, etc. alongside SPSS - nothing wrong!
The matter seems interesting to me. One thing (though no one touched) we the developing country authors should know the language that impact journals want to see in your write up. Secondly, using statistical analysis is not a matter as almost all mentioned, I also go with them. sometime we miss very simple thing before choosing the appropriate statistical software, such as SPSS and SEM. SPSS assume data normality where as SEM can deal with not-normal data. Even among SEM, variance-based (PLS is useful) and co-variance-based (AMOS is useful) framework will help you to determine the appropriate software. So, the assumption will help you to choose between the two stream software. Thirdly, as Hemlata mentioned "the clarity of your research questions, appropriateness of analysis", you need to bridge well between "research question" and "robust tool" with justification that will eventually address the objectives of your article. Finally, you must bring something new to either literature, methodology, or adding to theory (or modification), otherwise you can not stand your paper before the paper editor/reviewer to get into published in reputed outlet. One thing I also like to disclose (though rarely talked among academic elites), include few papers from the journal you target to publish, it really matter.
Hope these would be of use for an author from developing country.
Several of the responses say the software does not matter. This depends, so I disagree with those. Statistical and data visualization software do different things. Even for simple analyses, sometimes the defaults are different (e.g., some run a t-test assuming the variances are the same, some do not make this assumption). Even something as seemingly straightforward as quantiles will come out differently depending on what software you use (see http://127.0.0.1:30379/library/stats/html/quantile.html). For more complex analysis procedures often use different methods, so it is important to examine what each is doing, and of course some software does not have that many procedures available. While many of the data visualization packages allow you to mimic what is done in others, if they are using statistical data (e.g., confidence or credibility intervals), these may be calculated differently in these packages. This is why it is recommended to cite which packages/languages that you use (and for some projects include the code you use).
Most journals (there are journals specific to some packages, like Stata) should not worry about the package that you use so long as you say what you did and justify it. If you say you did X because the package could not do Y, the editor should likely tell you to use a package that can do Y (if one exists), but if X is what you want to do and say it can be done in Excel then editors should be alright with this.