You might try Rcommander . A former student of mine thinks it's a great. Package and his students seem to think so too. Best wishes, David Booth

Adriano Marcos Rodrigues Figueiredo I also prefer coding on RStudio to learn/do statistical analysis and modeling. Reproducible coding gives me opportunity for hands on learning experience.

If you understand what you want to do, you can spell it out. Learning to speak is also learning to think. This goes togehter. There is no simple point-and-click shurtcut to this road. Common, repetitive tasks can easily be automated, and even this is much simpler to achieve with a script that using a protocol explaining what to click when and which radiobuttons and checkboxes to select.

It is also an important state of thinking to keep data and analysis separate. Although this separation can be achieved practically also in point-and-click GUIs, mixing this (even when it's just looking like this for the user) destructs this kind of thinking and actually compromises good scientific practice.

I think if one really wants to do or learn (empirical) science in a professional way, there is no way around some kind of coding. Some tasks may eventually (in practical work) be solved with any tool, including point-and-click or some stats-made-easy software (what is absolutely ok for that purpose). But a serious empirical scientists must be able to understand and to formulate the concept of data analysis, and there is really not any big step to use some kind of programming / scripting / statistical language for this.

In my experience, I face a lot of (young and senior!) scientists who never really learned to look and to work with their data, and they largely don't understand their data. They use inappropriate models to analyze their data, test wrong or simply uninteresting statistical hypotheses within their models and they often over- or misinterpret the results of their analyses. There is a huge lack of understanding, and to my opinion this is in part due to the fact that the difficult task of learning to think about data is largely avoided by producing quick results after clicking a few menu items and dialog boxes. They learn using a hammer, and after this everythink seems to look like a nail... This is surely not the only reason for the problems, and learning a programming/scripting/statistical language is not a general remedy, but it is one puzzle piece in this game, that should not be disregarded just because it seems more difficult. The great difficulty lies in "learning to think", and bypassing this means to eventually produce mediocre scientists. This should not be our aim.

When our Statistics Department switched its social science track to use the SPSS GUI (they use JMP now), there was much worry about students wildly poking at menu choices that were poorly understood. What we saw in the Statistical Consulting Center that I managed was that the GUI users had more time to learn to think about statistics. I commonly heard students that "learned" code say that they spent so much of their time debugging it that they had little time to focus on what it all meant. As a result, I became a fan of GUIs for people who don't have the time to learn to code. I review all the GUIs for R here: https://r4stats.com/articles/software-reviews/r-gui-comparison/