nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Could JMP improve the online manual?

Fulcrum

I do quite like JMP, compared to other statistical software. I fell that what holds it back, to a significant extent, is the lack of dexcription it provides in the online help manual. The hands-on experience is good, IMO, but how to understand and interpret the output is poor. The expamples in the online book aren't always helpful. If the idea of this software is to bring powerful statistical analyses to the masses and those learning with great GUI software (which this really is), I think it would be great to enhance the manual with details that explain everything. It would really help make this software (i.e. all the outputs) so much easier to use and unbeatble. This is my suggestion after being a long term fan. Thank you.

Accepted answers

Mark_Bailey

There is no need to apologize. It is not about lack of awareness or lack of good intentions. It is all about limited resources and balance among the myriad efforts, as I said before. Well-documented tools of any kind are always welcome. Software documentation, though, is generally not designed to teach. That is the role for training courses and textbooks.

Of course, there is a lot to learn. You cannot learn much in a 10-minute sound bite, a 15-minute video or two pages in documentation. You cannot spend 14 hours in a short course and become a PhD organic chemist.

The discussion area is meant to be helpful. Please continue to use it. Return with your specific questions, not just about "how to" in JMP, but also "why this or why not that?"

All comments

txnelson

Could you provide more specifics? If you could take a specific output, select the ? tool and then click on the area of the output you want to use as your example. Then provide your feedback as to what additional information you would like to see on the displayed help file. I think that would be very helpful in understanding exactly what you are looking for.

Mark_Bailey

A lot of effort has been spent on continually improving the documentation. Your comments are welcome. Remember, too, that many different users have different expectations of software documentation. @txnelson's suggestion is good. Please let us know when you find an unsatisfactory or a missing entry in the documentation. There is a total of 6626 pages in the complete PDF file, so focusing the effort is more likely to achieve the goal, albeit it one topic at a time.

Please do not post such findings here in this thread or new ones. Add them to the Wish List on the JMP Community site that is provided to collect ideas about improvements.

statman

Your suggestion for continuous improvement is, as Mark says, welcome, although I don't have anything to do with it (nor do I have any knowledge of what work might be going on to this end).

If I understand your suggestion correctly, you would like to have the manual "translated" into a "language" with examples that make sense to you.

Communicating how to "interpret" results that are statistical in nature to users that may have little to no background in statistics in not a trivial effort (even to those with a statistical background which inevitably is biased to their educational/professional experience). Additionally, many (perhaps all) of the statistical platforms are developed by statisticians and therefore are written in terms they understand. It is a challenge to explain difficult concepts, balancing practicality/usefulness with being "technically correct" and this is compounded by there is no one right way to do analysis in any given situation (certainly not one method that all statisticians agree to).

It would be interesting to know if the folks designing/writing the manuals are experimenting (or executing sampling plans) to assess the robustness of the manual to users...

Fulcrum

Hi txnelson, thank you for offering to help out. There's nothing really specific and I can usually figure things out, but it would be really good if the manual provided more descriptions on what the outputs mean with a with a lot of the statistical tests. An example would be random forest plots and the output you get. The manual just provides examples with a bit of interpretation. I wish I was a statistician but I am not. JMP makes statistics really easy to perform and I'd just like to learn what complex outputs mean from guidance in the manual so that I can apply it.

I have a good example of what I am not explaining very well here with StatsDirect softwars. It's not exactly as sophisticated or specialised as JMP, but it does make the software really self-contained. I haven't found any of the books written for JMP to be really helpful in providing detailed explanations behind the outputs for the statistics.

Fulcrum

Thank you, Mark. My apologies - I know it's a really big task.

Fulcrum

Thank you. Yes, I understand, it is specialised. The software is so incredibly easy to use that you get novices (such as myself!) that use it and find themselves suddenly able to perform really powerful statistics. It's like the cart-before-the-horse for me.

Mark_Bailey

The discussion area is meant to be helpful. Please continue to use it. Return with your specific questions, not just about "how to" in JMP, but also "why this or why not that?"

louv

My two cents....Manuals are very useful and helpful but not a replacement for a statistical eye. At my company I was supported by statisticians who could be consulted with. Unfortunately not all folks have this luxury and a shout out that organizations need to invest in statisticians.

P_Bartell

I'll add my two cents to all that @louv , @Mark_Bailey , @txnelson and @statman have contributed. SAS and JMP offer a very rich set of resources and content such as textbooks, online videos, courses, webinars, that JMP users can access, many free of charge. Very often, the specific content is structured with two end goals in mind; 1. Demonstrate 'How to in JMP' AND 2. Interpretation of JMP output, tips, tricks, and things to consider. So my recommendation is if you find the documentation lacking...search the JMP web site for content centered on your specific problem, data, and analytics/visualization techniques and chances are you'll get that explanation in either statisticallese or more plain English.

During my tenure with the JMP division at SAS, I often worked either solo or with my colleagues to create this type of content. Our goal was to always cover the 'How to in JMP...?" as well as the interpretation and meaning of specific JMP output. I used to say, "The chart is just a chart if you don't understand what it's telling you."