Means Diamond on Oneway Analysis - Does not respect preferences

[#1]

The oneway analysis platform does not respect the preference setting of means diamonds.  I have deselected it in preferences and it always is presented, requiring me to turn it off.

I had previously reported this bug and it must by close to 20 years ago.  This bug is now old enough to vote. It is still in JMP 16 and technical support says yes we know about it and basically "we have no intention of fixing it."

If you guys want me to keep paying for your product you might want to fix some of the bugs.

Andy

[#2]

I suggest that you present your sentiments directly to the support folks at
support@jmp.com
and to the JMP Wish List
Expressing it to the Discussion Community informs the JMP User community, but doesn't guarantee that it gets to the decision makers from JMP.

[#3]

I have. Given that it is a long know outstanding bug I think there is no interest in fixing it. So I am trying some public shaming.

[#4]

The preference here is working as designed. The preferences available correspond to the options available in the red triangle menu and control which of those options is on as the platform is initially launched.

Some of menu options activate more than one other option, as is the case here for the Means/Anova option. It turns on the Means Diamonds and JMP provides no option to specify otherwise.

 

I have recorded your request to try to provide a preference to not display the means diamonds with the Means/Anova option. 

[#5]

The purpose and meaning of the preference that you tried is if you want the mean diamond graphic automatically added to the plot on its own or not. It is a separate command in the context menu (red triangle). Independent of the graphic control and its preference, using the ANOVA command will add this graphic, too, along with several tables of numerical results.

The same pattern of behavior is exhibited by the horizontal lines (graphics) that represent one standard deviation above and below the group mean. There is a separate command for the graphic but it is automatically added by both the Means and Standard Deviations command and the Unequal Variances command.

[#6]

My understanding in preferences is that it allows me to set the graphic elements that I would like by default to be included.

To that end I have made selections including Jitter on and turned others off such grand mean line.

When I do a oneway analysis the points are jittered as requested, but the means diamonds are INCLUDED. I then have to turn them off every time by going to the red triangle and going to display options and deselecting means diamonds.

Maybe I am confused about the meaning of preferences and I assumed that this would control the default display when using that platform.

[#7]

Do you see the mean diamonds when you launch Oneway? That behavior is what the preference controls. If you enable it, then you see the graphic upon launch. If you disable it, then you do not see the graphic upon launch. That preference has nothing to do with the ANOVA command and its results.

[#8]

I do see it when I launch oneway. That is crux of my issue. I have set a preference for oneway analysis and when with means diamonds deselected I still get it as is show in my screenshot. I had first report this bug to JMP to the best of my recollection in about 2003.

[#9]

---

@Hegedus1 wrote:
I do see it when I launch oneway.

---

We may need a clarification of terms here. @Mark_Bailey may have been imprecise when he asked, "Do you see the mean diamonds when you launch Oneway?" I suspect he mean when you launch Fit Y by X with a categorical X and a continuous Y.

I think you mean when you choose Means/Anova from the red triangle menu in the Oneway platform.

The preference controls what is displayed when you launch Fit Y by X, not when you choose Means/Anova.

[#10]

What I mean is exactly launch Fit Y by X and I get means diamonds. The first display of the graph without selecting any means/anova tests from the disclosure triangle. I think what is happening if means is also selected in preferences then the means diamonds preference is ignored. You can further go to the disclosure triangle and deselect means diamonds, but since I have that set in my preferences I would most kindly request that I can have BOTH means selected and means diamonds deselected and get that result as the preferred graph when I do Fit Y by X with a categorical X and continuous Y.