Per @markbailey 's advice, I suggest taking a look at the JMP online documentation for Contingency Table analysis within the Fit Y by X platform. There are examples for you to follow.   Contingency Tables in Fit Y by X Platform  ... View more

It took awhile...but during my tenure in industry...I was always trusting of the plan we articulated for the execution of a designed experiment was, in fact, what happened during execution, and to vehemently agree with @statman 's earlier reply, measurement of responses. Things like, randomization, blocking, did they really set the factors at the prescribed levels, any noticeable noise variables we didn't anticipate??? and on and on. Only after discovering that all too frequently our plan was NOT executed that I decided that to help with 'trustworthiness', I had to be present for experimental design execution and measurement work. Meant alot of long nights on 'C' shift, or learning new processes to ask the critical questions...but we ALL felt that our data was more trustworthy.   In the Toyota Production System there is a concept called 'Gemba'...which loosely translated means 'go to where the work is done and see.' That's what I tried to do. ... View more

Without question...W. Edwards Deming. I was newly degreed (MS in Applied Mathematical Statistics) and had the opportunity to attend Dr. Deming's famed 4 day seminar. I had a pretty inflated opinion of my skills and abilities and was ready to kick some butt at work.   Dr. Deming began the event by asking the audience a simple question, "Why are you in business?" An attendee then responded, "We're in business to make money. To earn profits." Deming's retort was, "Those of you in business to make money, to earn profits, at the expense of customers, will soon enough have neither of both." I was scratching my head. What?   By the time I had experienced The Funnel, Red Bead, Deming's 14 points, Seven Deadly Diseases, and System of Profound Knowledge...I was humbled...and realized my degree was just a piece of paper signifying, well, not much. I had learned by the close of the seminar I had so much still to learn to call myself a competent statistician. Hopefully I made some progress by the time I retired? ... View more

There are many confusing or unclear points in your original post. So rather than parse through them I'll focus my comments on question #1. Answer: Yes. Have you taken a look at JMP documentation? There are literally scores of regression and non regression methods in JMP...ranging from good old fashioned ordinary least squares, to more elegant/flexible general linear models, to completely non linear/non regression based methods such as the Partition and Neural Network platforms. ... View more

@Ella  The detailed steps to execute the workflow I suggest in my latest reply to this thread are far too many and detailed and lengthy, and some depend highly on domain expertise more than a statistical theory or technique to articulate in a JMP User Community Discussion Forum thread. I have two suggestions for you from this point forward:   1. Since it sounds like you have little to no previous experience using JMP, DOE and modeling for process optimization problems find someone in your organization, or hire a consultant, who has done this before and has some experience. If for nothing else they'll be able to guide you through the platform workflows, offer suggestions along the way, and ask thought provoking questions that will help you continue to frame your problem and approaches to a solution.   2. If you are truly all alone and have no other technical resources available for a detailed workflow for what I'm suggesting, I recommend looking at this JMP Blog and video written by my former (I'm retired) colleague @robert_anderson . The blog post shows the basic workflow and offers some guidance on navigating the JMP ecosystem to operationalize the workflow I'm suggesting. Here's a link to the blog post: Finding the best process operating conditions using 2 optimisation approaches    ... View more

@Ella If your target value is somewhere between 0.35 and 0.36, your predicted variation is not -0.3 to 0.3 but closer to +/- 0.1. And I suspect the largest contributor to the residual variation in this space is due to factor changes...not LOF. Take a look at the residual plot in the neighborhood of 0.35 and 0.36 to see my point. Here's a follow up thought for you:   I suggest taking your analysis all the way through to the JMP Prediction Profiler and then using JMP's Desirability, and Desirability optimization capabilities to find a factor space optimum centered on 0.355 with upper and lower JMP Desirability Function bounds at 0.35 and 0.36. Then if you have some process knowledge of how much to expect the factors to vary at those factor recommendation/optimums, run a Monte Carlo simulation within the Prediction Profiler to see what the variation in your response is. If your simulated factor variation in the simulation is relatively large, then LOF will be the least of your issues. ... View more

I concur with all that @Dan_Obermiller contributed. From your residual plot it looks like the root cause of the lack of fit is a relatively small replicate response variance compared to the total response variance.   Here's another thought for you to consider...what is the purpose of experiment/study? If it's prediction across the factor space, how impactful is the magnitude of the prediction error across the experimental space from a PRACTICAL point of view? If it makes little practical difference, then, quite frankly, I couldn't care less that you have LOF. ... View more

To reinforce @Dan_Obermiller and @statman 's advice...avoid viewing statistical significance as a binary cliff,,,,either it is or is not significant. Suppose you had a critical p value of 0.05 as so many adopt. Are you willing to say that an effect's p value of 0.051 is NOT significant, and willing adopt significance if it's 0.048? One or two decimal places off in a response could easily make something like this happen.   And what does your graphical analysis of the data suggest? Please tell me you plotted the data six ways from Sunday BEFORE you did any modeling. ... View more

I agree with all the @markbailey has shared. To add my two cents, if you don't have JMP Pro, here's an idea for you to consider where you don't have to invoke informative missing on a platform dialog specification window, but get the moral equivalent of said invocation in the modeling analysis...the 'missingness' will be sort of hidden and not as efficiently/neatly displayed in tabular or data visualizations...but it will be in there.   First a bit of an explanation of how JMP handles 'missingness' for nominal variables/effects. In general the concept is called 'informative missing' throughout the JMP ecosystem. The primary presumption wrt to informative missing is that there is INFORMATION within the system of study that leads to the observation being missing. Maybe an underlying cause? Commonality? Or lurking common variable that leads to the missing observation?   For example, suppose one were to survey a bunch of fishermen about how many fish they caught and you were trying to model effects that influence the number of fish caught. In the survey one of the questions is 'fishing method'. And there are three options for respondents to pick from; 1. trolling, 2. spin casting, 3. still fishing. Respondents are given the option to leave a field blank...that is it's now 'missing' within the observation. Now let's suppose for some reason the respondents that tend to have the highest fish caught counts are also those that chose to NOT ANSWER THE QUESTION. And this makes some sense doesn't it? Fisherman are notorious for keeping their secrets so there is information wrt to the study when respondents leave out the answer. So now one would want to try and understand the underlying mechanism behind why the data is missing to better understand the entire system.   Now here's my suggestion within JMP. If you are willing to presume there is information in the 'missingness'...and this is a big presumption...then you can replace the missing cells with a psuedo value nominal value that is common in each missing cell. For analysis purposes, informative missing in JMP/JMP Pro generally treats the missingness as another level of the nominal variable. Hence it will be included in any effects that include that term. Here's the JMP documentation regarding informative missing within the Partition platform that is NOT part of JMP Pro. So make sure you closely examine each modeling platform dialogue specification box you are thinking of using...you may NOT have to mess with your data table after all:   https://www.jmp.com/support/help/en/15.2/#page/jmp/informative-missing-2.shtml       ... View more

Have you tried a simple I/mR chart of daily tweet count? ... View more

Here's my two cents...I welcome commentary:   In one sense, what you may be describing is the process of sequential experimentation as a guiding principle to problem solving. If so, there is much in the literature on this process. I suggest reading the first few chapters of Box, Hunter and Hunter "Statistics for Experimenters" as a starting place. In a very real sense, all practitioners of sequential experimental design as a process always leverage earlier experimentation with subsequent experimentation. No quarrel at all with this method. All up all in my experience is it's the most efficient problem solving process for DOE centric investigations.   If what you are really after is merging two or more experimental designs into one Mother of All Experiments and then analyzing, this method is fraught with many many pitfalls and watch outs. Not the least of which could be inconsistent hypotheses, experimental goals, and objectives for each of the ingredient experiments...with the hypothesis, goals and objectives of the 'Mother' inspired investigation. So above and beyond maybe using the 'Mother' experiment for exploratory data analysis I'd be leery of much beyond that. ... View more

To add to @markbailey and @statman 's comments/counsel have you plotted the data in a meaningful fashion BEFORE modeling? Since you conducted an experimental design, Factor and block vs. Response plots, simple distribution of the responses, and a response vs. experimental execution order are a minimum set of plots I'd suggest you create. With an eye towards answering the following questions, which ultimately are used to build process understanding AND help insure we've got clean data for modeling:   1. Do the factor/block plots suggest something that flies in the face of known biological/physical phenomena? Back in the day, I'd suggest to the engineers we look at these plots. If they show the moral equivalent of 'the plots suggest water runs up hill, well, we've got a problem'.   2. Are there any response outliers that suggest unanticipated lurking variables that may have crept into the conduct of the experiment?   3. Does the experimental execution order plot of responses suggest a trend or suspicious pattern that again there may have been a lurking variable within experimental conduct?     ... View more

I guess I’m OK with curious people that want to fully understand the math behind these sorts of statistics. But for decision making, process understanding and communicating and sharing the data, conclusions and recommendations in the process capability space,,,, Simply put there are three rules for successful data analysis: 1. plot the data 2. Plot The Data 3. PLOT THE DATA There are so many useful data visualization techniques in JMP’s Process Capability and Process Screening platforms, only those suffering a severe case of ‘mononumerosis’ would bother with capability indices. ... View more

@David_Burnham: I had to chuckle. Back in the day when teaching basic quality and statistical methods I used to say, “The two most abused statistics in Statistics are R^2 and Cp/Cpk. Buy me a beer at the bar tonight and sit back and listen to my rant.” ... View more

My suggestion will offer guidance on which variables are correlated, and note, correlation is not causation, especially with happenstance data, which I presume you have. Just look for changes in the x’s that align in time (noting there may be lags) with changes in the response. ... View more

When you say, 'trend'...I'm presuming you mean a time based systematic change in the response. For the continuous predictor variables I suggest a simple overlay plot in Graph Builder leveraging the time sequence of collection of the responses (maybe with a local data filter to highlight specific x's efficiently). For the categorical variables is there still a lurking time element? Say, like batch 1, 2, 3, 4, etc. where each subsequent batch is bounded in a time sequence? If you've got a time sequence to the categorical variables, use the Value Ordering column property to get them to show up in the correct time sequence in Graph Builder. If you all you are really interested in is detecting this time based trend...you should be able to 'see it' in Graph Builder and not even need to go to any modeling. ... View more

I agree 100% with all of @markbailey 's input and opinions. @34South , you many also want to watch this online webcast: https://www.jmp.com/en_us/events/ondemand/analytically-speaking/to-explain-or-predict-that-is-the-question.html Dr. Shmeuli covers lots of different but important topics wrt to selecting modeling methods.   Lastly, if you have JMP Pro, you'll find the Model Comparison and Formula Depot platforms especially helpful for efficiently comparing the performance of multiple modeling outcomes. And you'll also have access to additional modeling platforms that are adept at handling categorical response data. For example, the penalized regression and PLS platforms may be very helpful. ... View more

Because the default number of observations to create each moving range data point is 2, and those 2 observations are successive wrt to the time series sequence. ... View more

You ALWAYS have to worry about the size of the data set so long as the data set is a sample from some population(s) of interest. But it's not just the size...size influences statistical risk, but contributes little to nothing wrt to representation risk.    Why do you have to always worry about the size? Because there will always be risk of Type I or Type II errors regardless of sample size. I remember years ago, in an advertising brochure for the Master of Science in Applied Mathematical Statistics program at Rochester Institute of Technology, in the brochure they defined 'statistics' as the 'Science of decision making in the face of uncertainty.' It's this uncertainty, which is made up of two components, statistical risk, which is influenced/controlled in part by sample size, and representation risk. Sample size speaks only to statistical risk...not representation risk.   A small but illustrative example of the interplay of statistical and representation risk: Suppose we have oodles (that's an unscientific term for a LARGE amount of data) of observations on pilot equipment and scale and we've got p values to die for...all but absolute certainty that we have ALL the key x variables and none of the non influential variables in our model. BUT, and here's a big BUT, we're now going to make PREDICTIONS about manufacturing processes and manufacturing scale...say going from beaker (pilot) scale, to rail car (manufacturing) scale. And our results are found to be not even close to the predictions. Then sample SIZE meant nothing. Representation risk bit your behind big time.   Why people focus on sample size and all but ignore representation risk is just mind boggling to me. I get it...it's easy to talk about statistical risk, alpha, beta, difference to detect, Type I and II errors, all easy peasy to get your mind around and we've now got computers to tell us what we need to know...no longer need those pesky ASQ/ANSI look up tables.   But NEVER forget about representation risk...rant over. ... View more

I'll leave the cross validation conversation to others. Here's my take on the applicability of neural networks,,,and I'll be philosophical as well. I think the issue of applicability without context wrt to the practical problem at hand, the data (is the data set wide/narrow, multicollinear etc.), risks both statistical and representation, and end use case for the model, will all have a greater influence on the applicability of the method compared to the outright sample size alone all by itself.   For example: What's the purpose of the study? Prediction or exploration? Is there the opportunity for additional study...perhaps in the context of designed experiments? And on and on.   For some insightful commentary on the prediction vs. exploration construct, I suggest taking a look at this:   https://www.jmp.com/en_us/events/ondemand/analytically-speaking/to-explain-or-predict-that-is-the-question.html   ... View more

@natalie_ You don't need to install a JMP Add In. The calculator you saw in the Youtube video ships natively with JMP...it's found here:   From the JMP Main Menu choose: Help->Sample Data (Then within the Teaching Resources dialogue box, select Teaching Scripts) - > Interactive Teaching Modules -> Sampling Distribution of Sample Means ... View more

To complement @markbailey 's advice I also suggest watching this on demand webinar from my former colleague, @JerryFish    https://www.jmp.com/en_us/events/mastering/topics/evaluating-and-monitoring-your-process-using-msa-and-spc.html     ... View more

@irosso Yes indeed, the "?" hover, right click, "Help" is a cool feature in JMP. Or as my colleague @gail_massari schooled me many years ago, "Right click is your friend in JMP!"   When I used to teach JMP to new users I'd always recommend, "If ever you are stuck and want to change/modify/remove/edit anything in JMP, and you don't know where to turn, hover over the thing you want to change, then right click. Chances are good a pathway to your desired change is now apparent."   Good luck! ... View more

Have you looked through the JMP documentation? I also suggest perusing the the JMP User Community File Exchange in case someone has written a script or Add In that can perform these techniques directly.   Near as I can tell native JMP or JMP Pro does not have these specific capabilities. But it appears to me that all three are primarily variable selection techniques and JMP and especially JMP Pro have a pretty rich set of variable selection techniques...regression based, non linear (tree, neural nets,,,,). So you might be able to solve your specific problem using these native techniques. ... View more