What inspired this Wish List Request?
Multiple requests from customers for a complete set of statistics from Orthogonal Regression. They are comparing us to Minitab on this front and we are viewed as lacking. There are other opportunities where we can further enhance our capabilities on this front (see: Add Orthogonal Regression to Response Screening) but this is has been highlighted as a serious deficiency in their comparisons. On investigation there are multiple requests - internal and external - for this stretching back as far as 2017, possibly further. See:
https://jmpdiscovery.atlassian.net/browse/JMP-8670
https://jmpdiscovery.atlassian.net/browse/JMP-8617
https://jmpdiscovery.atlassian.net/browse/JMP-13667
https://jmpdiscovery.atlassian.net/browse/JMP-8639
Please describe the new or improved feature or solution you would like to see.
Mark Bailey's suggestion (https://jmpdiscovery.atlassian.net/browse/JMP-13667) echos the requests from the customers. The only addition that I would make to it is to have confidence intervals on all estimates and statistics.
"The report for Fit Orthogonal omits much of the information that is available for other fits such as Fit Line. (See image with Linear Fit and Orthogonal Fit examples.) Investigate how additional information can be added to provide user with more results. For example, Fit Orthogonal is a common method for assessing if a new test method is comparable to a standard method. The only regression results available since JMP 4 are the estimate of the intercept and the point and interval estimate of the slope, A method comparison involves a decision about both the intercept and the slope.
Add appropriate information as Summary of Fit, Analysis of Variance, and Parameter Estimates like the other fits."
What would the value be to you and/or other users if this new feature is provided?
Orthogonal regression is a method commonly used for tool - tool matching in metrology for semiconductor manufacturing. Having the CI's for the intercept and slope are critical these analyses. We are fairly unique in not providing both of these statistics and forcing the customer to do a long hand calculation to get the CI's we don't provide.
Is this an external customer request?
Entegris, Seagate, Samsung.
Is there a competitive software package that has this feature or solution that we should be aware of?
Minitab, Python based solutions, etc... pretty much everyone that does Orthogonal / Deming Regression except us.
Will this impact a renewal, expansion, or adoption at a customer, and does this put an account at risk if we do not address this?
We are trying to integrate ourselves in the training programs for these accounts and organically displace competitors in the process. Not having core functionalities adequately covered will slow / impede the activity - impacting expansion and adoption efforts.
