Control Charts are easy in JMP 10

 

Photo by Tesla Motors

JMP 10 introduces a snazzy way to create control charts with a new platform called Control Chart Builder. JMP's prior Control Charts were designed over two decades ago for the classic process engineer in manufacturing and were past due for an overhaul.

The more we listened to which customers were running control charts and how they were using them, the more I was surprised to learn that people from nearly any field, including health care, medicine, sales, banking, scientific fields and education were running control charts. This was when we realized JMP needed something easily accessible to everyone.

The new Control Chart Builder platform in JMP 10 works much like Graph Builder with a drag-and-drop interface. It displays an Individual and Moving Range chart the instant you drag over a Y, and switches to an XBar and R chart when you drag over a subgroup column.

To replicate the chart below in JMP:

  • Open Cars Sold.jmp, which you can download from the JMP File Exchange on the JMP website
  • From the Analyze menu, choose Quality and Process->Control Chart Builder
  • Drag Volume ($M) sold to the Y-axis area of the chart
  •  To see the labels on the X-axis, drag the column Date to the X-axis area of the chart to see date labels on the X-axis

Since sales managers are on alert for low-volume months and always looking to repeat high-volume months, the limits on Control Charts can be useful in sifting out the noise from the more interesting events.

The car sales volume data in this example comes from the U.S. Census Bureau, Economic Indicators.

In the plot above, we first notice the drastic drop in volume for June and July 1998, caused in part by the General Motors labor strike. In 2008, we see the deep continued drop in sales resulting from the 2008 recession. The blue reference line is set at December, 2007, which the US National Bureau of Economic Research states was the beginning of the late 2000’s recession.

To see just the top (Individuals) chart, simply click on the axis label for Moving Range and drag the label out.

We hope you will find Control Charts are more fun to drive now.

tags: Control Charts, Data Visualization, JMP 10, Quality Improvement

6 Comments

  1. Ryan Ong
    Posted March 21, 2012 at 6:57 pm | Permalink

    Will JMP 10's Control Chart Builder have a way to switch to Levey Jennings or other types of control charts?

    • Posted March 22, 2012 at 9:14 am | Permalink

      Yes! The Control Chart Builder turns the control chart inside out, letting you set the statistics used for the points and limits separately, which makes the control charts more flexible. To switch to a Levey Jennings chart, you need only change the sigma used to create the limits. To do this, right-click on the graph and choose Limits->Sigma, then Levey Jennings.

      Take a look at both the Points->Statistic and the Limits->Sigma menus and you'll see many other types of control chart choices.

  2. Christine Huang
    Posted July 16, 2013 at 5:48 pm | Permalink

    How did you get the date on the x-axis? I only see "subgroup" on the x-axis. When I go to put the date on the x-axis for the independent and moving range chart, the graph changes to an Xbar and R chart. I want to maintain an independent and moving range chart with time on the x-axis.

  3. Posted July 26, 2013 at 9:42 am | Permalink

    Thanks for your question. You’re correct that the variable on x-axis acts as a subgroup variable. In this example, all the dates were unique, so the subgroup size was one. The only time it switches from an Individual and Moving Range set to an Xbar and Range set of charts is when the subgroup size is greater than one.

  4. Lisa Hall
    Posted November 8, 2013 at 1:50 am | Permalink

    Is it possible to show a trend line on an SPC chart, similar to what can be added to a line chart in Excel?

    • Annie Dudley Zangi Annie Dudley Zangi
      Posted February 5, 2014 at 3:02 pm | Permalink

      Lisa, While neither of the Control Chart platforms in JMP have an option to plot a trend line, you can get the trend line thru Bivariate, then add the line to any graph using the Customize.. option. Here's an example for the Y Function used to plot a regression line on a graph:

      Y Function(131.49624 - 2.3084042*x,x);

2 Trackbacks

  1. [...] my last blog post, I noted that the Control Chart Builder is an overhaul of the older Control Chart platform. One of [...]

  2. [...] my first Control Chart Builder blog post “Control Charts are easy in JMP 10,” I used a column named “Date” on the X-axis. When I downloaded the data, it didn’t arrive [...]

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <p> <pre lang="" line="" escaped=""> <q cite=""> <strike> <strong>