JMP can import files from a variety of formats. From the JMP menu select File > Open. In the Open Data File dialog navigate to the desired folder where ADK_Newborns_2014.xslx is stored and select “Excel files” from the file type drop-down menu. This will display the available Excel files. Select ADK_Newborns_2014.xslx then click OK and the Excel Import Wizard will appear as shown in Figure 11.1 Excel Import Wizard.

The wizard shows a preview of how the Excel data will appear in a JMP data table. Adjustments can be made in the wizard to indicate the number of rows in the Excel file that contain headers and the row and column where the data begins. Click the Import button to open the JMP data table.

Once the data is imported into JMP, the columns should be checked to be sure that they have the correct data and modeling types. There is no guarantee that the data elements as stored in the native file format are as needed for the JMP analysis. Other column formatting such as List Check may need to be applied.

The Data Table Columns Viewer is a convenient way to quickly review the modeling types that were assigned to each column on import. Select Cols > Column Viewer and a list of columns appears. The column Operating Certification Number is an identifier and should be changed to a nominal modeling type. This can be done by right clicking over the modeling type icon and choosing “Nominal” as shown in Figure 11.2 Changing the Modeling Type in Data Table Columns Viewer.

Some of the columns, such as APR Risk of Mortality, contain ordinal variables. The levels or ordinal variables can be ordered using the List Check column property. This is done for APR Risk of Mortality by right clicking in the column header and choosing Column Info and from that dialog select List Check from the Column Properties drop-down menu. Figure 11.3 List Check Dialog for APR Risk of Mortality Column shows the List Check dialog.

Use the buttons at the right to reorder the levels as desired.

The Data Table Columns Viewer provides a table of summary statistics for all columns in the data table. This table facilitates an initial review of the data to ascertain the amount of missing data and data anomalies. In the Data Table Columns Viewer select all of the columns and click Show Summary. A portion of the Summary Statistics table is shown in Figure 11.4 Columns Viewer Summary Statistics Table.

The sample size and number of missing values are given for each variable. The number of categories are displayed for nominal and ordinal columns. Descriptive statistics are displayed for continuous variables which allows identification of outliers.

Several columns such as Health Service Area have only one category. Selecting that column and clicking the Distribution button will display the JMP Distribution platform which shows all rows in this column have the value “Capital/Adiron”. This column will not be useful in the analysis as it has no variation and if desired can be deleted to reduce the size of the data table.

There are pairs of related variables such as CCS Diagnosis Code and CCS Diagnosis Description. Codes facilitate data input but are not useful to analysts or stakeholders unfamiliar with the meaning of the code numbers. A separate column gives the description of the code. This architecture is a best practice in data management. If a coded column is not useful in the analysis or presentation of results, it can be deleted from the JMP data table.

A JMP project is a convenient way to maintain all documents related to an analysis. This is consistent with best practices of reproducible analysis and is efficient for the analyst or researcher who wishes to revisit this work in the future. A number of objects can be included in a JMP project including documents, URLs, JMP windows, and database queries.

To create a JMP project select File > New > Project and click on the New Project icon. A project is added with the name “Untitled.” Right click the drop-down menu as shown in Figure 11.5 JMP Project Menu to view the list of actions that can be taken.

Figure 11.6 Populated JMP Project shows the contents of the project named “ADK_Newborns_2014” which includes the URL to access the SPARCS 2014 data dictionary, the Excel file downloaded from SPARCS, the processed JMP file, and a JMP Distribution analysis of Birthweight. The URL was lengthy, so the entry was renamed by right clicking over the URL and selecting Rename from the menu.

The file ADK_Newborns_2014.jmp contains a subset of the downloaded Excel data that will be used in this and subsequent cases. Several of the facility names have been shortened using the JMP Recode feature. This will improve the readability of graphs that use Facility Name. The data is now prepared and supporting documents are organized so the exploratory analysis can begin. Several of the JMP column names have been edited to include units.