StatTutor: Race of mother and low birth weight (Question 2)

Description of activity and Excel instructions: StatTutor Race of mother and low birth weight (Question 2)

Dataset: Direct link for low_birth_weight.xls not available.

Exploratory analyses

Use Datapilot in Calc to create a table of counts and percents:

• Highlight the column(s) of data you want to use.
• Click Data > DataPilot > Start....

The Select Source dialog displays.

• Click Current selection.
• Click OK.

The DataPilot dialog displays. Now we have to tell Calc how to construct the table:

To the right of the various fields windows you should see a button for each of the variables selected.

• Drag and drop the desired response variable into the Column Fields window.
• Drag and drop the desired explanatory variable into the Row Fields window.
• Drag and drop either variable into the Data Fields window.
• Double click the variable in the Data Fields window or select the Options button on the right side.

The Data Field dialog displays.

• Select Count in the Function list.
• Click OK.

You now need to specify where to put the table.

• Click the More button in the bottom right corner of the DataPilot dialog.
• In Results to choose - new sheet - from the drop down menu.

• Click OK to exit the DataPilot dialog.

You should now have a table in a new sheet that shows the number of each type of data entry (including empty for missing data) along with a total. This is the count of each of these entries.

• Rename the sheet "Q2".

Let's create a second table with entries that are percent of total.

• Right click anywhere in the table.
• Select Start....

The DataPilot dialog displays.

• Select the variable in the Data Fields window.
• Click Options... on the right.

The Data Field dialog displays.

• Click More in the bottom right.
• In the Type dropdown menu, select % of row.
• Click OK.

Now let's specify a location for this new version.

• Click the More button in the bottom right corner.
• In the Results to row, use the shrink button to specify the top left corner of the table, say cell A12.
• Click OK.

A second version of the table is displayed, but, sadly, with the proportion in each cell.

• Change the proportions to percents using cell formatting.

Return to the StatTutor exercise for Question 2 to report on the relationship between the two variables: describe the key features of the data display and support your description with numerical measures. (Be sure to include the numerical results in your description.)

Calculate p-value for chi-square test

To calculate the chi-square test, you need two tables, one with the observed counts (which you created in the exploratory analyses above) and one with the expected counts. Make a new table with expected counts to the right of the table with observed counts. Appropriately label the columns and rows and use the following formula to calculate the expected count for each cell:

expected count = (column total*row total)/table total

Using the tables of observed and expected counts, calculate the p-value for the chi-square test.

• Pick an empty cell and type "p-value".
• In the cell to the right type, "=CHITEST([observed range],[expected range]), where [observed range] is the range of data in the table of observed counts (without row and column headers or totals), and [expected range] is the range of data in the table of expected counts (again, without row and column headers or totals).

Return to the StatTutor exercise for Question 2 to report results and draw conclusions.