Changing tables in Prism
Go to parent GraphPad Prism statistical analyses
Exercise 4: Adding column names to a table
In table Exercise2 (created in Exercise 2A):
- X-values represent temperatures in °C
- Y values represent predominant methods of disease transmission:
- Group A in a dry climate
- Group B in a humid climate
Look at the scatter plot of the data. |
---|
When you look at the navigator you see that the data that is open in the main window is highlighted in bold in the Data Tables section of the navigator (red). If you are viewing the data in the main window the name is not only highlighted in bold but also in blue. The corresponding graph (that represents the data in the table) is also highlighted in bold in the Graphs section of the navigator (green). This way you can always easily find the graph that corresponds to a certain table.
Open the Exercise2 graph in the main window. The first time that you open a graph, the Change Graph Type window is opened automatically:
Here you can define the type of graph you want to create. For this exercise:
A scatter plot of the data is generated but there is no annotation: no titles on the axes, no legend.... |
Note that when you look at a graph the content of the upper toolbar is quite different than when you look at a data table !
Add the annotation to the data table. |
---|
Just open the data table again in the main window and type the annotation in the column headers:
|
Look again at the scatter plot of the data. |
---|
Just open the graph again in the main window:
You see that the annotation has been added to the graph too: you now have a legend and a title for the X-axis. If you also want to include a title for the Y-axis just double click the title and type a new text e.g. Predominant method of disease transmission.
If you're finished typing just click anywhere outside the text box of the title. |
Exercise 5: Sorting rows
First we will sort the rows of a table in ascending order.
In Prism changing the order of data is regarded as a change of the data. This is why you can find the tools to change the order of your data in the Change section of the top toolbar.
In table Exercise2, sort the rows according to X-values in ascending order. |
---|
This sorts the data in the way you want:
|
Now we will reverse the order of the rows of a table. We will check if this has any impact on the graph of the data.
In table Exercise3, add annotation to the data (X = Time (h) and Y = Length (mm)) and look at the scatter plot. |
---|
|
Reverse the order of the rows. |
---|
This sorts the data in the way you want:
|
Look at the scatter plot of the reversed data. |
---|
In the Graphs section of the navigator click Exercise3. The graph looks exactly the same. |
Exercise 6: Excluding data values
Prism allows you to exclude values in a data table. Excluding data is seen as a change of the data set, not a transformation so you can find the Exclude button in the Change section of the upper toolbar.
Exclusion is mainly used to remove outliers from the data set. Outliers are extreme values that are substantially different from all other values in the data set. They can occur just by chance or they can be the result of a measurement error. The problem with outliers is that they can distort the statistical measures of a data set, misleading the interpretation of the results. The mean of a data set for instance is heavily influenced by the presence of outliers while the median is not.
Nevertheless, excluding outliers is very controversial especially in small data sets. We use the following rule of thumb: you can only delete outliers if you have a good reason for doing so e.g. if you know for a fact that a measurement error occurred.
As an example we will remove a data value from the Exercise2 data set. In this example we can remove the outlier because we know for sure a typing error has occurred.
Again we will check the impact of the transformation on the graph of the data. So first of all look at the data graph before you do the exclusion:
Remove the most extreme data value of the Humid climate data. |
---|
When you look at the data table you indeed see an outlier in the column of the Humid climate data: 18.
Since there are only 5 methods of transmission 1, 2, 3, 4 and 5, we know that "18" must be a typing error, so we may remove this value from the data set. To do so:
|
Is the data value still shown in the data table. |
---|
Yes, when you look at the data table, you see that the excluded value is shown in blue italics:
|
Is the data value still shown in the data graph. |
---|
No, the excluded value is no longer shown on the graph:
Note that the scale of the Y-axis has changed with respect to the graph of the original data so the graph might look very different but actually isn't. |
Important: do not exclude data values unless you have a good reason to do so