|
I m a g i n e I T T o o l b a r Example on exploring data graphically Looking for correlations between body dimensions The spreadsheet below contains measurements that some students made on each other. This is how it looked.
Setting up the spreadsheet The data was typed in first of all. Then the table of data values (without the headings)
was highlighted by dragging over it with the mouse. Next the new
table button With any cell in the table selected, the Specify
table labels button To look for correlations between armspan and height
a scattergraph was produced, by selecting any cell in the armspan
column, then clicking the x-y scattergraph button Next, a bar chart showing heights was produced by
selecting any cell in the height column, clicking the Vertical bar
chart button The additional plot of armspan was added by selecting
any cell in the armspan column, clicking the Add plot button Downloading the example spreadsheets To download the spreadsheet already set up, right-click (Mac: control-click) this link, choosing to save to your hard disc, and then open the downloaded file from within Excel. If instead you want to download just the data and practise setting up the table and graph for yourself, use this link:
Changing what is plotted To change what is plotted on either axis of the
scattergraph, first make bring the graph to the front (by clicking
on it) and then click either
|
was clicked to indicate that these cells form the data table.
The default yellow colour for the background helps remember where
this data table is.
was then clicked and the rows chosen for graph
titles, labels for the axes, and labels for individual data plots
(green colour).
, choosing a normal Excel graph when asked, choosing height
for the x-axis, and finally the option not to label data plots.
, and choosing to have names on the x-axis, and data plots
labelled. 
, and then choosing the bar chart from
the alternative graphs displayed in turn.
or 
. The axis you have changed is automatically re-labelled. In
this way you can very rapidly plot any quantity against any other
in order to search for correlations, as shown below. You can also
change both axes at once and/or which rows are plotted by clicking
.
