Plot the balance of data in an experimental design
This function provides easy visualization of the balance of data in a data set given a specified experimental design. This function is useful for identifying missing data and other issues (see examples).
ezDesign(
data
, x
, y
, row = NULL
, col = NULL
, cell_border_size = 10
)data |
Data frame containing the data to be visualized. |
x |
Name of the variable to plot on the x-axis. |
y |
Name of the variable to plot on the y-axis. |
row |
Name of a variable by which to split the data into facet rows. |
col |
Name of a variable by which to split the data into facet columns. |
cell_border_size |
Numeric value specifying the size of the border seperating cells (0 specifies no border) |
The function works by counting the number of rows in data in each cell of the design specified by the factorial combination of x, y, row, col variables.
A printable/modifiable ggplot2 object.
Michael A. Lawrence mike.lwrnc@gmail.com
Visit the ez development site at http://github.com/mike-lawrence/ez
for the bug/issue tracker and the link to the mailing list.
#Read in the ANT2 data (see ?ANT2).
data(ANT2)
head(ANT2)
ezPrecis(ANT2)
#toss NA trials
ANT2 = ANT2[!is.na(ANT2$rt),]
ezDesign(
data = ANT2
, x = trial
, y = subnum
, row = block
, col = group
)
#subnum #7 is missing data from the last half of the experiment
## Not run:
ezDesign(
data = ANT2
, x = flank
, y = subnum
, row = cue
)
#again, subnum#7 has half the data as the rest
#now look at error rates, which affect the number of RTs we can use
ezDesign(
data = ANT2[ANT2$error==0,]
, x = flank
, y = subnum
, row = cue
)
#again, subnum#7 stands out because they have half the data as the rest
#also, subnum#14 has no data in any incongruent cells, suggesting that
##they made all errors in this condition
#finally, subnum#12 has virtually no data, suggesting that they mistakenly
##swapped responses
## End(Not run)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.