Are outliers possible with categorical data Just want to make sure that I understand the meaning of an outlier. 

Question: Can you have an outlier of categorical data? 

I think that to have an outlier you must first have some sort of measurement.  My reason is that any data point > 3*IQR (Interquartile range) is used to identifiy an outliner. 
However, there is no measurement with categorical data, as I understand. 
 A: Suppose you have 1000 people choose between apples and oranges. If 999 choose oranges and only one person chooses apple, I would say that that person is an outlier. 
We use measurement as a way to detect anomalies. With categorical data you have to explain why choosing an apple is considered an anomaly (that data point does not behave as the rest 99.9% of the population). 
There are also papers that talk about outliers in categorical data, for example http://www.cs.umn.edu/tech_reports_upload/tr2008/08-008.pdf.  
A: I don't know what the standard treatment of this problem is, however I have a remark about the question. In order for the concept outlier to have any meaning you need to be able to define a distance between the values, that in this case may not be trivial i.e. is an apple closer to an orange or a pear?
A: After scanning the paper "Understanding Categorical Similarity Measures for Outlier Detection" by Chandola, Boriah, and Kumar, it seems that the answer is yes, you can have outliers in categorical data.
From a data point perspective, we can have a data set with data points containing categorical values. One example of similarity is the number of values that are the same (called overlap) in two data points; more overlap then more similarity. There are techniques other than overlap, and which one is best depends on the problems, however some (such as eskin) have been observed to work well across a larger number of problems.
Another way to think about categorical outliers is if a categorical value within a collection of values from that categorical variable is an outlier. One way to interpret a categorical value outlier might be in frequency. Categorical values with that occur very infrequently might be thought of as outliers.  One way you might use this is simplify the problem by grouping infrequently occurring categories together.
