Repeated-measure contingency table with two variables with many levels? I am trying to analyze the results of a survey on what would make people of different ages visit certain tourist places.
However, age was collected as age-group (not my fault). Hence, it is a categorical variable.
I have two variables: AGE GROUP (5 levels) and PREFERENCE (8 levels).
However, participants could indicate more than one preference.
Hence, PREFERENCE is a repeated-measure variable (i.e. one participant could contribute to two values, e.g. PREFERENCE1 and PREFERENCE2).
My table looks as below

I thought I could use a contingency table to analyze the results. However, I clearly violate the assumption of independence.
Is there a way to analyze the above table?
Please, I am just a social science student, and I am not very familiar with jargon.
 A: First some clarification: a contingency table is a type of table in a matrix format that displays the (multivariate) frequency distribution of the variables. It is not related to any test.
Second, you should be asking what test / model should I use for my data, given my research question? And in this case it appears that a Chi-square test is what you are looking for. And a Chi-square test takes the data in a contingency table form.
A Chi-square test works for arbitrary 2-dimensional tables, it is not restricted to the 2x2 case (binary variables).
However, you also have repeated measures, since some participants could vote more than once. So yes, the independence assumption of a Chi-square test are violated.
In this case, the general next step is to use McNemar's test, which is a Chi-square test for repeated measures. However, this only works for square matrices, so does not work for your case.
Lastly, an Expanded McNemar test (McNemar–Bowkera) or a Poisson mixed model could be used (a little more complicated), if you are interested specifically in which categories are different, since the Chi-square / McNemar's test only return a global p-value.
