Govt. organisation favouritism in its complaint handling I am analysing some data for a government organisation and trying to convince them that their complaint decisions are biased. I believe they favour certain organisations. 
I have put together the following data, and done a chi-square test on it. The chi-square value is 25.34. I believe that the outcomes are not entirely independent. Is that the correct way? Additionally, is there any robust way to determine which organisation they are favouring?
Company      Upheld   Not Upheld

Comp.1               11       14
Comp.2              13        47
Comp.3              13        111
Comp.4               4        20
Comp.5               6        57
Comp.6              17        42
Comp.7               7        18
 A: Summarizing from the comments and adding some more information.
As @NickCox pointed out, all chi-square does is show that you can reject the null hypothesis that the proportion of complaints upheld is the same across companies. If this is all the data you have, there is little more that you can do. 
You could add an assumption that complaints against each company are equally likely to be valid or justified. Depending on the particulars of the situation, that might be a reasonable assumption, a somewhat reasonable one, or a ludicrous one, but there is nothing in what you have posted that lets you show how reasonable it is.
If you have more data and, in particular, if you have data on how justified each complaint is, then there would be more you could do. You could model Pr(upheld) as a function of the company and the justification rating. The specifics of this model would depend on the nature of the data. 
I could see some situations in which the same people are making complaints against multiple companies or the same company multiple times. If that is the case, you would have to account for that.
