Is there statistically significant evidence that the distrib
 Is there statistically significant evidence that the distribution of ratings is different for the 4 cars?
Can you use a X^2 test?
| Car | Good | Bad | OK | Total | 
| Honda | 17 | 22 | 77 | 116 | 
| Toyota | 3 | 23 | 64 | 90 | 
| Ford | 4 | 18 | 31 | 53 | 
| Mazda | 13 | 15 | 40 | 68 | 
| Total | 37 | 78 | 212 | 327 | 
Solution
Doing an Expected Value Chart,          
           
 13.12538226   27.66972477   75.20489297  
 10.18348624   21.46788991   58.34862385  
 5.996941896   12.64220183   34.36085627  
 7.694189602   16.22018349   44.08562691  
           
 Using chi^2 = Sum[(O - E)^2/E],          
           
 chi^2 =    15.46596829      
           
 With df = (a - 1)(b - 1), where a and b are the number of categories of each variable,          
           
 a =    3      
 b =    4      
           
 df =    6      
           
 Thus, the critical value is          
           
 significance level =    0.05      
           
 chi^2(critical) =    12.59158724      
           
 Also, the p value is          
           
 P =    0.016926242      
           
 As chi^2 > 12.5916, and P < 0.05, we   REJECT THE NULL HYPOTHESIS.      
           
 Thus, there is a statistically significant evidence that the distribution of ratings is different for the 4 cars at 0.05 level. [CONCLUSION]

