Thursday, 31 July, 2008

Extreme nonsense

An article titled Dump theory, act on facts, the Mumbai Mirror (July 31, 08) reports the counsel for the ICSE board arguing in court that the percentile system of evaluating students marks (introduced this year for admissions to colleges in Maharashtra) is not justifiable because 'this year the ICSE topper got 98.29 percent and the SSC topper 97.84 percent. The difference is hardly 0.4 percent.” 

Was he deliberately misleading the judges or is devoid of commonsense, not to speak of the preliminaries of statistics? Either has to be true. Or why would one quote the maximum marks? 

To do so here is no different than arguing that the Dutch (average height 6' 0.8" for men) are no different from the Chinese (men's average height 5' 4.9") because the tallest Dutch (7' 7.33") is shorter than the tallest Chinese (9' 5.39"). 

The idea, as far as I know, is never to look at the extremes but the middle measures (mean, median, mode) while comparing two groups, and then ask if they are dissimilar enough, and if they are, which one is exceeds the other on average.  

(I got the heights from the Net, and they are probably as wrong as the lawyer's logic.) 

No comments: