another article I found interesting-Law Grads Are Angry

[bigs5068]

I would also like if your website included Median and Mean salaries. I think a median is more favorable to the schools and the mean salary would account for the low salaries that some graduates are receiving to give a more accurate picture.

[nealric]

Accuracy and truthful reporting would be extremely difficult to verify, and at least when it comes from the schools there's an aspect of accountability in how they report.  We also think schools would find it hypocritical if an organization that aims for honest reporting is willing to accept information from unverified sources.  It also duplicates the work schools already do in the way of data collection; of the nine components in the LST standard, career services offices already collect 7 of them for each graduate in answering the NALP/USNews surveys.   

I would think of self reporting as a bit of a "stick". You can tell the schools: either you release your data (accurate data), or we are going to present the survey data (which I suspect may be even worse than average due to the JD underground effect).

[IPFreely]

On top of that if I remember my 7th grade math properly you would be much more interested in the mean opposed to the median salary. Correct me if I am wrong, but MEDIAN would be like this if 7 salaries were reported as 25,000, 30,000, 35,000, 90,000, 100,000, 100,000 150,000, . The median salary would be 90,000. while the mean would be around 60,000. Reporting a median makes a school look a lot better if they can say 90,000 instead of averaging in people with very low salaries.

You fail at math.

You can create distorted data to argue for any of the three values -- mean, median, mode.  In any typical salary distribution, median would be the best number to use.  Unfortunately, we have that pesky bimodal distribution -- a consequence of the biglaw "we'll pay $160K to anyone we hire" salary-matching system -- that makes all of them equally useless.

You want transparency in this situation, the only way you can get it is to show a breakdown that separates biglaw out.  Something along the lines of "x% got biglaw jobs;  of the remaining y%, the median salary is $NN with a standard deviation of $X."

[bigs5068]

Yea mean, median, and mode can all be manipulated so why not give all three and fine your system makes sense. The bottom line is the data is completely inaccurate from Harvard to Cooley and it borders on intentional misrepresentation for any school to list their salaries the way they do. Again, this is not limited to law school all schools really exaggerate their stats, which goes to my other post that education is pretty much a scam, but one you have to deal with.