Average v. Median Salary Forum

[Kobe_Teeth]

Which one matters more?


I feel like average skews and median is more accurate. Is this correct?

[sissyclark]

I would think it would be the other way around.

[cardnal124]

Mean salary skews data if there are a significant number of outliers, i.e. a lot of salaries at 160k or 40k.

Median gives a better feel for the middle salary people are actually making, and is not skewed by outliers, so is probably a better measure.

[Vincent Vega]

I think median is a better measure, as long as reported salary information is absolutely accurate.

[Kobe_Teeth]

I think median is a better measure, as long as reported salary information is absolutely accurate.

School had 72% report. I'd say that's pretty good.

[darknightbegins]

The median is likely what you can expect coming out from a law school. You are right that an average can skew more since, by definition, it is a combination of all the salaries students report then divided by the number reported. The median is a single score and is less likely to skew. People who graduate from law school tend to have a vast range of salaries they pull down, from people who can't even get a job and are waiting tables to those doing public interest work to the BigLaw job. Look at the median.

Also, check how much of the class is reporting a salary. Some schools are up in the 90+ percent. If it is any lower than that than even your median score could be skewed. Salary data is self-reported and those who have a bad starting salary are less likely to report it. So if I look at a school that has a median starting salary of 80K but only 70 percent is reporting and look at a school that has a median starting salary of 60K but 95 percent is reporting I am likely to trust I'll be pulling down 60K at the second law school. At the first school I am more inclined to believe I'll be making quite a bit less than 70K coming out.

Elementary statistics my dear Watson! Your undergrad never taught you median vs avg?

[darknightbegins]

Mean salary skews data if there are a significant number of outliers, i.e. a lot of salaries at 160k or 40k.

Median gives a better feel for the middle salary people are actually making, and is not skewed by outliers, so is probably a better measure.

Mostly right. The Median can still be skewed if you have small percent reporting starting salary. Anything below 90 or 85 percent is what I consider a "small percent".

[Rowinguy2009]

The median is likely what you can expect coming out from a law school. You are right that an average can skew more since, by definition, it is a combination of all the salaries students report then divided by the number reported. The median is a single score and is less likely to skew. People who graduate from law school tend to have a vast range of salaries they pull down, from people who can't even get a job and are waiting tables to those doing public interest work to the BigLaw job. Look at the median.

Also, check how much of the class is reporting a salary. Some schools are up in the 90+ percent. If it is any lower than that than even your median score could be skewed. Salary data is self-reported and those who have a bad starting salary are less likely to report it. So if I look at a school that has a median starting salary of 80K but only 70 percent is reporting and look at a school that has a median starting salary of 60K but 95 percent is reporting I am likely to trust I'll be pulling down 60K at the second law school. At the first school I am more inclined to believe I'll be making quite a bit less than 70K coming out.

Elementary statistics my dear Watson! Your undergrad never taught you median vs avg?

I have always questioned the bolded. If I had a horrible salary, I think I would make extra sure to report it due to bitterness towards my school. Maybe that's just me though; it's absolutely true that a median salary with a low percentage of grads reporting should not be trusted too much because you have absolutely idea what the non-reporting grads are making/doing.

[Aberzombie1892]

Mean salary skews data if there are a significant number of outliers, i.e. a lot of salaries at 160k or 40k.

Median gives a better feel for the middle salary people are actually making, and is not skewed by outliers, so is probably a better measure.

Mostly right. The Median can still be skewed if you have small percent reporting starting salary. Anything below 90 or 85 percent is what I consider a "small percent".

Have you seen the US News salary data? 99% of schools have less than 80% reporting. Some have less than 40%.[

quote="Rowinguy2009"]
I have always questioned the bolded. If I had a horrible salary, I think I would make extra sure to report it due to bitterness towards my school. Maybe that's just me though; it's absolutely true that a median salary with a low percentage of grads reporting should not be trusted too much because you have absolutely idea what the non-reporting grads are making/doing.[/quote]

To be complete honest with you, I think schools lie about it (I have no evidence, just a feeling).

If you report your data, and the schools 25% salary is relatively low, you will assume your data was counted. No single student is going to go around contacting all of his/her classmates 9 months after graduation to ensure that the school isn't lying.

[cardnal124]

Mean salary skews data if there are a significant number of outliers, i.e. a lot of salaries at 160k or 40k.

Median gives a better feel for the middle salary people are actually making, and is not skewed by outliers, so is probably a better measure.

Mostly right. The Median can still be skewed if you have small percent reporting starting salary. Anything below 90 or 85 percent is what I consider a "small percent".

But in that case, the mean wouldn't be good either, so median could still be the better measure, although neither are probably accurate.

[General Tso]

25th - 75th is the best measure....too bad USNWR only does it for private law firms (which only makes up 50-60% of most classes)

[Kobe_Teeth]

I have always questioned the bolded. If I had a horrible salary, I think I would make extra sure to report it due to bitterness towards my school. Maybe that's just me though; it's absolutely true that a median salary with a low percentage of grads reporting should not be trusted too much because you have absolutely idea what the non-reporting grads are making/doing.

+1

Last year I got something in the mail twice to report my employment to my UG school and I didn't do it. Just through shear laziness.

[darknightbegins]

Mean salary skews data if there are a significant number of outliers, i.e. a lot of salaries at 160k or 40k.

Median gives a better feel for the middle salary people are actually making, and is not skewed by outliers, so is probably a better measure.

Mostly right. The Median can still be skewed if you have small percent reporting starting salary. Anything below 90 or 85 percent is what I consider a "small percent".

But in that case, the mean wouldn't be good either, so median could still be the better measure, although neither are probably accurate.

I wouldn't argue that the mean would not be skewed just that the median would also be skewed. Overall if you don't have at least 90 percent of the class reporting I would be very suspicious of the starting salary for the school.

[Panther7]

I have always questioned the bolded. If I had a horrible salary, I think I would make extra sure to report it due to bitterness towards my school. Maybe that's just me though; it's absolutely true that a median salary with a low percentage of grads reporting should not be trusted too much because you have absolutely idea what the non-reporting grads are making/doing.

I don't.

I was unemployed, and consequently never got a questionnaire from my undergrad.

To this day they have no shyness about asking me for money, but never request my status (although I am since employed).