Math geeks

[Reggie Cleveland]

When dealing with a survey, what do you understand the difference, if any, to be between "mean" and "average?"

[Blues Fan in TX]

Originally posted by Reggie Cleveland@Aug 23 2005, 06:00 PM
When dealing with a survey, what do you understand the difference, if any, to be between "mean" and "average?"

Quoted post

Mean

In general, the arithmetic mean and the average are identical...there are other statistical mean values, however, that don't necessarily have an analogous counterpart in a list or array context...(i.e. harmonic mean, root-mean-squared, population mean for random variables, etc...)

[Reggie Cleveland]

Thank you very much.

Let's say we did a survey of the Lounge, and asked members to state exactly how they spent 100 percent of their time. Let's say these were the following results, labelled as "means":

Sucking dick -- 62
Sucking two dicks -- 19
Sucking three dicks -- 10
Sucking four dicks -- 6
Sucking five dicks -- 3

Is this any different than an average, both as individual line items, and as a composite portrait of the "typical" lounger?

[Airshark]

Originally posted by Reggie Cleveland@Aug 23 2005, 08:27 PM
Thank you very much.

Let's say we did a survey of the Lounge, and asked members to state exactly how they spent 100 percent of their time. Let's say these were the following results, labelled as "means":

Sucking dick -- 62
Sucking two dicks -- 19
Sucking three dicks -- 10
Sucking four dicks -- 6
Sucking five dicks -- 3

Is this any different than an average, both as individual line items, and as a composite portrait of the "typical" lounger?

Quoted post

You have the average lounger as pretty liberal there, Reg.

[chiguy]

Originally posted by Reggie Cleveland@Aug 23 2005, 08:27 PM
Thank you very much.

Let's say we did a survey of the Lounge, and asked members to state exactly how they spent 100 percent of their time. Let's say these were the following results, labelled as "means":

Sucking dick -- 62
Sucking two dicks -- 19
Sucking three dicks -- 10
Sucking four dicks -- 6
Sucking five dicks -- 3

Is this any different than an average, both as individual line items, and as a composite portrait of the "typical" lounger?

Quoted post

Those aren't averages or means, they're tabulations. The average number of dicks sucked would be: [(1x62) + (19x2) + etc.]/(62+19+...)

And there can be a difference between a mean and an average. An average is just what you would think -- the sum of the observations divided by the number of obsertations. It is a description of a set of observations. Conceptually, the mean is an estimate of some number based on a set of observations. When you're talking about surveys, for instance random sample phone surveys, the mean can differ from the average if there are sampling weights employed in the survey.

For example, it we're talking about a random sample survey of Americans conducted via random digit dialing, then it is a virtual guarantee that the percentage of respondents who are black will be lower than what you should get. Through a set of complex procedures, a sampling weight can be derived that counts information from each black respondent more than it does a white respondent. Say the weight is 1.5, in that case the black person's response would be (response * 1.5) while the white person's response is (response * 1) when calculating the mean.

I'm not an expert in survey sampling, but this is one way that the mean and the average can differ. Its worth noting that popular descriptions of surveys almost never tell you whether they are weighting the data, so you don't know whether you're getting the mean or the average. As pollsters learned back in the 1920s, little things like that can potentially produce discrepancies that matter.

[Reggie Cleveland]

Thanks. But if you're NOT weighting the data, then mean = average for all practical purposes, right?

[backstop]

Reg, for everything you and I deal with on a day-to-day basis, mean=average.

VORP VORP VORP VORP

[Dwight Schrute]

The mean is the middle number in a sample. The average is all of the numbers added up and divided by the total in the sample.
There can be a big difference.

For example the average annual baseball salary should be much higher than the mean salary. The reason is because the high end guys make a hell of a lot more than the majority. It shoves the average up but does not really relfect what the average (or middle) guy makes.

[Fishbone]

Originally posted by Beetlebop70@Aug 24 2005, 09:32 AM
The mean is the middle number in a sample. The average is all of the numbers added up and divided by the total in the sample.
There can be a big difference.

For example the average annual baseball salary should be much higher than the mean salary. The reason is because the high end guys make a hell of a lot more than the majority. It shoves the average up but does not really relfect what the average (or middle) guy makes.

Quoted post

You're confusing median with mean.

[Reggie Cleveland]

Reg, for everything you and I deal with on a day-to-day basis, mean=average.

That's what I thought.

[Dwight Schrute]

Originally posted by Fishbone+Aug 24 2005, 09:33 AM-->

QUOTE(Fishbone @ Aug 24 2005, 09:33 AM)

[phantom]

Originally posted by Beetlebop70+Aug 24 2005, 08:35 AM-->

QUOTE(Beetlebop70 @ Aug 24 2005, 08:35 AM)

Originally posted by [email protected] 24 2005, 09:33 AM

[Blue Chicago]

Here, this should help: LINK

[chiguy]

Originally posted by Reggie Cleveland@Aug 24 2005, 08:20 AM
Thanks. But if you're NOT weighting the data, then mean = average for all practical purposes, right?

Quoted post

Its hard to say without knowing more about the context of the problem. There are lots of ways where that might not be the case when you're dealing with survey data. But, its unlikely that in any application that you should treat these things as terribly different in practice.

[Reggie Cleveland]

Thanks.