Google| Subject | Re: Colonial Photo & Hobby |
| From | Sandman |
| Date | 04/20/2014 17:23 (04/20/2014 17:23) |
| Message-ID | <slrnll7poj.sun.mr@irc.sandman.net> |
| Client | |
| Newsgroups | rec.photo.digital |
| Follows | Tony Cooper |
In article <tpj7l998fbjsf55639ta3so9jrcllvi8j9@4ax.com>, Tony Cooper wrote:
"data point" != "data".Tony CooperSandman
See how you had to make it plural to make the sentence work?
If you want to argue grammar, it was already plural. Data is the plural of datum. A data point is a datum that makes up the data.
A data point is an identifiable element in a data set. It's singular.
Incorrect.SandmanTony Cooper
Because "a single data point" isn't statistics.
Certainly it is.
To have data, you must have data points.You just used the plural form of "data point" above, Tony. Surely even you can recognize the difference? No?
All statistics are based on an accumulation of datum, and each instance of datum is part of the whole that is used to compile or project the statistic.But one instance alone is not statistics.
You cannot analyze or project a single data point statistically, but without it your statistic doesn't exist. It is part of the whole.If there is a whole, which there wasn't in this case.
No, I am correctly stating that statistics consists of more than one data point.SandmanTony Cooper
You have to have more than one data point for it to be statistics, and then one single data point of those many data points become important.
Well, not quite. You're implying that multiple data points are statistics.
You can have an unlimited number of data points without that group of data points being a statistic.No one has claimed otherwise.
Yes, quite.SandmanTony Cooper
I have four doors in my office. That's not statistics. If I were to compare it to the number of doors in other offices, I would have more data points and voila - statistics.
Again, not quite.
Remember, statistics is the *study* of accumulated data.If you have data. If you have one data point, you don't have data.
(Unless we say "statistics are", in which case we mean the results of the study) We can do a statistical study on four doors. We can put a counter on each door, count the number of openings/closing per work day, collect this data in sufficient quantity, and project which door is likely to used the most in future days within an expected probability of error.You just added more data points to the analogy, and thus - you have required information to create statistics.
The number of doors in one office or in many offices is not "statistics".In one office, no. In many offices - it certainly can be.
What you *do* with the number is what determines whether or not it will result in statistics.Of course.
Sandman
Incorrect. As long as you have one datapoint, the data is worthless, statistically.
Tony CooperYes, quite.
Not quite.
You have to have one, but what you mean is you can't have *only* one.Which "as long as you have one" means.
It can be argued, though, if you have a "data point", you have more than one. A data point is an instance within the data.You're learning. Good.
Incorrect.Tony CooperEric StevensSandman
Simple denial without explanation. I don't think you know much about statistics.
Ironic.
But a seemingly accurate observation.
-- Sandman[.net]
04/14/2014-
