July 29, 2016, 02:10:27 AM

Author Topic: 5D Mk3 Web's Most Popular Camera, based on true online EXIF info.  (Read 1855 times)

unfocused

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 3135
    • Mark Gordon Communications
Re: 5D Mk3 Web's Most Popular Camera, based on true online EXIF info.
« Reply #15 on: July 28, 2016, 12:36:25 PM »
...The iphone is certainly represented in this random sample...

Let's try this again. This is NOT a random sample and it does not claim to be.

People consistently use "random sample" without any concept of what it means. As I explained above when Dilbert incorrectly used the term, this is a compilation of information from multiple photo sharing sites. They did not randomly sample anything, they simply compiled the data. (To their credit, they never claimed it was a random sample.)

To conduct a random sample, you must first decide on the universe you are going to test and them select a sample from that universe using a process of random selection. There are a number of ways you can randomly select a sample, but they all have one thing in common: In order to be accurate, every member of the universe being sampled must have an equal opportunity to be selected.

canon rumors FORUM

Re: 5D Mk3 Web's Most Popular Camera, based on true online EXIF info.
« Reply #15 on: July 28, 2016, 12:36:25 PM »

Maiaibing

  • 5D Mark III
  • ******
  • Posts: 648
    • Copenhagen Cycle Chic Pictures
Re: 5D Mk3 Web's Most Popular Camera, based on true online EXIF info.
« Reply #16 on: July 28, 2016, 05:20:09 PM »
...The iphone is certainly represented in this random sample...

Let's try this again. This is NOT a random sample and it does not claim to be.


You can write that 100 times and it will be wrong every time if you are talking about statistics. This is not about your everyday random where anything deliberate is non-random. Its about whether the sampling is relevant or not - and this is certainly not a sample that will say anything at all about the most popular camera on the web.

Thus it is a random sample because the methodology makes it random. Random is not a statistical abstract. This sample is random or non-random according to the claim that the numbers say something about the most popular camera on the web (the sites heading).

To spell it out: Its random in at least two statistical dimensions: 1) the selection of web sites do not representative of pictures posted on the web making it random to the finding the web's most popular cameras (which is the relevant measure here according to the site) and 2) the methodology does not take into account the presence/availability of some/all/correct exif.

And to top off we can put the sample to a simple practical test: are the results believable/close/accurate or anything like this when it comes to the claim we are being presented stats of the web's most popular camera - no.

Its so far off its laughable to even sit here and start arguing with someone that this is a representative sampling of the web's most popular camera. It is patently clear that it is not. The unweighted/corrected selection of sites and unweighted/uncorrected use of exif data and pictures with exif data determines this.

Yeah stats can be fun, but these do not - and cannot - support any legitimate claims about the web's most popular cameras (or their settings for that matter).

unfocused

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 3135
    • Mark Gordon Communications
Re: 5D Mk3 Web's Most Popular Camera, based on true online EXIF info.
« Reply #17 on: July 28, 2016, 06:44:00 PM »
It is a random sampling which is the correct method to use.

It's not a random sample and doesn't claim to be. It's a compilation of statistics from specific photo sharing sites. Nothing random and nothing sampled.



...The iphone is certainly represented in this random sample...

Let's try this again. This is NOT a random sample and it does not claim to be.


You can write that 100 times and it will be wrong every time if you are talking about statistics...

Quote from: Wikipedia
In statistics, a simple random sample is a subset of individuals (a sample) chosen from a larger set (a population). Each individual is chosen randomly and entirely by chance, such that each individual has the same probability of being chosen at any stage during the sampling process, and each subset of k individuals has the same probability of being chosen for the sample as any other subset of k individuals.[1] This process and technique is known as simple random sampling, and should not be confused with systematic random sampling. A simple random sample is an unbiased surveying technique.

I give up. That sound you hear is me banging my head against the wall.


canon rumors FORUM

Re: 5D Mk3 Web's Most Popular Camera, based on true online EXIF info.
« Reply #17 on: July 28, 2016, 06:44:00 PM »