POLL - which D800 model would YOU buy

Started Jan 7, 2012 | Discussions thread
bobn2
Forum ProPosts: 28,371
Like?
Re: Halftime Score
In reply to CriticalI, Jan 9, 2012

CriticalI wrote:

Nikon will have done a lot more well founded and rigorous surveys to establish the market for a new camera, and will use those to make their decision. If you like, what I'm doing is just registering the thought that this kind of exercise, while some may enjoy it, contributes no real information.

Sorry, am I under some sort of "obligation" to contribute information or can I just do something to satisfy my own curiosity? Judging by the response, quite a few other people find it interesting as well.

You are under no obligation to do anything at all. However what you are doing is in a public space, and you are putting forward opinions claiming this 'survey' as validation for them. I and anyone else who cares, am quite at liberty to point out the holes in your deductive process.

If it was done properly, it would yield the information that a self selecting sample of people on DPR had a particular opinion, but this does not even do that.

Yes, it most assuredly does. It's not just one you are interested in.

No, it doesn't. The way you have designed it, with built in (and completely unnecessary, BTW) confounding variables means that it most assuredly doesn't.

I could post a survey:
Which do you prefer
Nikon D800: 16MP, 25600ISO, 6FPS, 51 AF points, $4000
Canon 5DIII; 32MP, 51200 ISO, 8FPS, 61 AF points, $3000

And then when the majority answers Canon 5DIII claim that the majority prefers Canon, or whatever.

Yep, that would be a completely useless survey.

Yet it is useless for exactly the same reasons and in exactly the same way that your survey is useless.

I have used DPR for surveys, but to answer more subtle questions like 'which factors are most important in IQ'. One of the questions the survey didn't include was 'which factors are most important in IQ'.

Er, because that is not what I was interested in.

No, but it was what I was interested in. The point was that i was interested in answering a much more focussed, but more difficult, question, and the survey to produce a sensible answer needed to be somewhat more sophisticated than just asking them.

I was interested in what class of camera most were interested in based on the two major debates raging here, megapixels and video and four hypothetical cameras, all of which Nikon could build.

So, why four out of a huge number of possible specification variants. Why are your axes of classification? Why do your four choices contain arbitrarily different selections of features. Is the 'same price' a realistic assumption? What about the unstated specifications, are you expecting people to make assumptions about those? If so, what are those assumptions?

Since it was my survey I guess I get to choose what I want to ask. Is that OK with you Bob?

Of course it's OK, but it's also OK that I get to point out why there is no useful information coming out of what you have done.

It would be impossible short of running a Gallup pole of several thousand photographers to assess anything more accurately.

Nonsense. All you have to do is eliminate some of the elementary errors that you have made in survey design and you will get some valid information, you just have to know first what it is you are trying to find out. Of course, the more participants you have the more valid it will be, but things can have statistical significance with quite small samples.

But this survey is still evidence of a sort, just not conclusive.

No, it is not evidence of any sort. Even with a million participants, you would not be able to draw conclusions from it, conclusive or otherwise.

Conversations I have with other photographers and dealers simply back it up.

I'm sure they do. That is no better evidence.

I never claimed otherwise. You are simply constructing a strawman argument.

I am not constructing an argument, I am just pointing out the manifest flaws in your survey, which prevent any conclusions being drawn.

As for manufacturers surveys, there are plenty of examples of where they have been an unmitigated disaster.

I'm not aware of any. I am aware that management decisions taken as a result of surveys have been an unmitigated disaster, but that is quite a different thing.
--
Bob

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
D800New
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow