Here are the results of 74 voters
|
| Author | Content |
|---|---|
| ComputerBob Dec 19, 2011 9:57 AM EDT |
Really??? In a two week period, only 74 visitors to your blog voted in your poll? And you thought that the results of such a tiny poll were worth reporting to the world? Ridiculous. |
| darkduck Dec 19, 2011 11:58 AM EDT |
Show me better results... |
| Fettoosh Dec 19, 2011 12:15 PM EDT |
Quoting:Show me better results... Lack of better results doesn't give any credibility to yours. A sample of 74 visitors is just not good enough to make a sound statistical decision. That is all. |
| number6x Dec 19, 2011 1:15 PM EDT |
@DD, Thanks for publishing the results of your poll. I'd especially like to thank you for publishing the numbers instead of just percentages. Knowing that 31 / 74 respondents chose Xubuntu is much more informative that you saying that about 42% of respondents chose Xubuntu. Publishing the numbers let's me calculate that the standard deviation is on order of the average number of respondents per choice. This lets me guess that your sample size is way too small to be used in any strict, meaningful way. You cannot force people to visit your blog, and you cannot force those who visit to take the poll. It is nice that you offered the poll and that you published the results in a format that is useful. I hope you continue to offer polls like this. If your site becomes more popular and your visitors contribute their opinions at a greater rate, your results will become more meaningful. Even if you don't get enough visitors to have statistically significant results, you still offer a forum for visitors to discuss issues important to them regardless of how many vistors there are. |
| jdixon Dec 19, 2011 1:16 PM EDT |
> A sample of 74 visitors is just not good enough to make a sound statistical decision. That is all. That depends entirely on the size of the base. If you only have 75 users, then a sample size of 74 is more than adequate. |
| darkduck Dec 19, 2011 1:37 PM EDT |
to number6x: I have already made similar voting for best KDE distro. And there were almost as much voters before first announcement of results as after that. It made me to publish second round of results. And you know what? Numbers were different. Proportions were almost the same. It still means that statistically we're more or less correct, even if number of votes can looks small. These links are in my current post, but I repeat them here anyway. http://linuxblog.darkduck.com/2011/08/users-voted-for-best-k... http://linuxblog.darkduck.com/2011/09/users-voted-for-best-k... |
| number6x Dec 19, 2011 1:51 PM EDT |
Quoting:That depends entirely on the size of the base. If you only have 75 users, then a sample size of 74 is more than adequate. I was speaking of this case where the difference in standard deviation is about as great as the average number of respondents per choice. I would usually take that to mean I should keep polling to look for more data. I know there are more XFCE users out there. The first thing I wanted to respond with was : "@ComputerBob and @Fetoosh, then run your own poll if you don't like DD's". That would just be more whining and sound childish, so I thought it would be nicer to offer some constructive comments. Adding to the back and forth complaining is a waste of everyone's time. People here are usually not too 'whiney' and there was no reason for me to take their comments in the wrong way. They both regularly contribute to the conversation around here in a helpful and informative way. This is LXer after all, I like to think we are a little more civil and respectful of each other than some of the other news and comment sites :) |
| jdixon Dec 19, 2011 3:17 PM EDT |
> I was speaking of this case where the difference in standard deviation is about as great as the average number of respondents per choice. I would usually take that to mean I should keep polling to look for more data. I know there are more XFCE users out there. I'm sure there are. :) But Fettoosh made a general statement which is not universally correct. And if people will insist on snarking, sometimes I can't resist snarking back. |
| ComputerBob Dec 19, 2011 5:46 PM EDT |
It's not snarking to point out that, with 7 billion people in the world, and hundreds of thousands of people using Linux (???), the results of a 2-week poll that got only 74 responses are incredibly self-selected and useless for any real purpose other than to artificially increase traffic to DD's blog. In fact, if the LXer summary had stated up-front that there were only 74 respondents, I wouldn't have even wasted my time visiting and reading the blog. BTW, I flipped a coin 10 times today. It came up "heads" 7 times. So, to some of you, I suppose that proves that the odds are 7 out of 10 that when you flip a coin, you will get "heads." I'll write up a catchy "news" headline about it and submit it to LXer later today, so that lots of people will visit my blog. /sarcasm (for those who would misunderstand this post without it) |
| jdixon Dec 19, 2011 9:44 PM EDT |
> t's not snarking to point out that, with 7 billion people in the world, and hundreds of thousands of people using Linux (???) And how many use XFCE? I personally have no idea. > ...the results of a 2-week poll that got only 74 responses are incredibly self-selected and useless for any real purpose other than to artificially increase traffic to DD's blog. All online polls which allow open participation are self-selected. And most of them are useless. So? You don't like DarkDuck's articles. We've got that Bob. So stop reading them. I don't read stories by Ken Hess. > So, to some of you, I suppose that proves that the odds are 7 out of 10 that when you flip a coin, you will get "heads." Depends on the coin. Not all coins give a 50/50 result. Not all polls require a large sample size. Is this one in that group? Probably not, but I haven't seen anyone but number6x try to make that argument. |
| Grishnakh Dec 19, 2011 11:40 PM EDT |
My problem with these "best distro for $DE" is that I have to wonder how many voters have actually tried multiple distros (with that same DE), so that they'd be in a position to make an informed vote. Personally, I wouldn't vote in any of these polls, because I don't feel like I have sufficient experience to say "$distro is the best distro for using $DE!". I'm not one of these people that tries out every distro out there; I only use two at the moment (Kubuntu and Linux Mint KDE Ed), and I used to use SUSE years ago, and I've worked with Red Hat (before there was a Fedora). Given that I've only used 3 KDE distros, and only two of those recently (the SUSE one was KDE 3.5.x; that should tell you how old my experience is there), I would never put myself out there as someone competent to judge KDE distros. I've never tried Magieaeaeia (sp?), I have no desire to try the latest SUSE as long as that stupid patent agreement with MS is in effect, I haven't tried Fedora with KDE, and I haven't tried the other ones like Chakra, so who am I to say that one of the distros I've used is better than any of these? I imagine that most voters don't think this way, and just vote for what they use, so it really becomes a popularity contest, with no basis in informed opinion. This is why I much prefer articles where the author himself reviews various distros, and presents the good and bad about each one, so I don't have to go download and install 8 different distros myself to see which are good and which suck. |
| darkduck Dec 20, 2011 6:14 AM EDT |
@Grishnakh
I have tried different XFCE distros. Some in Live mode, but couple in installed. Links are in the article text. Maybe more to come (!#, for example)
And after all these tests I run the voting. Voting results confirmed my opinion that Xubuntu has valid reasons to dominate that market segment. Would it be useful if I write another post where I compare different distros face-to-face, not scattering views across different posts? |
| linuxsavvy Dec 20, 2011 6:35 AM EDT |
Quoting:Would it be useful if I write another post where I compare different distros face-to-face, not scattering views across different posts?Try putting up a comparison chart for several distros, based on well-researched information. |
| jdixon Dec 20, 2011 7:20 AM EDT |
> My problem with these "best distro for $DE" is that I have to wonder how many voters have actually tried multiple distros (with that same DE), so that they'd be in a position to make an informed vote Probably only a handful of the 74. But again, that would almost certainly be true with any open Internet based poll. > ...based on well-researched information. Well, that's a rather difficult to define condition, especially when the evaluation criteria are almost certainly going to be subjective. |
| ComputerBob Dec 20, 2011 10:52 AM EDT |
Quoting:You don't like DarkDuck's articles. We've got that Bob. So stop reading them.As I stated above, I would not have wasted my time reading this DD article if its summary had been up-front about the fact that it was a poll of only 74 users. Fool me once... |
| caitlyn Dec 20, 2011 10:57 AM EDT |
Quoting:Show me better results...DistroWatch page hit rankings would qualify. Xubuntu ranks behind Vector Linux, another (and IMHO better) XFCE based distribution. |
| jdixon Dec 20, 2011 12:30 PM EDT |
> Xubuntu ranks behind Vector Linux... Vector Linux <-- Slackware. Fast, relatively bug free, and stable. Xubuntu <-- Ubuntu. Slower, buggier, and less stable than Slackware. Hmm. I wonder ever why? |
| Grishnakh Dec 20, 2011 2:42 PM EDT |
jdixon wrote:> ...based on well-researched information. Well, that's a rather difficult to define condition, especially when the evaluation criteria are almost certainly going to be subjective. Nothing's perfect, but I'm much more interested in the well-informed opinion of someone who's clearly taken the time to do their research and back up their opinion with facts, rather than some random poll of random people who probably haven't done any research at all. Even in scientific writings, not everything is hard science and fully reproducible. Some fields just don't afford you that luxury, such as archaeology, so you have to take what you can get, because it's better than nothing. With comparisons of Linux distros, I don't see how you could ever have a comparison that's not subjective; it's just the nature of the beast. Different people like different things. There's even a few wackos and morons out there who like these new dumbed-down UIs! But with a full-length article wherein the author compares different distros and points out strengths and weaknesses of each one to back up his opinion, this information is much more useful, even if I disagree with him, than a simple popularity poll or anything else that lacks in-depth information and research. |
| Koriel Dec 20, 2011 3:45 PM EDT |
I just tried Xubuntu 11.10, installed from the 32bit x86 CD iso. Problem 1: It does not support networking natively in Thunar out of the box had to google to fix that problem. Solution is to install a missing package gvfs-backends. Problem 2: If you want perfect theme integration across all apps you can choose any theme as long as its "greybird". Any other theme causes some apps such as gnome-system-monitor to look like it was made for Windows 3.1 a quick google shows others have the same problem with other apps such as pavucontrol not theming either. Solution: No perfect solution yet. Problem 3: No central system settings area that I can find on the base install, I consider this a must have on any desktop OS regardless of how minimalist it is. Please note I said desktop OS, Im also a Slackware user and i am well used to doing without these niceties but when it comes to my desktop I expect simple that my late grandma could use. I use PCLinuxOS Phoenix (XFCE) Edition as my main os none of these are a problem, I can use any theme i like without any hardship. Control Centre is there for all your hardware configuration needs. Thunar networks out of the box. Memory usage from a cold start is 150 Mb with MySql and Vmware drivers running and 121MB without. Why on earth would i use Xubuntu? I would have to assume his readers are masochists as I like my OS to work pretty much out of the box, |
| caitlyn Dec 20, 2011 5:32 PM EDT |
Quoting:My problem with these "best distro for $DE" is that I have to wonder how many voters have actually tried multiple distros (with that same DE), so that they'd be in a position to make an informed vote.One of the reasons I suggested the page hit ranking on DistroWatch is the users of that site largely represent the membership of Distro Hoppers Anonymous. I write for DW now and again and if you ever read the comments section (difficult at times, I know) you certainly know that a lot of those people change distros the way most people change socks. Seriously, I've grumbled about the "perfect" this and "best" that articles for a long time. No offense to the author, but this 74 vote survey is about as accurate as a Ouija board. |
| Koriel Dec 20, 2011 5:48 PM EDT |
Im reminded of the /. byline on their polls This whole thing is wildly inaccurate. Rounding errors, ballot stuffers, dynamic IPs, firewalls. If you're using these numbers to do anything important, you're insane. |
| BernardSwiss Dec 20, 2011 9:18 PM EDT |
Easily fixed. Suggest to DarkDuck that he re-label his informal survey as an "informal survey", or if calling it a poll is important, that he call it a "straw poll". If you ask nice, he might even add a few caveats specifying that it's not a mathematically or methodologically reliable, formal study. Problem solved. :-P Not that I see it as a "real" problem, in the first place. I doubt any one (including DarkDuck) sees his blog -- or this poll -- as anything other than a fairly casual and informal presentation of Linux. That may be considered a fairly significant flaw, but it might also be considered its chief strength, and its principle charm. Perhaps that approach is not to many LXers' taste (at least not the vocal ones) -- but so what? |
| ComputerBob Dec 21, 2011 10:57 AM EDT |
@BernardSwiss - my suggestion is even simpler: change the LXer headline from Users voted for best XFCE-based Linux distribution to 74 users voted for best XFCE-based Linux distribution |
| darkduck Dec 22, 2011 10:53 AM EDT |
@ComputerBob:
Quoting:change the LXer headlinePointless. Whatever the number is before the "users", there will always be people who are not happy with poll results. Thus, they'll find any possible hole to say the poll is incorrect. How many voters would you consider as reliable source? 100? 200? 1000? 10000? Even then they will remember about firewalls, anonymizers, calculation methodology, users who only tried one and thus can't compare with others. Flaws are everywhere, if you are looking for them. That's why having "Users voted", "74 users voted" or "1074 users voted" won't make any difference. Low number of votes? There were about 260 people who read that particular post before the publishing date. Only 74 voted - others were not interested in voting. I can't make them, it's a free world. But maybe they don't have their preferred distro? Those who care - they showed their vote. In fact, this is my second poll. And there were exactly the same arguments about results of my first one about KDE. First round of results had 88 votes. There were lots of comments. Then I had new results published with 140 votes. Different number of votes, but still same leaders and even proportions between systems. It shows that statistics was correct even the first time. So, whether you trust figures or not - it is subjective topic, not mathematical one. |
| ComputerBob Dec 22, 2011 11:54 AM EDT |
@DarkDuck, 74 votes may seem significant to you, because your blog hardly gets any visitors, but IMHO, a poll with the results of 74 uknown visitors to an unpopular blog contains no information of value to me, and doesn't belong on a real news site like LXer. After I read your blog, I felt like you had wasted my time. I suspect that I'm not the only person who read it and then thought "That *might* be interesting information for the tiny group of people who regularly visit DarkDuck's blog, but why did he post it on LXer?" I fully understand that you don't care that I feel that way -- and that your standards for what should be considered "newsworthy" are much lower than mine. And that you apparently have no knowledge of, or interest in, the basic principles of statistics. Therefore, I won't make the mistake of reading your blog again. |
| caitlyn Dec 22, 2011 1:58 PM EDT |
Of course 1074 votes is more significant than 74. I take it you never took a statistics class, darkduck. |
| darkduck Dec 22, 2011 6:09 PM EDT |
@caitlyn:
In fact I had statistics class. The teacher there was one of the most boring in whole uni. But I still passed my exam. Of course, remember no formulas since then, because never used them since exam. I understand that 1074 votes would be more significant, but not necessarily with different results. That's the point I want to explain. |
| caitlyn Dec 22, 2011 6:22 PM EDT |
...and not necessarily with the same results. The fact is that you don't know. The fact is that I am no better informed for reading your blog post. |
| darkduck Dec 22, 2011 7:46 PM EDT |
... and I have already given you example where difference in proportion given by 88 and 140 voters is insignificant. |
| Fettoosh Dec 22, 2011 8:37 PM EDT |
I think these links should help you agree on something. Let's hope these are enough. :-) http://www.sciencebuddies.org/science-fair-projects/project_... http://www.surveysystem.com/sscalc.htm http://www.robertniles.com/stats/sample.shtml http://www.raosoft.com/samplesize.html |
| ComputerBob Dec 22, 2011 9:39 PM EDT |
I just finished a 4-day poll of everyone sitting at my desk. 100% of them will never read DarkDuck's blog again. Maybe I should post it as a news story on a really popular Linux news site, because - even though virtually no one ever visits my blog - I'm sure that the whole world will want to see the results of my really important survey. /sarcasm (again, for those who would otherwise misunderstand). |
| jacog Dec 23, 2011 7:25 AM EDT |
No CBob, it's just you being childish and rude. |
| montezuma Dec 23, 2011 8:23 AM EDT |
> Of course 1074 votes is more significant than 74. I take it you never took a statistics class, darkduck. Well since "everyone" else is being picky I would like to point out that this statement may also be incorrect. Reason being that if the large sample is not *representative* and the small sample is then the latter is more significant. This issue of representativeness is particularly acute in self selecting surveys that you find on the internet. Having said that my guess is that DarkDucks very small survey has also not been checked for representativeness and so is likely of very limited value. Happy Holidays Everyone. |
| tracyanne Dec 23, 2011 5:07 PM EDT |
Unbelievable. Are you people still arguing over this? |
| Jeff91 Dec 23, 2011 5:45 PM EDT |
Take this down if it is offensive... But whenever I see an online argument my mind goes right to this saying - http://farm2.staticflickr.com/1368/987251565_22ea2338dd.jpg ~Jeff |
| BernardSwiss Dec 23, 2011 7:18 PM EDT |
Up to this point I was manfully resisting the urge to post this link: http://xkcd.com/386/ |
| darkduck Dec 23, 2011 7:55 PM EDT |
montezuma wrote:Having said that my guess is that DarkDucks very small survey has also not been checked for representativeness and so is likely of very limited value.just proves my previous statement darkduck wrote:Even then they will remember about firewalls, anonymizers, calculation methodology, users who only tried one and thus can't compare with others. Flaws are everywhere, if you are looking for them. Enjoyed Bernard's joke. Did not enjoy Jeff's, what is usual for me. Parted from this discussion. |
| slacker_mike Dec 24, 2011 3:35 AM EDT |
@Jeff91: Your picture is offensive and in poor taste. You show a great deal of immaturity for someone who is the main developer of an Ubuntu spin. We get it you don't like darkduck, fine, but really you've got nothing better to contribute but a cheap gag at the expense of mentally challenged people? Arguing on the internet is far less productive than a mentally challenged person's sense of accomplishment in partaking in the Special Olympics. So carry on Jeff why don't you write another great article slagging off the Fedora IRC channel. |
Posting in this forum is limited to members of the group: [ForumMods, SITEADMINS, MEMBERS.]
Becoming a member of LXer is easy and free. Join Us!