Hi, I realized today that surprinsingly the Russian Wikiversity is better ranked by Alexa than the English version (ru.wikiversity.org = 53.66% of wikiveritsy.org´s traffic, en.wikiversity.org only 33.28%). The description says: "(...) it is relatively popular among users in the city of Velikiy Novgorod (where it is ranked #121). While roughly 11% of visitors to Wikiversity.org come from Russia, where it is ranked #6,106, it is also popular in Algeria, where it is ranked #4,570." http://www.alexa.com/siteinfo/wikiversity.org#
On the other hand, according to Wikimedia stats, the English Wikiversity site has much more Views/hr (4,068) than the Russian one (271) and it also has much more articles (15,718 vs. 1,917). http://stats.wikimedia.org/wikiversity/EN/
So, I wonder what´s the reason for this discrepancy. Does anybody have an idea? Best,
René
Hi, I realized today that surprinsingly the Russian Wikiversity is better ranked by Alexa than the English version (ru.wikiversity.org = 53.66% of wikiveritsy.org´s traffic, en.wikiversity.org only 33.28%). The description says: "(...) it is relatively popular among users in the city of Velikiy Novgorod (where it is ranked #121). While roughly 11% of visitors to Wikiversity.org come from Russia, where it is ranked #6,106, it is also popular in Algeria, where it is ranked #4,570." http://www.alexa.com/siteinfo/wikiversity.org#
On the other hand, according to Wikimedia stats, the English Wikiversity site has much more Views/hr (4,068) than the Russian one (271) and it also has much more articles (15,718 vs. 1,917). http://stats.wikimedia.org/wikiversity/EN/
So, I wonder what´s the reason for this discrepancy. Does anybody have an idea? Best,
René
Hunger for learning in Russia and Algeria. English speakers have many other opportunities. Just a guess though... I'm sure peer group communication, or lack of it, is mainly responsible.
Fred
On Tue, May 31, 2011 at 4:26 AM, René König kontakt@renekoenig.eu wrote:
On the other hand, according to Wikimedia stats, the English Wikiversity site has much more Views/hr (4,068) than the Russian one (271) and it also has much more articles (15,718 vs. 1,917). http://stats.wikimedia.org/wikiversity/EN/
So, I wonder what´s the reason for this discrepancy. Does anybody have an idea? Best,
Alexa statstics can be easily manipulated to a degree. The discrepancy could be as simple as having one or two users from Russia who have the Alexa toolbar installed from Russia who visit regularly every day, whereas the americans who visit don't have Alexa installed. (Because the population that tends to have the Alexa toolbar installed tends to be relatively self selecting in terms of certain industries.)
Alexa data should always be taken with a huge grain of salt. I was able to push my own site up to the 2,500th most popular site in Australia meerly by visiting it every day.
Laura Hale, 30/05/2011 21:35:
Alexa statstics can be easily manipulated to a degree. The discrepancy could be as simple as having one or two users from Russia who have the Alexa toolbar installed from Russia who visit regularly every day, whereas the americans who visit don't have Alexa installed. (Because the population that tends to have the Alexa toolbar installed tends to be relatively self selecting in terms of certain industries.)
Alexa data should always be taken with a huge grain of salt. I was able to push my own site up to the 2,500th most popular site in Australia meerly by visiting it every day.
+1
René König, 30/05/2011 20:26:
On the other hand, according to Wikimedia stats, the English Wikiversity site has much more Views/hr (4,068) than the Russian one (271)
Wikimedia page view stats are not approximate; this simply proves that Alexa is wrong.
Nemo
Thanks everyone! I knew that one should be careful with Alexa but I didn´t expect it could be THAT wrong. It´s a pity because otherwise it could be a convenient and interesting tool. The next question is: is this rather an exception or is Alexa in general not reliable? Maybe I shouldn´t use it for academic papers at all. Best,
René
Am 31.05.2011 08:57, schrieb Federico Leva (Nemo):
Laura Hale, 30/05/2011 21:35:
Alexa statstics can be easily manipulated to a degree. The discrepancy could be as simple as having one or two users from Russia who have the Alexa toolbar installed from Russia who visit regularly every day, whereas the americans who visit don't have Alexa installed. (Because the population that tends to have the Alexa toolbar installed tends to be relatively self selecting in terms of certain industries.)
Alexa data should always be taken with a huge grain of salt. I was able to push my own site up to the 2,500th most popular site in Australia meerly by visiting it every day.
+1
René König, 30/05/2011 20:26:
On the other hand, according to Wikimedia stats, the English Wikiversity site has much more Views/hr (4,068) than the Russian one (271)
Wikimedia page view stats are not approximate; this simply proves that Alexa is wrong.
Nemo
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Thanks everyone! I knew that one should be careful with Alexa but I didn´t expect it could be THAT wrong. It´s a pity because otherwise it could be a convenient and interesting tool. The next question is: is this rather an exception or is Alexa in general not reliable? Maybe I shouldn´t use it for academic papers at all. Best,
René
As in physics the nature and limitations of any measuring tool and concept is part of your research. Personal use of the Alexa toolbar affects measurements only when what one, or a few, persons do matters in the statistics.
Fred
Am 31.05.2011 08:57, schrieb Federico Leva (Nemo):
Laura Hale, 30/05/2011 21:35:
Alexa statstics can be easily manipulated to a degree. The discrepancy could be as simple as having one or two users from Russia who have the Alexa toolbar installed from Russia who visit regularly every day, whereas the americans who visit don't have Alexa installed. (Because the population that tends to have the Alexa toolbar installed tends to be relatively self selecting in terms of certain industries.)
Alexa data should always be taken with a huge grain of salt. I was able to push my own site up to the 2,500th most popular site in Australia meerly by visiting it every day.
+1
René König, 30/05/2011 20:26:
On the other hand, according to Wikimedia stats, the English
Wikiversity
site has much more Views/hr (4,068) than the Russian one (271)
Wikimedia page view stats are not approximate; this simply proves that Alexa is wrong.
Nemo
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
On Tue, May 31, 2011 at 12:36 PM, Fred Bauder fredbaud@fairpoint.net wrote:
As in physics the nature and limitations of any measuring tool and concept is part of your research. Personal use of the Alexa toolbar affects measurements only when what one, or a few, persons do matters in the statistics.
That's not really true - apart from the statistical error, there are also systematic errors, and those remain when a single person is insignificant enough not to influence the statistics. For example, it used to be (I haven't checked recently) that Korean files came out relatively high in the Alexa rankings - presumably because the Alexa toolbar usage was high in Korea.
Am 31.05.2011 12:36, schrieb Fred Bauder:
Thanks everyone! I knew that one should be careful with Alexa but I didn´t expect it could be THAT wrong. It´s a pity because otherwise it could be a convenient and interesting tool. The next question is: is this rather an exception or is Alexa in general not reliable? Maybe I shouldn´t use it for academic papers at all. Best,
René
As in physics the nature and limitations of any measuring tool and concept is part of your research.
Right, the problem is just that Alexa is not "my" research tool and therefore it not transparent to me what it actually does. The same goes for other tools like Google Trends, BTW. Of course, it´s understandable that these companies do not provide much information about the methodology but as a researcher that´s kind of frustrating. R.
Personal use of the Alexa toolbar affects measurements only when what one, or a few, persons do matters in the statistics.
Fred
Am 31.05.2011 08:57, schrieb Federico Leva (Nemo):
Laura Hale, 30/05/2011 21:35:
Alexa statstics can be easily manipulated to a degree. The discrepancy could be as simple as having one or two users from Russia who have the Alexa toolbar installed from Russia who visit regularly every day, whereas the americans who visit don't have Alexa installed. (Because the population that tends to have the Alexa toolbar installed tends to be relatively self selecting in terms of certain industries.)
Alexa data should always be taken with a huge grain of salt. I was able to push my own site up to the 2,500th most popular site in Australia meerly by visiting it every day.
+1
René König, 30/05/2011 20:26:
On the other hand, according to Wikimedia stats, the English
Wikiversity
site has much more Views/hr (4,068) than the Russian one (271)
Wikimedia page view stats are not approximate; this simply proves that Alexa is wrong.
Nemo
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
On Tue, May 31, 2011 at 9:15 PM, René König kontakt@renekoenig.eu wrote:
Right, the problem is just that Alexa is not "my" research tool and therefore it not transparent to me what it actually does. The same goes for other tools like Google Trends, BTW. Of course, it´s understandable that these companies do not provide much information about the methodology but as a researcher that´s kind of frustrating.
The methodology is pretty straightforward. Users install the Alexa toolbar. They complete a survey that includes basic demographic data at the time of the install. Based on the IP address and the survey, the location that a user is from is determined. Once the toolbar is installed, the sites that the user visits are recorded and the amount of time the user spends on the website is recorded. This data is than compiled on a daily basis and global and national rankings are compiled. Other demographic data is also compiled for each site.
Problems with the methodology include that users with mobile devices are not included on Alexa's data and Safari browser users are not included. (Mac users can install the Alexa toolbar, provided they use Firefox. Historically, they have not been collected.)
Certain countries have smaller populations so the total rankings for the top sites in those countries is smaller. In the case of New Zealand, only the top 20,000 sites are listed. In the case of Australia, only the top 50,000 sites are listed.
Certain populations are more predisposed towards using Alexa than other populations. Internet marketers, website maintainers, researchers, social media professionals may be more likely to use it than other people because these parties have a vested interest in improving their website ranking. Thus, a higher ranking on Alexa than say Google Analytics indicates a site should have could be a result of selected population visiting.
Laura, thanks for the useful information! You have one more follower on Twitter :) Best,
René
Am 31.05.2011 23:30, schrieb Laura Hale:
On Tue, May 31, 2011 at 9:15 PM, René König <kontakt@renekoenig.eu mailto:kontakt@renekoenig.eu> wrote:
Right, the problem is just that Alexa is not "my" research tool and therefore it not transparent to me what it actually does. The same goes for other tools like Google Trends, BTW. Of course, it´s understandable that these companies do not provide much information about the methodology but as a researcher that´s kind of frustrating.The methodology is pretty straightforward. Users install the Alexa toolbar. They complete a survey that includes basic demographic data at the time of the install. Based on the IP address and the survey, the location that a user is from is determined. Once the toolbar is installed, the sites that the user visits are recorded and the amount of time the user spends on the website is recorded. This data is than compiled on a daily basis and global and national rankings are compiled. Other demographic data is also compiled for each site.
Problems with the methodology include that users with mobile devices are not included on Alexa's data and Safari browser users are not included. (Mac users can install the Alexa toolbar, provided they use Firefox. Historically, they have not been collected.)
Certain countries have smaller populations so the total rankings for the top sites in those countries is smaller. In the case of New Zealand, only the top 20,000 sites are listed. In the case of Australia, only the top 50,000 sites are listed.
Certain populations are more predisposed towards using Alexa than other populations. Internet marketers, website maintainers, researchers, social media professionals may be more likely to use it than other people because these parties have a vested interest in improving their website ranking. Thus, a higher ranking on Alexa than say Google Analytics indicates a site should have could be a result of selected population visiting.
-- twitter: purplepopple blog: ozziesport.com http://ozziesport.com
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
On Tue, May 31, 2011 at 8:21 PM, René König kontakt@renekoenig.eu wrote:
Thanks everyone! I knew that one should be careful with Alexa but I didn´t expect it could be THAT wrong. It´s a pity because otherwise it could be a convenient and interesting tool. The next question is: is this rather an exception or is Alexa in general not reliable? Maybe I shouldn´t use it for academic papers at all. Best,
I am using Alexa data in my dissertation and I think it can be reliable for some things. It is just an issue of understanding how Alexa gets its data, and then seeing if the results are similar on compete and Quantcast. In some cases, you don't really have much of a choice but to use Alexa. For Australian and New Zealand website rankings, I can't think of another site that provides publicly available free data. I've used it in a blog post at http://ozziesport.com/2011/02/new-zealand-sport-web-traffic-in-response-to-t... for general event tracking, it can be useful to see how events impacted a site's traffic amongst the group that uses it.
I just wouldn't use Alexa data in isolation involving other data. I used it in http://ozziesport.com/2011/01/derryn-hinch-journalist-traffic-versus-wikiped... has broken images) in a bit that I considered using for my dissertation chapter to try to guess at the amount of web traffic that a journalist got. So yeah, Alexa data can be used and in some cases should be used but against a backdrop of other information.
Sincerely, Laura Hale
wiki-research-l@lists.wikimedia.org