Nov 25 2007

Profile Image of Mana
Mana

The Not-So-Solid Science of Technorati

Posted at 1:09 pm under Science and Technology

Salient and Why Don’t You Blog have been carrying a very interesting discussion on blogging tools such as Technorati and Blogrolling that atheists use, and their potential lack of effectiveness lately. I have also commented on this with the idea that a change from Blogrolling to another blog rolling services that would not be as overwhelmed may be necessary.

As a result of this discussion Salient looked at some Technorati data to see if any trends were visible there:

“Despite the fact that Planet Atheism seems as busy as ever, the graph for “Blog posts mentioning atheism per day for the past 30 days” does appear to show a major decline.”

English posts that contain Atheism per day for the last 30 days.
technorati atheism data for Oct. 27 to Nov. 25 2007

First, the graph I pulled today from Technorati looks different than the data given to Salient Technorati yesterday. I blame this one on Technorati and the not-so-exact science of web statistics.

Also, I would love to see a comparison of this set of data to same period last year. Because technorati doesn’t give us data from last year, I decided to look at a 180 day period rather than just 30 days.

English posts that contain Atheism per day for the last 180 days.

technorati atheism data for May 30 to Nov. 25 2007

So, the great news is that there was a huge spike in atheist-themed blog posts that Technorati identified during the mid Sept. and mid Oct. time-span, which dwarfed the periods before and after. So I agree with Salient that the current decline is probably related to the Holiday season, but it also seems to be a return to statistics more in line with the pre-spike periods. But when I look at usage of word ’science’ I’m seeing a drop in stats similar to ‘atheism’, whereas Salient is seeing  an increase in the usage of science. So is Technorati making this data up on the fly?

How relevant is Technorati data? Here’s a review of the past 6-months of Technorati history:

I’ll let my readers decide on their own how much credence to put on technorati data at this point in time. It may be that a few months from now, when they’re done with data center changes and “economizing” we’ll see an increase in quality. Until then we can follow the advice on the Technorati troubleshooting page, write to them and ask for changes (which apparently has happened a lot lately), and/or we can find a better service.

7 responses so far

7 Responses to “The Not-So-Solid Science of Technorati”

  1. heatheron 25 Nov 2007 at 1:36 pm 1

    That is some great research you’ve done.

    It looks as if it correlates with the university terms doesn’t it?

    Students and lecturers using the academic facilities to blog, but taking summer and mid-term breaks.

  2. TWon 25 Nov 2007 at 1:49 pm 2

    Interesting post. It motivated me to go back and re-check the traffic logs on WhyDontYou (actually, I used feedburner. I am too lazy to FTP into the site, download the log files and analyse them myself…).

    While it is no where near as significant, the blog did get a spike in traffic late September / early October. There has been a monumental decline in traffic since the middle of November though (around half as many hits as we were getting in August / early September and about a quarter of the hits from late September).

    On a related note, I remember heather pointing out to me that the blogroll was going mental early September - the “Most recent” 25 blogs had a lifespan of about two hours from start to finish, while over the summer it could be days between posts before a blog was pushed off. Right now there is around 4 hours between oldest and newest on the blog, so it could well be that all these crests and troughs are based on school / college / university terms.

  3. salienton 25 Nov 2007 at 2:27 pm 3

    I think that the discrepancy in the graphs may reflect the time period covered. I noticed that the on-site graph and the graph whose html I ‘pasted’ looked different — presumably for this reason. However, it was late and I didn’t look closely.

    Interesting research! Your list of behind the scenes changes is interesting. I think that you have explained why the T search engine kept crashing.

  4. Manaon 25 Nov 2007 at 3:55 pm 4

    Just noticed WDYB is listed as new on the Atheist Blogroll, hurray! Something must have gone right today.

    Yes, I’m sure web traffic correlates with season, weather, holidays, school etc. And on top of that Technorati data depends also to their glitches…

    Salient, yes it’s odd, I notice too that the graph on T site looked different than the one linked through the html code they provide… As I said, how can we know which graph is right?

  5. Lugosion 06 Dec 2007 at 1:49 pm 5

    One problem I have with such measures is that they focus on specific words, in this case “atheism” or “atheist.” Quite honestly (though I’m NOT willing to swear on a Bible), I don’t think I’ve ever used either of those words on my site (aside from the Blogroll itself). But I think I’ve increased the percentage of my posts that criticize Intelligent Design, question the Catholic Church moral authority, and otherwise defend thought & reason.

  6. Manaon 06 Dec 2007 at 7:53 pm 6

    Lugosi, it is hard to say what a word such as “religion” tells us about who writes on religion. It could be both religious folk, or atheists. So you’re right. And in fact, web reporting is a very inexact science.

  7. batu bata presson 20 Nov 2013 at 1:15 am 7

    Hello Dear, are you actually visiting this site regularly,
    if so then you will without doubt get fastidious know-how.

Trackback URI | Comments RSS

Leave a Reply