Dennis Technology Labs: PC Total Protection Suites 2012

Correct, those sites are from w3school's website. That's why the whole demographic skew comes up. Statistics are skewed by the sample source and demographics exhibited by that sample source. I'd expect that the average unpatched grandma using IE7 would be less-likely to visit w3schools.com.

Do a test where you smell 100 people to see if they smell bad, for example. If your entire sample group consists of people who just finished fertilizing a field, chances are pretty good that the majority of them will stink. However at that point, saying "98% of people smell bad" would be bad work, since the sample source skews the results.

Just like "Most of the machines that come in for repair are unpatched" cannot be translated in any reliable way to the general population simply due to the fact that unpatched machines are more likely to need repair, while patched machines are not.

That location was given as an example. I do not intend to provide a comprehensive list of sources, since they are better found by performing web searches, however I can provide the following:

http://gs.statcounter.com/#browser_version-ww-monthly-201103-201203 lists March IE7 use as 2.91%. That is a global data range for every site that uses the Stacounter.com system for statistics.

I run two sites that are relatively low volume and specialized (which makes the demographic skew worse). Both use Google Analytics. On the art gallery site, IE7 is 5.56% of all -IE- users, and IE users are 26.5% of all visitors, for an end rate of 1.47% of all visitors using IE7. That is on a 1-year scale. Looking at the past rolling month specifically (and thereby a much smaller volume), IE is 31.66% of all users, and IE7 is 2.75% of all IE users, for a rate of 0.87% of all visitors using IE7 But due to the demographics of that site and the sample size being only 1337 visitors, I don't consider the sample or demographic to be doing it any favors.

The second site can be considered inaccurate due to granularity issues caused by an exceptionally small sample size (189 visits in the past rolling month), however for the sake of thoroughness, it recorded 31.22% IE use overall, and 1.69% of IE users at IE7. Literally, one of those 189 visitors was on IE7, for a rate of 0.529%. That data is next to useless though due to the low sample size.

Looking deeper, you could hit http://marketshare.hitslink.com/ where IE as a whole has a 53.8% market share. Dropping into the guts of that at http://marketshare.hitslink.com/browser-market-share.aspx?qprid=2&qpcustomd=0 shows IE7 specifically at a 4.48% market share. Curiously below IE6's 6.9% market share. That being said, when you dig much deeper, that site weights their stuff:
"For example, if our global data shows that Brazil represents 2 (percent) of our traffic, and the CIA table shows Brazil to represent 4 (percent) of global Internet traffic, we will count each unique visitor from Brazil twice. This is done to balance out our global data."

If you look at something like http://getclicky.com/marketshare/global/web-browsers/internet-explorer/ you even get to see some very interesting things: IE use is higher on weekdays and other browser use is higher on weekends on a very heartbeat-like graph. Business are prone to force the use of IE, which is liable to affect it, and businesses have been said to be the worst offenders at not patching or upgrading critical components.

The main consideration though is that no large sample group I found shows IE7 consistently over 5% of the userbase, and the majority of them show IE7 at around 2-3%. That's global internet use, not "Broken computers that come into the shop".

 

I have just been in contact with Webroot and I can assure you that the way the product works would lead us to run all necessary scans. This is because it walks the user through the process, rather than acting in the more traditional way.

As I wrote earlier, we can't cover every eventuality on how all these products work in the methodology. We have to make general rules and principles. And one principle is to act as a naive user. For this reason our WSC user would have run all scans.

 

This fully addresses and resolves my earlier reservations regarding Dennis Technology Labs' March 2012 tests.

Many thanks for looking into this. And many thanks for spending your weekend time to investigate the issue and post the result here.

Kind regards.

__

 

Dennis Tech Labs

Can another vendor (except Symantec) make it to the No.1 place?

Just wondering...

 

Simon, the number ten entry for Panda Global Protection 2012 should be revised...

 

With deep enough pockets, I don't see why not.

 

Thank you! I will ensure that typo is removed when I put out the revised report.
Simon

 

It depends what you mean by number one place. Kaspersky PURE scored the highest (just) in terms of the Protection Ratings. Kaspersky put no money into this test.

I know that some of you believe that the tests are biased in favour of the sponsor but I'll just state for the record that this is not so.

Simon

 

Good then, do consider to include Avast and Comodo next time to make the test more interesting

 

Wow now this is a really detailed test. (Because the sample size was small i guess but it was still very detailed)

 

it sure is

 

Thanks for going on the record.