This has been around for quite a while, first as an Advanced Search feature in 2010, and then as a sidebar feature.

In the March 2011 update thread [webmasterworld.com...] in a March 24, 2011 post, tedster mentioned it as one of the factors he was watching....

Some other on-page areas I'm exploring:

1. Reading level (Google rolled this information into the SERPs after they started work on Panda this past year)

2. Semantic richness, as detailed in the phrase-based indexing patents

3. Spelling and grammar.

There's also this discussion, among others...

Reading Level and Panda
Apr 19, 2011
http://www.webmasterworld.com/google/4300397.htm [webmasterworld.com]

I think that reading level needs to be appropriate to the kind of query vocabulary you're targeting. I'm sure it also ties in with user demographics.

I just checked numerous sites, not all my own, and found very consistently that sites with 'advanced' reading levels of more than a few % had been pandalised.

More curiously, part of one of the sites is a 'mash up' with about 40000 pages, each containing very similar text and identical reading level yet google had attributed the reading level for the pages as basic: 59%, intermediate 31% and advanced: 10%.

So if they are using their reading level as part of panda it's not using very accurate data.

Interesting find. It's a bit of an insight at how well G can tell a pages content, beyond topic.

FYI I have 0% advanced (which is a supprise being a tech site) and ive been Panda slapped.

Cool, many thanks. I have never noticed this.

I just checked my site -
Basic - 14%
Intermediate - 60%
Advanced - 26%

Is that bad or good? Is there some kind of ruler?
Should I dumb it down? Smarten it up?
I don't know. Can guidance be found somewhere?

BTW, I was not Pandaized, so I am not sure about the "few %" comment.
.

Sally, I understand that lots of sites were not judged at all by panda for various reasons - eg site age and size, subject matter, commercial intentions of the site and appearance in serps for commercial terms might all have helped decide which sites were assessed.

So having a high 'advanced' rating for some sites not affected by panda is quite predictable.

In any case reading level would presumably only play a part in the overall calculation. I'm not personally planning on dumbing-down 500 articles in an effort to escape panda.

It is interesting to consider whether a machine trying to judge the quality of a page would conclude that an article with long sentences, excessive punctuation and lots of long words (presumably the main factors that constitute 'advanced reading level') would in all probability also be a poor quality article, for example because spam, spun and auto-translated page share the same attributes.

I do not think there is a magical reading score that will work for every site. I do think that most consumers sites should create pages for the basic & intermediate levels since over 60% of the US population do not have a college degree.

Figure out who your target market is and create your content for them. If you are talking to scientists then it probably makes sense to have pages with higher reading levels due to the technical jargon. If your targets teenagers then you probably want pages with a lower reading score.

Here are a few reading scores for site: searches using Google sites. I think this helps to show there is not just one magical reading score.

site:Google.com
Basic 33%
Intermediate 51%
Advanced 16%

site:zagats.com
Basic 67%
Intermediate 32%
Advanced < 1%

site:googleblog.blogspot.com
Basic 29%
Intermediate 70%
Advanced < 1%

site:mattcutts.com
Basic 72%
Intermediate 28%
Advanced 0%

(As per the TOS please do not post private sites. I used these Google sites to provide some examples to discuss)

I don't know if it does but it could,
Is reading level included in the algorithm?

It should make sense to provide search results related to search terms using the same rating for the search terms, therefore matching both avg.

Figure out who your target market is and create your content for them

Good advice, but we should be all doing that anyway, regardless of whether Google cares about it or not.

It is interesting to consider whether a machine trying to judge the quality of a page would conclude that an article with long sentences, excessive punctuation and lots of long words (presumably the main factors that constitute 'advanced reading level') would in all probability also be a poor quality article, for example because spam, spun and auto-translated page share the same attributes.

You could also conclude that grammatically correct complex sentences, and a consistently complex style are likely to be the work of a competent human author.

I would concentrate on writing well for your audience, which will influence the number of incoming links, which we know is a key factor, rather than guessing what Google wants.

could also conclude that grammatically correct complex sentences, and a consistently complex style are likely to be the work of a competent human author.

On my planet a competent editor will not tolerate needless complexity.

A good editor will make prose easier to read and understand -- without changing what it means.

...sites with 'advanced' reading levels of more than a few % had been pandalised.

Not true with my site. I saw no loss with Panda, in fact we gained SERP placement for many pages. However, this may be because of authoritative content as hundreds of .edu's link to our pages.

Basic 30%
Intermediate 65%
Advanced:5%

On my planet a competent editor will not tolerate needless complexity.

Needlessly is the key word. Sometimes long words or sentences are needed. One long word can replace a short phrase. Sometimes it can be easier to read one long sentence that several short ones.

It depends on your audience, and on what you have to say. There are good reasons why,for example, academic papers are hard to read according to tests like this, even if they are well written and easy for their target audience to read (of course some are badly written).

My site is 1% basic and 45% advanced, but was not affected by Panda. My feedback from visitors has consistently been that it is well written and easy to understand.

I think the key issue in this is that Google must have sufficient objectivity to realise how bad they are at accurately assessing the quality of the written word.

Consider Google translate – a project entirely dedicated to language and it rarely returns a translated sentence that reads correctly to human eyes - more often that not it is just total nonsense.

Testing Google’s reading level assessment this morning with sites that use spun articles that really are so bad as to be unreadable, Google returns many results that are mainly intermediate and advanced levels. Why? Possibly because vocabulary must be part of the assessment algo and because spinning software searches for opportunities to replace words and inevitably uses less common synonyms, this exotic vocabulary may translate as an advanced reading level.

Google must know they can’t do this very well and whilst they certainly may continue to test and improve and integrate such emergent techniques, it doesn’t make sense for this metric to figure very strongly in the algo just yet.

Creating content for a general level would not seem important. As with many other factors I assume that there could be relevance between the vertical, keyword phrase and the "reading level".

For example, if the audience is researchers and the keyword phrase pertains to research intended for academics, the reading level should be advanced. If I were analyzing content as a search engine I'd find low reading levels for content of this type to be suspicious.