Lexiles in LibraryThing?

For a while now, LibraryThing members have asked for Lexiles. We are exploring adding them, in conjunction with the company that makes them.

Here are some threads about it:
http://www.librarything.com/topic/1260
http://www.librarything.com/topic/41263
http://www.librarything.com/topic/67550
http://www.librarything.com/topic/37242

Can I get opinions on:

1. Do members expect to be able to EDIT the Lexile level? It seems to me that you shouldn't want to--that the 100k-odd ISBNs they've evaluated are the only ones that should have them, and the works attached to them.
2. Apart from seeing them in your catalog, do you want pages showing all works with a given Lexile? What do you expect here?

Any other ideas?

I'd love to know how you decide which features to pursue at which times.

Randomly, of course.

Reading most of those threads it seems many members would positively oppose the inclusion of Lexiles...

I think that the often restrictive use of Lexiles is harmful to young readers rather than helpful, but I also support the right for others to make their own decisions, and I do think that including Lexiles helps LibraryThing's business model. Members will probably want to edit Lexiles, but there should be some sort of icon or other indication to differentiate "official" ratings. For those who value the system, I think that pages showing groupings would probably be helpful, and the ability to search by combinations of tags and Lexiles would probably be good too. (e.g., Johnny loves baseball, and Susie likes vampires.)

For those (like me) wondering what the heck a lexile is:

The Lexile Framework for Reading is an educational tool that uses a measure called a Lexile to match readers of all ages with books, articles and other leveled reading resources.

Recognized as the most widely adopted measure of reading ability, more than 28 million Lexile measures are reported from reading programs and assessments annually. Thus, about half of U.S. students in grades 3rd through 12th receive a Lexile measure each year. Lexile measures are being used across schools in all 50 states and abroad.

-- wikipedia

Lexiles are very commonly used in education in the US. Adding Lexiles will be a great help to all the school-room libraries and small-school libraries on LT. Others won't care, I think.

About 20% of LT books (books, not works) have Lexiles. School-ish libraries on LT go much higher, of course.

ISTR there have been other threads about reading level measures for books also with strong opposition.

Some points.

1) Lexile is a primarily a USian thing isn't it?
2) There are competing schemes. If one of the other schemes come knocking will they have to be added.
3) There are criticisms of Lexile scores - and if one looks at some individual volumes valid ones.
4) Many books, even those suitable for school-age readers, will not have a lexile score. So the over-reliance on having a score is limiting.

Opinion of one non-US user:

If US schools want them, and you can provide them, then go ahead. It won't affect me one way or the other, but I think anything that makes LibraryThing more attractive to libraries is generally a good idea.

And from another non-US user:

They sound like the spawn of the devil to me and I'd have thought that you would be actively campaigning against their use.

I am with EveleenM - if that will help the US-based people that care about them - why not. As long as they do not change the things for the people that do not care about them.

Apart from seeing them in your catalog, do you want pages showing all works with a given Lexile?

I don't care about Lexiles in particular, but I think you should generally be in the habit of making cumulative pages for all data. So yes.

Um..... Would I throw a fit if Lexiles were added to LibraryThing data? Probably not.

Do I think Lexiles are important data to have? No, I don't. Do I think they are stupid and wrong and potentially harmful, given the ways that many schools/libraries use them? Hell yes.

In the threads liked in the OP, it does seem like a lot of users share that opinion and wouldn't like having Lexiles on the site. I understand that if schools already use them, having the data here on LT would be a benefit.... But I'm concerned that schools/libraries that *don't* already use them might "take the suggestion" and start using them, if the data is easily available here. And imo, anything that encourages people to use Lexiles is a bad thing.

I dislike when lexiles are used to restrict reading, but don't oppose their mere existence. LT displaying lexiles will not affect how schools use lexiles.

This is not a high-priority addition for me.

While not personally in favour of such systems, I recognize their existence. So if it helps school libraries - and probably youth sections of public libraries - to make better use of LT: go for it.

Re #1: Do members expect to be able to EDIT the Lexile level? It seems to me that you shouldn't want to--that the 100k-odd ISBNs they've evaluated are the only ones that should have them, and the works attached to them.

The advantage would be that libraries in other parts of the world - where similar but different reading level systems are used - could easily adapt the system to their own use.

Apart from seeing them in your catalog, do you want pages showing all works with a given Lexile? What do you expect here?

I would expect people actually using such systems would also want to see them indicated on work pages somehow.

I'm for information about books. More of it!

A page with "lexiles" and lexile-type data might work comfortably with reviews and subject headings and classifications.

I like the idea of having lexile data show on the page with reviews, to show that it is a judgment about the book, rather than an objective description.

17>

I like the idea of having lexile data show on the page with reviews

I don't. I like the idea of having it show up on a separate page, so that people who are interested can see it for books where it's meaningful rather than having people who don't care need to see it (or a blank field) for books where it's meaningless every time they want to see a review.

Oh come on. Tim's certainly not going to create a new page to hang off the work page that only has the lexile on it. The page Tim was talking about in post #1 was a summary page that applied to multiple works.

18>

The page Tim was talking about in post #1 was a summary page that applied to multiple works.

I don't have a problem with that. I just don't see why reviews are the right place to put this data that applies only to a small fraction of works and a small fraction of users.

Take this with a grain of salt, since I don't know much about how Lexiles work....

But here on LT, multiple editions can be in one "work", so it makes little sense (imo) to attach Lexiles at the work-level. Especially attaching it to the main work-page. Different editions may have different Lexile ratings or not have been assigned any at all. Same editions, according to LibraryThing, include differences like extra footnotes, introductions, information, etc, that could easily change a Lexile-rating. I don't see how it would make sense to put the rating on the main work-page when it might only apply to like 25% of the editions in that work.

Different editions may have different Lexile ratings or not have been assigned any at all.

Is that true? My impression is that lexiles apply to children's books and that it applies to the main text. If the main text changes enough to change the lexile value then I would expect it not to be the same work.

Ditto to #22. Bowdlerized, significantly abridged, and "for children" editions are not the same works as the adult literature from which they are derived.

I don't know about lexiles, but many children's books that were written initially for children still get simplified in more recent editions. I'm not at all convinced that these automatically constitute new works.

First, they're not just assigned to books for your children. I've got books with Lexiles that are marked as for graduate students only. True, the center is lower, but it's more grade-school and junior-high, I think.

I've looked at the question of different Lexiles on different editions of the same work. In general, the numbers are the same or very close. Very rarely there will be a big gap but I think this always comes from a work being combined with simplified-English edition, or something rather different. But anyway, I'm applying at both levels, with a distinction between the two--green text for work-level ones and black for ISBN-level ones.

Lexile scores are the corporate property of MetaMetrics. Is that company going to charge LT to publish their scores or as unlikely as it sounds pay LT to publicize them? I believe that there is an equivalent open source scoring system called ATOS that correlates to the Advanced Reading System. Will they both be published?

Given that was substantial authorial opposition to having their works 'graded' and the wide range of reading abilities that children of age x display. I'd be against the entire idea. IF LT were to be complete with every other single feature installed, finished, working properly in all brouser vairents, and every single bug squished, then and only then, might it be worth devoting even a few neurons of staff time to this feature, but I can still probably think of more ponies that would be more worth doing.

Another perspective: I've got a son who is reading far beyond his age level. It sounds like Lexile scores would be helpful for me when choosing books for him.

(Of course, I read the books my self and wouldn't be restricted to those scores.)

#28

Except that it doesn't always work for example the wikipedia page mentions
"The Library Mouse" by Daniel Kirk, is a 32-page children’s picture book rated by Amazon.com as “for ages 4-8” and has a Lexile score of 830.
Twilight by Stephenie Meyer has a Lexile score of 720.
Jurassic Park by Michael Crichton only has a score of 710.
Ramona Quimby, Age 8 by Beverly Cleary has a score of 860
The Wind in the Willows has a score of 1140.
The Lion, Witch And The Wardrobe has a score of 940.
The Magician's Nephew has a score of 790
The Fellowship of the Ring is 890 and The Two Towers is 810
The Hobbit is 1000.

I think those scores don't give me confidence in the algorithm used.

Also the lexile measure quite specifically doesn't measure suitability of content (for example sex or objectionable violent scenes) or non-textual complexity so shouldn't be used purely to filter in/out suitable books.

It's not intended to filter suitable reading. It is strictly measuring reading level - as in reading difficulty. How difficult is the vocabulary? How many multi-syllabic words?

If you know your child's reading level - how many words he/she recognizes - you can find which books will be too easy to maintain interest, which books will challenge and strengthen the reading.

Content suitability is quite another matter. It is subjective, usually, and depends on the parent and/or teacher knowing the maturity of the individual child.

It's totally puzzling that The Hobbit should have a distinctly higher score than The Two Towers.

The Wind in the Willows, on the other hand, probably deserves a score exceeding all of the others listed in #29.

29>

Those numbers look like they're chosen completely at random, even if we throw out Wind in the Willows as a freak outlier.

Just for curiosity: is the higher or the lower score supposed to indicate easier reading?

> 31/32: Maybe that system also takes the use of archaic - or otherwise unusual - words into account?

The lower the number the easier it is to read.

I am positive that there is going to be some algorithm (trade secret though unlike ATOS that TLCrawford mentioned) behind it all depending on sentence length, word length, number of adjectives and adverbs etc.

#34

Possibly but there are oddities all around

Treasure Island = 1100
War of the Worlds = 1170
Robinson Crusoe = 1320
The Hound of the Baskervilles = 1090
Adventures of Huckleberry Finn = 990
Picture of Dorian Gray = 920
On The Beach = 780
Grapes of Wrath = 680

I would imagine that sentence length hits Robinson Crusoe very badly.

I think that Lexile is calculated on a section of the text and not the full text. How that works with books which have differing sections in terms of textual complexity I don't know.

This would be useful for people, it's a quick add, throw it in. Though as above you are going to want to consider having more than one line for more than one system.

I've never understood the complete objection peole have about this categorising books by age. Yeah I wouldn't have liked to be restricted to books of only my age, but it was helpful for my aunts and uncles to be told that 8yr old Lindsay reads books at a 12-14yr old level, or 12yr old Gary reads at 7yr level (because he's dyslexic). They are only guidelines for adults to chose books they know nothing about for children.

(1) Agree with 37. It's useful to teachers and parents to have someone look at works and evaluate them on difficulty of reading. Of course, like most readers and writers, we are opposed to the idea of using these kinds of things prescriptively -- "You're only 7yo, so you can only read these books!" -- but I question whether the lexiles are used that way.

(2) 27 > With respect, different authors like or dislike all kinds of things, and it's really not incumbent on us to do catalog according to the minimum set of information that all authors agree on. Some authors don't like libraries circulating their books. Some authors don't like particular critical interpretations of their works. Some authors don't like the subject headings assigned by the Library of Congress. Some don't like having their work excerpted for study. Many authors don't like particular covers or illustrators assigned to their works.

Nevertheless, it is the job of "catalogs" to operate for *users*, to assist users in locating materials. Some people like to locate materials according to "reading level", i.e., size of words, complexity of sentences, etc. I apply something like the Sony-Betamax test to cataloging data: If there is any useful, non-censorious purpose for a type of cataloging data, then I say, go for it.

(3) I would certainly hope that LibraryThing would not exclusively ally itself with a single proprietary rating system. I would in fact expect that if there's an open source / open content of cataloging data of *any* sort, then LT would use it -- so up with ATOS.

38> And, as we've seen from Tim's postings, some authors demand their books be de-listed from LT!

34>

I'd assume it does, but that doesn't explain The Hobbit ranking as significantly "harder" than Lord of the Rings, the huge difference between the two Narnia books listed, or Ramona ranking as "harder" than Jurassic Park.

really? delisted? I've never come across that. Hilarious.

41> Well, there was some question over whether or not they actually understand LibraryThing was a cataloging site and not a pirate book downloading site. But yes, they did.

There seems to be a correlation to the works age but there are other factors at work. I wonder what Hemingway or Stephen Donaldson would score.

1719, Robinson Crusoe = 1320
1883, Treasure Island = 1100
1885, Adventures of Huckleberry Finn = 990
1890, Picture of Dorian Gray = 920
1898, War of the Worlds = 1170
1901, The Hound of the Baskervilles = 1090
1937, The Hobbit =1000
1939, Grapes of Wrath = 680
1950, The Lion, the Witch and the Wardrobe =940
1954, The Fellowship of the Ring =890
1954, The Two Towers =810
1957, On The Beach = 780
1990, Jurassic Park =710
2005, Twilight =720

#43

The Sun Also Rises is 610
A Farewell to Arms is 730 although other Hemingway books are a bit higher.
The left hand of darkness is 970
Crime And Punishment is 990 (although I have also seen 850)
Count of Monte Cristo is 930
Harry Potter and the half-blood prince is 1030
Around the world in eighty days is 1070
The Jungle Book is 1140
Equal Rites is 880
And Then There Were None is 570
The Songs Of Distant Earth is 1420

It just seems so useless - a suitable book may be coming from a very wide point range.

The entire Lexile range is supposed to range up to about 2000. The only thing I have seen above 1500 is non-fiction.

As these are relatively neutral reading comprehension expressions (in contrast to the usually warped content ratings), I like to see them included.

Playing around on their website, I see their aspiration level as rather undemanding (Julius Caesar's Caesar's The Civil War in English is rated 1570L, way out of recommended 12th grade reading. In a classical education, it is rather common to read the (unrated) Bello Gallico in its Latin version. Joseph J. Ellis' His Excellency, George Washington with 1450L, perhaps as a punishment for the author's lies about his military service, is also way out of student reach.)

Please display it on a page only if you actually have a rating. Otherwise all pages acquire another neglected broken window element.

For their use, I see a combined tag and lexile range search page as the best option:
Display books tagged "animals" in the Lexile range 600 to 800

Did I miss an announcement about this? I notice there's now Lexile information showing in the zeitgeist.

If there was an announcement, I missed it too.

Me too.

Just in the Zeitgeist. I'll probably pull it back soon. Anyway, the data's all in; it's a question of displaying it a few places.

Did this ever get added? I would love Lexiles on my Library!

Yes, there is a field for Lexiles when you edit your book's details.

Crime And Punishment is 990 (although I have also seen 850)

For translated works, presumably different translations will have different measures because one translator happened to use more archaic words or the like?

IIUC, the system is only for English-language books? Otherwise it migh be interesting to see if there's any systematic tendency for translations to get a higher (or lower) score than originals.

>52 AndreasJ: - I'm pretty sure it's only for English-language books at the moment, though they may have included some Spanish-language ones. But yes, Lexile measures only apply to a specific ISBN, meaning that different editions of a work could have different measures.