Leveled Readers

From OLPC
Jump to: navigation, search

I depend on others in the community to address whether these ideas would have merit in schools outside the U.S. My thought is that they would be useful in teaching English, especially if audio tracks were associated with them. I don't know if literacy experts in non-English speaking areas use this approach to teach children to read, and it seems that the mere translation of simple English texts might not have the same difficulty in another language.

Finished Leveled Readers

File:Frog facts.pdf (html/css+png version coming soon)

What is a Leveled Reader and Why are They Needed

Leveled readers are reading texts that have been rated as to their difficulty. U.S. teachers of literacy use leveled readers to guide literacy learners through ascending levels of difficulty in the texts they attempt to read on their own. It has been found (citation) that this improves motivation.

There are several systems for scoring reading texts. A correlation table relating several such systems can be found here [1]. A paper describing the history of scoring reading material and assessing two such systems is found here: [2].


Providing leveled text on or accessible to the XO would aid the teaching of English literacy. The form of these texts might be short books or perhaps comic strips.

Evaluating the level of a text entails recognizing various classes of words:

  1. Sight words -- These are common words that teachers often ask children to memorize. For frequently used lists of these words see [3] and [4]
  2. Simple words (consonant, vowel, consonant) with no inflections or perhaps made plural.
  3. One syllable words with consonant combinations (th, sh, etc.)
  4. Contractions
  5. Multi-syllable words which are simple words lengthened by common prefixes and suffixes
  6. Words with uncommon letter sequences (-irt, -urt, -eigh, for example.)
  7. Special words such as proper nouns or onomatopoeic words.

Texts containing mostly words from the beginning of this list of classes are simpler than those having many words from the higher numbered classes.

Level ranking also depends on stylistic clues such as repetition and rhyming. It depends on context: are there pictures giving clues to the text? Is the font large, with short sentences fitting entirely on the page?

On the OLPC XO several activities might take the place of the lowest leveled readers. In particular, gCompris reading activities which allow the child to pick words to label images, or to fill in letters spelling the name of images seem to be roughly equivalent to "caption books".

Other Rankings and Progressions

In addition to the difficulty ranking of the content, emphasizing different phonic concepts is useful. Children could then progress across various practice texts to build their decoding skills, since this is recognized to be beneficial in learning to read. Perhaps additional sequences might prove important if the texts or activities were being used for English language learning. (For example, progressions of vocabulary or tenses or uses of words containing difficult sounds.)

Infrastructure for Creating Leveled Readers

To produce the subsequent levels of short readers, the community might cooperate in creating simplified versions of stories found within the Gutenberg corpus or the Internet Archive, or non-fiction readers based on Wikipedia articles. Together with appropriately licensed graphics from Wikipedia or such sites as Flickr, a bookroom of these leveled texts would be available to teachers. The simplest way to create this content might be to link to a [Bookroom] page on this wiki. Classification could be done by other community members if there were exemplars for each level in place.

It might be helpful to create a web application that allowed images and text to be associated in a "page" that would then be stored in a form providing integration with pootle translation.

  • Some open source passages with reading level scoring.
  • Manual for online tool in previous link.

Further Examples

Dinosaur_book is intended to be the first example of a leveled reader created from copyleft materials by the greater-OLPC community. Its goal is to not only create content, but hopefully a process/framework for further creation of content.