The ERF Reading Scale has the virtue of being solely based on the publisher-reported headword count for each reader. The method of determining the list of headwords used in a text, however, may vary considerably depending on the publisher’s list of word frequencies employed and what frequency range of headwords are used to determine the headword list for a particular book. For example, a reader might have a headword count of “200” but those 200 words might have been drawn from any words in the top 1000 most frequent words on the publisher’s list, in some cases.
Text difficulty may also vary depending on factors other than the headword count:
- There may be other, more difficult words outside of their headword count. that have been glossed or illustrated as an aid to comprehension.
- The quality and frequency of illustrations might aid comprehension
- Background knowledge of the topic aids comprehension. Stories, for example, based in other cultures might prove more difficult.
- The number of characters in the story and the student’s familiarity with the names employed.
- The complexity of the plot.
- Events out of chronological sequence.
- The total length of the book. Other factors being equal, longer books tend to be more difficult.
- The size and style of the font used.
Note that headword counts are only available for books claimed by their publishers to be “graded readers”. Books published for native English readers, therefore, have no ERF Scale rating.
There are other reading scales that include a human judgement of the relative difficulty of text. Two are listed below. Both include ratings for a wide range of texts, both graded readers and native “youth readers”.
The “Yomiyasusa Scale”
The “Yomiyasusa Scale” (Japanese for “ease of reading”) was originally based on the combined judgements of a group of teachers and graded reader enthusiasts, who judged their difficulty (for Japanese learners) by comparison with the established values of other readers on the scale. It is reported to correlate well with student-perceived difficulty as reported by Holster et al. (2017). We know of no research, however, on how well these difficulty ratings work for students from other language backgrounds.
The MReader “Kyoto scale”
This 10-point scale roughly based on headword counts and modified by reference to other similar books. While not as accurate as the Yomiyasusa scale, the “Kyoto Scale” has the advantage of grouping the books into 10 discrete levels, which allows library collections to color code and arrange the books by level, which greatly aids student selection of suitable books.
The Lexile Scale
The Lexile Scale, by Metametrics, is a readability scale that is purely based on textual factors such as average word length, average sentence length, word difficulty and other such factors based on their own proprietary (read ‘secret’) formula. It is widely employed in the U.S. education industry but does not yield consistent results for English learners.