How readable is your writing? If it’s too complex, you risk losing your readers. For those interested in promoting their websites, that would be disastrous. Measuring a qualitative factor such as readability is no simple task — a complete method would take into account sentence structure, diction, and other factors that cannot be quantified. However, algorithms do exist to measure a small subset of these factors and while they will not guarantee the level of readability they claim your writing is, they are helpful in giving a general idea of where the actual readability stands and what improvements, if any, need to be made.
One such algorithm, the Automated Readability Index (ARI), was developed for the US Air Force in 1967. The ARI determines the readability of a piece of text using the number of characters per word and the number of words per sentence. In other words, longer words and longer sentences equate to reduced readability. The output is the estimated US grade level education needed to understand the text.
Another index very similar to the ARI is the Coleman-Liau Index (CLI) which was developed in 1975 to estimate the readability of a printed work without necessitating actual character recognition. Like the ARI, it uses characters per word and words per sentence and outputs a US grade level. However, it puts greater emphasis on the number of words per sentences in determining readability.
Other indices are more advanced and use not the number of characters per word, but the number of syllables per word. Examples of this include the Flesch-Kincaid readability tests consisting of two individual indices: the Flesh Reading Ease Score (FRES), a score out of 100, and the Flesh-Kincaid Grade Level (FKGL), a US grade level. These two indices are approximately inverses of each other — a high FRES indicates simplistic language corresponding to a low FKGL.
A similar index is the Gunning Fog Index (GFI), which, like the others, factors in the number of words per sentence. In addition, it takes into account the number of complex words — words that are not proper names or compound words and have at least 3 syllables minus any suffixes. Other words are therefore assumed to be equal in difficulty, an assumption that is generally true in the English language. Like most of the other indices, the GFI is measured in US grade level.
It would be quite a pain to count up everything manually. Luckily, there are automated tools available to help. Several of these are online such as The Readability Test Tool. Other tools can be used offline such as Flesh which is free and open source. For WordPress, the Word Stats plugin calculates several readability indices while a post is being composed. This immediate feedback can help with improving the readability of a post as it is being written.
The final, yet perhaps most important issue is determining what index to aim for. As expected, this is highly dependent on your target audience. For material intended for a global audience, the grade level should be kept below 5 or 6. For academic material, the index will likely be higher but should be kept at a level at most the grade level where the material is typically covered. Research papers and legal material typically score the worst readability where accuracy and precision has top priority. This blog aims for a grade level index of less than 13 on its posts so that its content can be understood by most educated individuals.
Above all, remember that these indices are only numbers that can only hint at the level of a piece of writing. If readability is important, the writing must be proofread by an actual human to corroborate the indices. While text consisting of short sentences and short words will be deemed readable by the indices, it may still have a structure rendering it impossible to decipher, even with close scrutiny.