You probably consider yourself to be a good reader - this is, after all, a magazine that demands a high level of reading ability. And because you are a good reader, you will likely assume that you read via a fluent wave of ocular stimulation, smoothly picking off words in turn in each sentence. That assumption would be wrong, says Holly Joseph, associate professor at the University of Reading’s Institute of Education.

“A lot of people think that their eyes move smoothly over the lines of text, so when I tell them that they actually jump from word to word, they are often surprised,” she explains. When we read, our eyes actually move in a series of jumps (saccades) and brief pauses (fixations), she says.

It is during these fixations that we extract the information we need from the text to start to process it linguistically - ie, to make sense of what we are reading. So reading is less Michael Jackson moonwalk and more Super Mario bouncing from block to block to reach the other side of a crevice. We don’t “step” completely on those “blocks”, either.

“Most people look a little to the left of the centre of the word if they are reading fluently - which, for most, will be from around the age of 7 - and they then see the immediate companion letters in their vision, and out from that the words get more blurry.

“Hence, when we read, we are making guesses as to what the word is. So with the word ‘because’, you would focus on the ‘c’ and the ‘e’ will be right on the edge of your vision. Adults’ fixations are generally 180-200 milliseconds and typically developing children’s fixations (depending on their age) are closer to 250 milliseconds.”

These fixations are when multiple processes take place. “You’re visually processing what you’re reading and, at the same time, your oculomotor system is making decisions about when and where to move your eyes to next,” explains Joseph. “You’re having to try and match these squiggles that you see in front of you with information you have stored in your memory about letter shapes, and then you have to combine those letters to make words and access the meaning of those words.

“And that’s before you even start making sense of the grammatical structure and building an overall meaning of the sentences. All of these very complicated processes are happening at once.”

Her research is of particular interest to schools as it can reveal these hidden processes to teachers to assist in diagnosing and supporting problems that children have when reading. For example, her recent focus has been on building a picture of what children do when they encounter a word they do not know, which feeds into the host of recent research around the importance of vocabulary knowledge in accessing the curriculum (for more on this see “Ceasefire in the reading wars”, Tes, 15 June).

“We often think of learning new words as something that happens in early childhood and, at that early point in your life, you learn words through spoken language from your parents, etc, but from the age of 9 onwards, the majority of words that children learn, they learn from reading,” she explains.

“So, I was interested in looking at what happens when a child is reading a story and they come across a word they don’t know? What do they do to try to understand the meaning of that word and then, when they encounter the word again, how do they build on the first experience they have of the word?”

Eye-tracker technology

Using eye-tracker technology, Joseph has designed a method of capturing this complex process in a laboratory setting, allowing her to look at the extent to which reading times on new words reduce with multiple exposures to these words. She intends to use this information to create what she describes as an “index of learning”.

“In my study, children encountered a number of rare words that they didn’t know in sentences that told them something about what the word means,” she reveals. “For example, ‘confabulated’ might appear in a sentence like, ‘The children confabulated for ages as they waited for the bus’. So the first time they read that word in a sentence, they spend ages reading it. They’ve never seen it before, they don’t know what it means, it’s long and it’s hard to read, so they maybe fixate on it for 600 milliseconds.

“Then they read it in another sentence and it’s still hard, they still don’t really know what it means, but the word is a little bit more familiar and they are starting to carve out a vague meaning for it. This time the reading time goes down to 500 milliseconds and the next time it maybe goes to 450 and then 400. So over these multiple exposures to this word within sentences, reading times gradually go down.”

Joseph says that this gradual reduction in reading times over repeated exposures to a new word can be seen as an index of learning: longer reading times are associated with processing difficulty, so if we see reading times decreasing each time the word is read, then this tells us that the child is learning something about the word, about its spelling and its meaning.

She adds that some children are more efficient learners (their reading times reduce more dramatically with each exposure) and some children reduce their reading times very slowly. “This tells us something that is interesting for researchers but also potentially useful for teachers, too, in working out why children may be struggling,” she says.

“If we look at children who we know have poorer reading comprehension (so-called ‘poor comprehenders’) compared to those who have better reading comprehension - and we give them particular kinds of sentences or passages where they look longer or don’t look longer compared to those who have better comprehension - it might tell us something about why they are having difficulty understanding what they are reading.”

One area where this knowledge could come in particularly useful is inference, which many poor comprehenders typically struggle with. “So you might have a sentence like: ‘Mr Jones looked up at the dark, grey clouds and looked for his umbrella. When he got to work, he was soaking,’” says Joseph. “When you get to the word ‘soaking’, if that is going to make sense to you then you need to have made the inference that the dark, grey clouds and the umbrella mean it was going to rain.

“If you’ve got good reading comprehension skills, you might look longer at the word ‘soaking’ as you link it back to the umbrella and grey clouds, and make the inference that it had rained but Mr Jones couldn’t find his umbrella. If you’re a poor comprehender, you might not do this, for various reasons. You might not know the word ‘soaking’, or you might not have linked it with ‘umbrella’ and ‘grey clouds’ to make the inference.

“If we compare reading times on key words like ‘soaking’ in this type of passage, this might give us some clues as to why these children are failing to understand and this might influence what teachers can do to help them.”

Joseph has already completed some research work around inferences and children and adolescents with autistic spectrum disorder (ASD), and she found that there were some subtle differences between those with and without ASD. “In our study, the young people with ASD took a bit longer reading words like ‘soaking’ and went back to re-read the passage more than those in the control group. In other words, they needed a bit of extra time, but they got to where the other children got to eventually. So, they did make those inferences but, as they were reading text, their reading behaviour differed a bit [compared to children without ASD].”

Joseph says she also made some interesting discoveries during a study of children who speak English as an additional language (EAL) that could prove beneficial to teachers.

“In one of our word-learning studies, when we compared [the eye movements of] children who speak English as an additional language with monolingual English speakers, what we found was that those children’s eye movements [relating to] reading times reduced much more dramatically than those of the monolingual children as they read new words over 10 exposures,” says Joseph. “I presented this data to some teachers recently and they were really excited.”

The key takeaway for teachers from this would be that although, in general, this group of children tends to have low English vocabulary and low reading comprehension because the language is new, Joseph’s data shows that, through doing more reading, children with EAL seem able to increase their vocabulary very efficiently.

“I wouldn’t say that teachers should just tell these children to go home and read a lot but, given that teachers can’t spend hours every week explicitly teaching new vocabulary, this is one way that might work unexpectedly well for this group who, in our studies, are shown to be efficient learners,” she says.

Joseph is working more with teachers to make sure that her research is more directly relevant to what goes on in the classroom, because she believes that her specialist area can offer a wide range of unique insights.

“Eye movements shed light on processes that are mostly unconscious during reading and can therefore tell us when and why a child is struggling to understand what they are reading, even if they cannot articulate this themselves,” says Joseph.

“Understanding more about the process of reading comprehension rather than just the end product can help us to support children who struggle with reading for many different reasons,” she adds.

Simon Creasey is a freelance journalist