Reading comprehension suffers when words can’t be revisited
In the brave new digital world of reading, words flash by one at a time on the tiny screens of smart watches and phones. This portable, pageless story doesn’t end well for people keen on understanding what they’ve read, say psychologist Elizabeth Schotter of the University of California, San Diego and her colleagues.
Rereading words salvages understanding of initially confusing passages, Schotter’s team reports April 18 in Psychological Science. Software that presents words one at a time makes it impossible to scan previously read words and phrases, undermining text comprehension, the scientists say.
Other studies have shown that readers use eye movements to glean information about upcoming words in a sentence before visually fixating on those words, Schotter adds. People deciphering messages from strings of single words lack visual hints about what they’re about to read and can’t take second looks at the words they’ve just read.
“Eye movements are a crucial part of reading, and our ability to control the timing and sequence of information about a text is important for comprehension,” Schotter says.
The new results “provide a really nice demonstration that poor comprehension will be a major problem for any technology such as Spritz that controls the rate and sequencing of words arriving to the language system,” comments psycholinguist Klinton Bicknell of Northwestern University in Evanston, Ill., who has collaborated with Schotter but was not involved in the new study.
Researchers have found that rereading text takes up 10 to 15 percent of the time devoted to reading. Schotter’s team suspected that rereading helps to clarify confusing passages. But rereading is difficult to study and experimentally manipulate.
In the new study, 40 college students faced a computer monitor equipped with an eye-tracking device and read 40 sentences that lacked punctuation. Twenty sentences were ambiguous (including “While the man drank the water that was clear and cold overflowed from the toilet”) and 20 were straightforward (“While the man slept the water that was clear and cold overflowed from the toilet.”) Eye tracking noted whether students reread any material.
But the volunteers could not reread material in some sentences because x’s masked each word immediately after a person read it. Upcoming words remained visible as clusters of x’s sprouted behind them. A comprehension question followed each sentence, such as “Was the man drinking water from the toilet?”
When words weren’t masked, volunteers understood an average of 84 percent of all sentences when words weren’t reread and 78 percent when they did reread material, a difference no greater than expected by chance. People apparently resorted to rereading only when confused about a sentence’s meaning, and backtracking pulled their comprehension back to the level attained for sentences that were understood on the first pass, Schotter proposes.
In contrast, comprehension fell to an average of 71 percent when volunteers read sentences with masked words, a statistically significant reduction. In a few instances where students tried to reread masked words, comprehension dropped even further.
Participants showed comparable declines in their understanding of ambiguous and straightforward sentences that couldn’t be reread. That result indicates that reading words one at a time impairs understanding of both simple and complex messages, Schotter says. Reading accuracy may especially suffer when dense texts, such as business contracts, are “Spritzed,” she suspects.
The company that developed the technology is working on ways to enable rereading of words while Spritzing, says mechanical engineer Maik Maurer, cofounder and chief technology officer of Boston-based Spritz. But Schotter’s investigation examined only readers who moved their eyes while reading sentences, while Spritz removes all eye movements when reading, based on research Maurer says he has conducted with colleagues. “This new study isn’t relevant to the Spritz methodology.”
Spritz positions each word so that the eyes can fixate on it without moving, triggering recognition of the word and its meaning, he says. Proper positioning makes words easily readable without advance visual cues from eye movements.
Research to support those claims has not appeared in any peer-reviewed scientific journals, Schotter responds. Rather than describing experimental methods or presenting data, she says, Maurer makes “vague, unsubstantiated claims” about Spritz’s effects on reading.
E. Schotter, R. Tran and K. Rayner. Don’t believe what you read (only once): Comprehension is supported by regressions during reading. Psychological Science. Published online April 18, 2014. doi:10.1177/0956797614531148.
B. Bower. Read all about it. Science News. Vol. 167, April 30, 2005, p. 280.