The Working Memory Told Them Apart

What the research found

A child can be gifted and dyslexic at the same time. Educators call this pattern twice-exceptional, and it is easy to miss. The reasoning ability hides the reading difficulty. The reading difficulty masks the reasoning ability. A new study from Greece set out to describe, in numbers, what this combined profile actually looks like.

Susana Padeliadu and Athina Voulgari tested 150 students in Grades 4 to 6. They sorted them into three groups. Thirty-six were gifted with a learning disability in reading. Thirty-one were gifted without one. Eighty-three had a diagnosis of dyslexia and cognitive scores in the average range. Every child sat the WISC-V, a widely used intelligence test, and a Greek reading battery called DADA, which measures decoding, fluency and comprehension.

The first finding concerns reasoning. On the reasoning indices of the WISC-V, the gifted students with a learning disability were indistinguishable from their gifted peers. All the differences fell short of statistical significance. Their verbal comprehension score averaged 115.4, well above the population mean of 100. Whatever their reading struggles, their capacity to reason was intact and high.

The second finding concerns working memory, the mental scratchpad that holds information in mind while you use it. Here the two gifted groups parted ways sharply. The gifted students with a learning disability scored an average of 93.5 on the Working Memory Index, against 107.9 for the gifted students without one. In the language of effect sizes, the gap was large: Cohen’s d of −1.29. Working memory turned out to be the single cleanest way to tell the two groups apart. A statistical test of its sorting accuracy returned an area under the curve of 0.82, which counts as good discrimination. Processing speed, by contrast, sorted the groups poorly, at 0.63.

The third finding concerns reading. When the gifted students with a learning disability tried to decode words, they looked like the dyslexic group, not the gifted one. Their decoding sat well below their gifted peers, a very large gap (d of −1.77). Their reading fluency showed no significant difference from the dyslexic students. And yet their reading comprehension told a different story. On understanding what they read, they scored meaningfully higher than the dyslexic group (d of 0.67), while still trailing their gifted peers. The authors read this as partial compensation: strong reasoning propping up comprehension even when the mechanics of decoding remain weak.

Put the three findings together and a portrait emerges. These children reason like the gifted, decode like the dyslexic, and comprehend in the space between. It is a genuinely mixed profile, not a diluted version of either.

What it means for you and your child

If your child seems both unusually capable and unexpectedly stuck on reading, this study offers a vocabulary for what you may be seeing. High reasoning and weak decoding can live in the same child. One does not disprove the other. A bright child who reads slowly, or dreads reading aloud, is not necessarily careless or lazy, and is not necessarily less able than their comprehension suggests.

The study also points to where the tension tends to hide. Decoding problems are visible. A child stumbles over words, and adults notice. Working memory problems are quieter. They show up as losing the thread mid-instruction, forgetting the start of a sentence by its end, or needing steps broken down. In this sample, working memory was the feature that most reliably distinguished the twice-exceptional children from their gifted peers. It is worth asking whether an assessment has looked at it directly, rather than inferring ability from a single overall score that can average the highs and lows into a misleading middle.

The comprehension result carries a practical warning. Because these children understand what they read better than their decoding would predict, a comprehension test alone can make the reading difficulty disappear on paper. The difficulty is real; it is simply being compensated for, at a cost in effort and speed. A thorough evaluation looks at decoding and fluency in their own right, not only at whether the child eventually grasps the meaning.

Be careful about what this single study can and cannot tell you. It is a snapshot, not a film. Because the researchers measured everyone at one moment, they cannot say whether this compensation holds up over years, or whether it frays under the heavier reading loads of secondary school. The authors are candid about this. They also flag that their groups were unequal in size, that the children reached the study by different referral routes, and that they did not account for factors such as family background or co-occurring conditions. The compensation story, they stress, is an interpretation, not a proven mechanism.

There is a further limit worth naming for English-speaking families. Greek is a transparent language: letters map onto sounds with unusual regularity. English is not. It is plausible that decoding difficulties look somewhat different, and possibly heavier, in an opaque orthography like English. The broad pattern here, high reasoning alongside weak decoding and propped-up comprehension, is likely to travel. The precise numbers may not.

The sensible takeaway is modest and useful. If a child shows this split, an assessment should measure the parts separately, pay particular attention to working memory, and resist letting a strong comprehension score or a strong reasoning score paper over a genuine reading difficulty. Support can then target decoding and fluency directly, while giving the child’s reasoning the room it deserves.