What Is the Structure of Processing Speed in Children, and How Can This Ability Affect Reading?

 

I have always found the meaning of the term “processing speed” somewhat elusive. On the one hand, it refers to the speed of thinking—the speed at which a person performs cognitive operations, which may be simple or complex. On the other hand, we assess processing speed using extremely simple tests that involve very little thinking. The main reason for keeping the tests simple is the attempt to measure processing speed in a relatively pure form, with minimal influence from other abilities. However, this raises the question of how much we can infer from processing-speed test results about a child’s ability to perform more complex cognitive operations quickly.

Gerst and his colleagues attempted to investigate the structure of processing speed and the relationships between processing speed and reading.

The researchers defined a processing-speed test as “any test that involves some form of cognitive processing and imposes a time constraint.” They therefore included tests that are not usually classified as processing-speed measures. The researchers identified 16 such tests and classified them according to three dimensions: complexity, modality, and time measurement.

Within the complexity dimension, simple processing speed was defined as reaction time to a perceptual stimulus, without any additional cognitive demands. Individual differences in such tasks are due to differences in speed rather than accuracy, because performance on these tasks is almost perfectly accurate. One example of a test of simple processing speed is the fifth part of the Trail Making Test, or TMT, which assesses motor speed. Another example would be repeatedly and rapidly selecting the direction in which an arrow is pointing—right or left. Processing speed at an intermediate level of complexity was defined as response speed on a test requiring a small amount of cognitive processing, without switching, inhibition, or working-memory demands. Tests assessing processing speed at this level include coding tests, the Woodcock-Johnson Visual Matching test, the first part of the TMT, which requires the participant to connect numbers only, and a color-naming test. Processing speed at a high level of complexity was defined as speed on a test requiring goal-directed processing and cognitive efficiency. This type of processing speed is essentially executive functioning that emphasizes speed or efficiency rather than problem-solving, working memory, or cognitive flexibility. Phonemic and semantic verbal-fluency tests are examples for high complexity processing speed tests although they do require cognitive flexibility. This is reflected in the ability to switch between different aspects of a category—for example, moving from sea animals to farm animals. The second part of the TMT, in which the child is asked to alternate between letters and num, bers, assesses processing speed at a high level of complexity. The Tower of London/Tower of Hanoi test, shown in the image below is also assigned to this category.

 

In this test, the child is asked to arrange the rings on peg C in exactly the same order in which they are arranged on peg A, without placing a larger ring on top of a smaller ring. The measure of complex processing speed in this test was defined as the time required to make the first move—to move the first ring—when the child completed the entire task correctly. The time required to make the first move serves as an indicator of the amount of planning the child invested in the task.

Within the modality dimension, processing-speed tests were divided into groups according to the type of stimulus: alphanumeric stimuli, such as letters and words, and nonalphanumeric stimuli, such as shapes and colors. Alphanumeric tests include phonemic-fluency tests (quickly finding words that begin with a specific letter), both parts of the TMT, and the Visual Matching 2 subtest from the Woodcock-Johnson battery. Nonalphanumeric tests include rapid color naming and rapid picture identification. The tests were also classified according to the type of response required from the child: verbal or motor.

Within the time dimension, the tests were divided into three groups: (1) Tests measuring the time between the presentation of a stimulus and the response to a single item. One example is the first move in the Tower of London/Tower of Hanoi test. (2) Tests measuring accuracy and speed within a predetermined time limit. Examples include verbal-fluency tests, Visual Matching 2, and coding tests. (3) Tests measuring the time required to complete the task. Examples include both parts of the TMT and cancellation tests.

The participants were 844 children in Grades 3–5. The children completed 16 different processing-speed tests, as well as reading tests. The reading measures assessed untimed single-word reading, single-word reading fluency or speed, and reading comprehension. The children were also assessed using a Woodcock-Johnson oral-language comprehension test (the child listens to a short passage and supplies a word that is missing from it), as well as tests of phonological awareness and rapid letter naming.

Findings

All three reading tasks (timed and untimed single-word reading and reading comprehension) were related to performance on the rapid automatized naming (RAN) letters test. Single-word reading was also related to phonological processing, but not to oral-language comprehension. Reading comprehension was also related to oral-language comprehension, but not to phonological processing.

Two structures of processing speed were found to fit the data: a structure based on complexity and a structure based on the time dimension.

Within the complexity dimension, the model that fit the data contained two levels of complexity: simple processing speed and complex processing speed. Complex processing speed incorporated both the intermediate and high levels of complexity described above. No evidence was found to support a distinction between tests assessing processing speed at intermediate and high levels of complexity. Thus, this model does not distinguish between tests such as Visual Matching or Coding and tests such as verbal fluency or the letters-and-numbers condition of the TMT, or TMT Part B. This is despite the fact that, intuitively, verbal-fluency tests—both semantic and phonemic—and TMT Part B appear far more complex than Visual Matching and Coding tests. The complexity of verbal-fluency tests and TMT Part B results from their executive component. They therefore lie at the boundary between processing speed and executive functioning.

Within the time dimension, the model that fit the data contained two components: a response-to-a-single-stimulus component and an efficiency component. The efficiency component included both tests that must be completed within a specified time and tests in which the time required to complete the task is measured.

The researchers state that they prefer the complexity model because it has a stronger theoretical basis.

Single-word reading, single-word reading fluency or speed, and reading comprehension were all related to processing speed on complex tasks. Simple processing speed was inversely related to single-word reading fluency. In other words, the higher the child’s simple processing speed, the lower the child’s single-word reading fluency. Simple processing speed was not related to untimed single-word reading or to reading comprehension. Processing speed contributes to reading because reading requires rapid matching between graphemes and phonemes. Fluent text reading also requires matching the visual pattern of an entire word with the phonemic sequence needed to pronounce it and with its meaning. Processing speed contributes to reading comprehension because reading comprehension requires words to be connected rapidly with their meanings and with the background knowledge that helps the reader understand the content of the text.

Nevertheless, processing speed made only a small contribution to reading in this study. It accounted for approximately two per cent of the variance in the reading measures beyond the variance explained by rapid letter naming (RAN), phonological awareness, and oral-language comprehension. However, the RAN test is itself a measure of processing speed (in addition to being a measure of retrieval fluency). RAN has high complexity (because of the need to switch between letters).

 

Gerst, E. H., Cirino, P. T., Macdonald, K. T., Miciak, J., Yoshida, H., Woods, S. P., & Gibbs, M. C. (2021). The structure of processing speed in children and its impact on reading. Journal of Cognition and Development, 1–24.