The use of dynamic learning tasks to predict growth in reading, identify children at risk of literacy and related difficulties and forecast response to intervention.

The use of dynamic learning tasks to predict growth in reading, identify children at risk of literacy and related difficulties and forecast response to intervention.

First Author: Hannah Nash -- University of Leeds
Keywords: Dynamic assessment, Language, Reading comprehension, Reading development, Mathematics
Abstract / Summary: 

The purpose of this symposium is to bring together current research on the use of dynamic assessment (DA) in predicting growth, identifying children at risk of difficulties and forecasting response to intervention (RTI). Dynamic tasks are designed to measure the ability to learn through teaching during the task, providing a measure of learning potential as opposed to current knowledge (Grigorenko, 2009). Therefore, they provide a promising supplement to conventional predictors of growth and screening measures.

Recent published research has focussed on DA of decoding (e.g., Petersen et al., 2018; Gellert & Elbro, 2018). In this symposium the use of DA is expanded to broader language skills (vocabulary - Nash, morphology - Gellert; Whylie, and narrative - Petersen) and their role in predicting growth in reading comprehension and identifying those at risk of difficulties and RTI. Finally, Cho presents a DA of math designed to predict calculation and word problem solving outcomes.

Symposium Papers: 

Morphological awareness in the first year of schooling contributes to reading comprehension after a year of formal literacy instruction

First Author/Chair:Sarah Whylie -- Coventry University
Additional authors/chairs: 
Anna J. Cunningham; Helen L. Breadmore; Clare Wood

Purpose: Past research has shown that morphological awareness (MA) predicts reading comprehension (Cunningham & Carroll, 2015) and, to a lesser extent, word reading in intermediate readers (Deacon & Kirby, 2004). However, there have been mixed findings regarding the importance of MA skills early in development, and a lack of research on the contribution of pre-literate MA to the subsequent development of literacy skills. This study examines the extent to which children’s MA in their first year of school contributes to reading attainment a year later, after formal literacy instruction has begun.
Method: Forty English-speaking children (4;8-5;8 years at time 1) completed measures of phonological awareness (PA), vocabulary, word reading and MA. MA was assessed using a novel dynamic task. An analogous nonword blending task controlled for the contribution of PA to MA task performance. The children also completed standardised literacy outcome measures (word reading and comprehension) a year later.
Results: Multiple regressions showed that the novel MA dynamic and PA analogous tasks, as well as vocabulary, all contributed unique variance in reading comprehension. However, only vocabulary and PA contributed to word reading.
Conclusions: Our results extend previous research indicating that MA contributes to comprehension, by illustrating that MA measured in preliterate children’s oral language is predictive of reading comprehension. This highlights the importance of MA for even very young children. On the other hand, MA did not contribute to word reading once the PA component of MA had been controlled for. This finding supports the view that MA is multifaceted.

Using kindergarten dynamic assessment of language to predict future reading comprehension difficulty

First Author/Chair:Douglas B. Petersen -- Brigham Young University
Additional authors/chairs: 
Maureen Staskowski

Purpose: The purpose of this study was to examine how well a kindergarten dynamic assessment of language predicted future reading comprehension difficulty at fourth grade.
Method: A pretest-teach-posttest dynamic assessment was administered to a large group of students in the middle of the kindergarten school year. At pretest and posttest, kindergarten students were administered narrative-based language assessments. The teaching phase entailed brief narrative-based language lessons led by the students’ general education teachers. Reading comprehension was assessed at the end of the students’ fourth grade school year using an authentic, high-stakes reading assessment.
Results: Preliminary results indicate that the dynamic assessment accounted for significant variance in reading ability, with fair to good area under the curve values and sensitivity and specificity.
Conclusions: Dynamic assessment of language is a promising approach to identifying future reading comprehension difficulty of young kindergarten students.

Can a dynamic vocabulary learning task add to the prediction of growth in reading comprehension and the identification of children with reading comprehension difficulties?

First Author/Chair:Hannah M. Nash -- University of Leeds
Additional authors/chairs: 
Paula J. Clarke; Anna S. Gellert; Anna Weighall; Christopher Dixon; Emily Oxley

Purpose: A dynamic test of vocabulary learning has been shown to predict growth in vocabulary (Gellert & Elbro, 2013) in Danish children. We used an adapted, computerised version of this task with a large sample of children from a diverse range of backgrounds in the UK. The aim was to test whether the task was less biased by the child’s background than static measures of reading related skills, whether learning predicts growth in vocabulary and reading comprehension and whether it could help identify those experiencing reading comprehension difficulties at the second time point.
Method: We used a longitudinal design with two time points. 407 children aged 8-9 years were assessed at time point 1, where they completed the dynamic vocabulary learning task and standardised measures of receptive vocabulary, reading accuracy and comprehension. Of the 407 children, 136 had EAL. At the second time point 12 months later, children completed the standardised measures of vocabulary and reading comprehension. In the dynamic vocabulary learning task children were taught the names of six aliens and semantic information about them.
Results: At the first time point the measures from the dynamic vocabulary learning task were not strongly correlated with socio-economic status or English vocabulary knowledge.
Conclusions: The lack of strong correlations with measures of the children’s background at the first time point suggest that the dynamic vocabulary learning task is assessing ability to learn and not existing knowledge. Data from the second time point will address whether learning predicts growth in vocabulary and reading comprehension and whether it can help identify those experiencing reading comprehension difficulties.

Examining the predictive validity of a dynamic assessment of morphological analysis to forecast response to a morphological vocabulary intervention

First Author/Chair:Anna S. Gellert -- University of Copenhagen
Additional authors/chairs: 
Elisabeth Arnbak

Purpose: To investigate the value of a dynamic assessment of morphological analysis for predicting responsiveness to a morphological vocabulary intervention.
Method: 111 fifth-graders participated in a morphological vocabulary intervention of 24 sessions and outperformed participants in a trained control group on several measures of knowledge of taught and untaught morphologically transparent words (Gellert et al., 2018). Before the intervention we administered traditional tests of word knowledge along with a combined static and dynamic assessment of morphological analysis. In the first static part of this assessment the child had to define morphologically transparent pseudowords (not taught during the intervention). The second dynamic part was administered immediately after the static part. In the dynamic part of the assessment the examiner re-administered all the items the child had failed on the static part and for each item used a series of prompts until the child answered correctly or the prompts were exhausted.
Results: Results from regression analyses indicated that the dynamic part of the assessment of morphological analysis added significantly to the prediction of children’s response to the morphological vocabulary intervention over and above the static part of this assessment as well as the traditional tests of word knowledge.
Conclusions: This study documents the added predictive value of a dynamic assessment of morphological analysis to forecast response to a morphological vocabulary intervention. The results are in line with previous results from studies indicating that dynamic assessment of decoding is predictive of responsiveness to reading intervention (Aravena et al., 2016; Cho et al., 2014).

Dynamic assessment for identifying Spanish-Speaking English Learners’ risk for mathematics disabilities: Does language of administration matter?

First Author/Chair:Eunsoo Cho -- Michigan State University
Additional authors/chairs: 
Lynn S. Fuchs; Pamela M. Seethaler; Doug Fuchs; Donald L. Compton

Purpose: The purpose of this study was to evaluate whether administering dynamic assessment (DA) of mathematics in Spanish-speaking English learners’ (ELs’) first language (Spanish) has benefit over English administration in forecasting two types of math outcomes, calculation and word problem solving. Additionally, we examined whether ELs’ language dominance moderates the effect of DA language (Spanish vs. English) on its predictive validity.
Method: At the start of 1st grade, ELs (N = 368) were randomly assigned to English-DA or Spanish-DA conditions, assessed on static mathematics measures and domain-general (language, reasoning) measures in English, and completed DA in their assigned language condition. At year’s end, they were assessed on calculation and word-problem solving. Multi-group path models were run to compare the effects of English vs. Spanish DA on predicting mathematics outcomes depending on ELs’ language dominance.
Results: Results indicated that both English- and Spanish-DAs have additional predictive validity for explaining year-end mathematics outcomes. However, ELs’ language dominance moderated DA’s predictive validity. Spanish-DA showed higher predictive validity in Spanish-dominant ELs than English-dominant ELs when predicting calculations but not word-problem solving. English-DA was predictive for both outcomes, regardless of ELs’ language dominance.
Conclusions: The study findings deepens our understanding of how mathematics DA may be designed to address the EL population’s linguistic characteristics. Spanish-dominant ELs’ potential for learning calculations may be underestimated when DA is administered in English. Thus, Spanish-DA appears preferable when predicting calculation outcomes. By contrast, English-DA appears preferable when predicting word problem solving outcome, which heavily rely on English language.