The aim of this study was to build on the work done in the three previous experiments on the grant that looked, in turn, at the development in children of primary memory, processing speed, and secondary memory. Measures taken from these earlier studies were used in this work to determine which of these factors underpinned working memory development in children, and the extent to which they mediated the relationship between working memory and academic attainment. A total of 218 children in school years 2 and 3 completed a range of experimental measures tapping primary memory, various aspects of processing speed, and secondary memory. Experimental measures of short-term memory ('simple span') and working memory ('complex span') were also given, as were a standardised test of mathematics [Numerical Operations subtest of the WIAT-II UK], a novel measure of reading comprehension, and a standardised test of classroom behaviour [Conners’ 3 Teacher Short Form] This data collection contains the data from the fourth of four studies conducted on the associated ESRC grant (see Related Resources).The aim of this project is to build on previous psychological research with both children and adults to provide the most comprehensive model to date of the factors involved in the development of working memory performance in children. In doing so, the project will investigate the extent to which these factors are separable or inter-related. Also the project will assess how these factors contribute to mediating the strong relationships commonly observed between working memory and academic attainment. The research has four specific objectives: To determine whether age-related changes in short-term memory capacity are related to working memory development. To determine how age-related changes in processing speed are related to working memory development. To determine whether age-related changes in long-term memory utilisation are related to working memory development. To determine which of the above factors mediate the relationship between working memory performance and educational attainment. These objectives will be met in a set of empirical studies, using both existing and novel experimental measures. These measures will be related to academic attainment and measures of classroom behaviour.  Each study will involve large samples of children in two age groups (around 5 and around 9 years of age).

This study used an empirical, experimental data collection method. All tasks, apart from the standardised measure of mathematics and the Conner’s 3 Teacher Form were programmed using Runtime Revolution software and presented on MacBook Pro laptop computers. The Conner’s 3 Teacher Form was completed by each participant’s class teacher. A total of 180 words were used in processing elements of the tasks, which were familiar concrete nouns, with age of acquisition of under 6.2 years. No words were repeated within or between tasks in a single testing session. All audio material was presented through the internal laptop speakers using pre-recorded male voices. Where a processing task was used in any of the tasks (including decision making in the memory scanning task), reaction times (RTs) were recorded when the child selected their response key (z or /) on the computer keyboard. ‘z’ corresponded to correct, and ‘/’ to incorrect in every task. Participants were assessed individually in three testing sessions lasting approximately 20 to 30 minutes each. In session 1, they completed the free recall task, the first assessments of i) verbal baseline processing, ii) visual choice, and iii) digit cancellation, and then the interleaved lists task, the digit span task, and final free recall, in that order. In session 2, they completed the articulation rate task, delayed free recall, and the forgetting task, in that order. In session 3, they completed the complex span task, the memory period task, and the second assessments of verbal baseline processing, visual choice, and digit cancellation, in that order. In addition to the memory measures given in these experimental testing sessions, all children were tested on a reading and mathematics test in a separate whole group session in their classroom, and each child’s class teacher was asked to complete the Conners’ 3 Teacher Rating Form. The full sample consisted of 218 children (95 males, mean age 7 years 8 months, range 6 years 8 months to 8 years 9 months, with 115 children in Year 2 and 103 in Year 3). Absences at the time when the reading and mathematics assessments were given meant that a full data set that also included these measures was only available for 205 children (109 in Year 2 and 96 in Year 3). Some teachers did not complete the Connors’ form (self selecting omissions), which resulted in data for 147 children (95 in Year 2, 52 in Year 3) being available for this measure.