Anne Fleur Kortekaas-Rijlaarsdam

CHAPTER 6 120 Practice phase The task started with 18 practice math problems (neither feedback nor reward were given) which were selected randomly from the intermediate level of a pool of 600 possible calculations (see below). A practice trial consisted of a white screen with a fixation cross (duration 1,000 ms), followed by the calculation presented on a white background (duration maximum 10,000 ms), followed by a 1,500 ms feedback screen. To minimize working memory load, trials were self-paced (with a maximum inter- problem interval of 10,000 ms) and the calculations were shown until a response was given (or the maximum inter-problem interval elapsed). Calculations differed in difficulty level ranging from 1 (easiest) to 12 (most difficult), for a specification and theoretical substantiation of difficulty levels, see below. Each child started at the intermediate level (Level 7). After each set of three trials, the level was adapted to the child’s performance. If the child answered all three calculations correctly, the level was increased (Level +1). If two calculations were completed correctly, the level remained unchanged. If the child answered two or three out of three calculations incorrectly, difficulty level was decreased (Level -1). The 18 practice trials calibrated the task to allow for each child to reach a level suitable for his or her age and level of mathematical skills. The level reached at the end of the practice phase determined the entry level (difficulty level) in the experimental phase. Experimental phase The experimental phase consisted of two conditions that were explained to the child prior to the start of the task, with trials presented in a random order: a neutral condition (21 unique calculations), and a motivation condition (21 unique calculations). Neutral trials were comparable to practice trials in setup: A white screen with a fixation cross (duration 1,000 ms), followed by the calculation presented on a white background (duration maximum 10,000 ms), followed by a 1,500 ms feedback screen. A motivation trial consisted of a similar setup but was identified by a green background (1,000 ms fixation cross on a green background), calculation on a green background (duration maximum 10,000 ms) and a green feedback screen (1,500 ms). For both conditions, if an incorrect answer was given, the screen remained empty. In the motivation condition, a correct response was followed in 85% of trials by a smiley, text (‘Correct’ ‘+1 point’ ‘Total score =’) and the sound of applause. Correct trials in the neutral condition were followed by an empty screen. Within each condition, an independent tracking algorithm (equal to the one used for the practice phase) adjusted the difficulty level for each child, to ensure that the number of correct responses and thus number of trials with rewarded feedback was approximately similar for each child. We used this algorithm to ensure that children gave on average 67% correct responses and 33% incorrect responses. In