Computerized Dynamic Adaptive Tests with Immediately Individualized Feedback for Primary School Mathematics Learning.

Introduction With the use of computer and communication technology, researchers have developed various assessment systems in improving students' learning achievement. The characteristic of learning or testing systems is to provide meaningful and personalized feedback that can accommodate a diversity of student personalization (Chen, 2009; Chen, 2011; Hwang, Panjaburee, Triampo, & Shih, 2013; Kao, Lin, & Chu, 2012; Luft, Gomes, Priori, & Takase, 2013; Moridis & Economides, 2012; Panjaburee, Hwang, Triampo & Shih, 2010). However, many computerized-based tests place greater emphasis on feedback on the final product of assessment tasks rather than the process of the learning. Moreover, all examinees received the same items and were required to complete all the items within the assessment procedure. These test systems do not capitalize on advantage of combining technology with testing, a computerized adaptive test (CAT). CATs offer examinees customized items suited to their aptitudes and cognitive status and shorten the length of the test (van der Linden & Pashley, 2010). Adaptive tests can provide equal measurement precision for most test-takers (Yen, Ho, Chen, Chou, & Chen, 2010; Yen, Ho, Liao, & Chen, 2012). Wu, Kuo, and Yang (2012) have developed a knowledge structure-based computerized adaptive test for diagnosing students' learning profiles. In the research of Wu et al. (2012), an adaptive test algorithm that can utilize fewer items than the original paper-based tests was proposed. The knowledge structure-based computerized adaptive test provides feedback regarding how well students performed on specific skills immediately after testing with fewer items (Wu et al., 2012); however, it is not sensitive to the process of students' learning. Meaningful feedbacks that can facilitate learning effectiveness are timely feedback and process-oriented feedback (Gabelica, Van Den Bossche, Segers, & Gijselaers, 2012; Harks, Rakoczy, Hattie, Besser, & Klieme, 2014; Hattie & Timperley, 2007; Parr & Timperley, 2010; Wang, 2011). Dynamic assessment is an interactive assessment that can provide students with both timely and process-oriented feedback (Campione & Brown, 1990; Haywood & Lidz, 2007). Dynamic assessment is grounded in Vygotsky's (1978) sociocultural theory, especially the Zone of Proximal Development (ZPD) concept, which describes how a student can go from his actual development level to his potential development level through continuous communication and consultation in the course of interactions with teachers, peers, or parents (Poehner, 2008). The ZPD relates to the gap between what the child can learn without help and what he or she can learn with adult guidance or in collaboration with more capable peers. Dynamic assessment embeds intervention within the assessment procedure. Typically there is a pre-test (to diagnose the characteristics of the student), an intervention (to provide appropriate feedback based on the student's characteristics and to help him or her learn), and then a post-test (to examine whether the student has learned the new concept after the intervention) (Haywood & Lidz, 2007). Despite the potential benefits of dynamic assessments, the need for experienced administrators and the amount of time involved have limited their use in applied educational settings (Frisby & Braden, 1992). Some researchers have reported the effectiveness of web-based dynamic assessments in enhancing students' academic performance (Chen et al., 2010; Wang, 2010; Wang, 2011). Computerized dynamic assessments can be applied to the majority of students and would greatly improve the feasibility of using dynamic assessments in the classroom. Based on the application of Vygotsky's theory, Campione and Brown (1987) have developed the graduated prompt procedure to implement dynamic assessments. In the graduated prompt procedure, predetermined prompts that range from general to specific are given to provide students with gradual assistance until they solve the problem. …