Till AIED konferensen 15th International Conference on Artificial Intelligence in Education, in Auckland, Nya Zeeland, presenterade Annika Silvervarg (Linköping Universitet) och kollegorna Lena Pareto, Magnus Haake, Thomas Strandberg samt Agneta Gulz (Lund Universitet) följande artikel 2011: An Educational Math Game with a Teachable Agent and a Social Chat (eftersom artikeln publiceras internationellt, är den på engelska.)
Abstrakt: We present an educational math game, including a teachable agent and a social chat, that trains basic arithmetic skills with a focus on grounding base-ten concepts in spatial representations. It employs a board-game design with a variety of different sub-games, game modes and levels of difficulty. [...] In other words, the basic game with the TA contains a form of on-task conversation between agent and student. But the game architecture also has been extended with a chat where the student can engage in conversation with the TA, writing freely by means of the keyboard and bring up basically any topic in a chat-like manner. [...] One reason to include off-task conversation is to enrich the game and its motivational qualities for the age group in question (12-14 year olds). Another is to be able to explore whether such a conversational module can enable pedagogical interventions, such as supporting pupils math self efficacy and change negative attitudes toward math in general. » Läs mer

”Sveriges bästa uppsats i systemvetenskap skrevs av studenten Ann Nilsson från Högskolan Väst. Därför tilldelas hon årets BEST-stipendium för uppsatsen Lärsituationens komplexitet för elever som upplever svårigheter i matematik – Rutiga Familjen i ny lärmiljö.”
Pressrelease JUSEK, lördag 14 november 2009 » Läs mer

Till AIED International Conference on Artificial Intelligence in Education, in Brighton, Storbritannien, presenterade Lena Pareto följande artikel 2009: Teachable Agents that Learn by Observing Game Playing Behavior (eftersom artikeln publiceras internationellt, är den på engelska.)
Abstrakt:  An educational game in mathematics targets cognitive skills oincluding pattern finding and generalization from example by teaching an agent. These can be difficult for young children, let alone teaching about them. Therefore, we have designed a new way of teaching an agent, the teach-by-guiding model. We describe the model and its basic implementation. Because this is a work-inprogress, we end by raising questions of how such an agent is perceived, and in what respect it may differ from other pedagogical agents. » Läs mer på s.31

För femte gången arrangerar Institutionen för Ingenjörsvetenskap vid Högskolan Väst Matematikens Dag, en fortbildningsdag för lärare som undervisar i matematik (förskola – gymnasium & komvux). Målet är att erbjuda kompetensutveckling i ämneskunskaper och skapa en möjlighet att knyta kontakter.

Rutiga Familjen presenteras och senaste forskningsresultaten visas:
Datorspel med agenter som lär sig matematik utan symboler

En ny artikel ”A Teachable Agent Game Engaging Primary School Children to Learn Arithmetic Concepts and Reasoning” har publicerats i en internationell tidskrift om intelligenta spel och andra digitala verktyg i skolmiljö: International Journal of Artificial Intelligence in Education av Springer

(eftersom artikeln publiceras internationellt, är den på engelska).
Abstrakt:  In this paper we will describe a learning environment designed to foster conceptual understanding and reasoning in mathematics among younger school children. The learning environment consists of 48 2-player game variants based on a graphical model of arithmetic where the mathematical content is intrinsically interwoven with the game idea. The environment also features teachable agents, which are computer programs that can be taught and behave according to their knowledge. Thus, the environment provides both learning-by-doing (playing the game) and learning-by-teaching (teaching the agent to play). It differs from other learning-by-teaching systems 1) by targeting basic mathematics and primary grade students; 2) by using teachable agents as an extension to educational games in order to leverage engagement, reflection and learning; and 3) by using an agent-driven question dialogue to challenge students’ mathematical thinking, to role-model learner behaviour and to transfer game knowledge to out-of-game mathematics. The teachable agent game is described and evaluated in an authentic classroom study enrolling 443 students from 22 classes in 9 schools. Students range from 2^nd to 6^th grade of mainstream classes and 7^th to 8^th grade for students with difficulties in mathematics. Part of the study was designed as a quasi-experimental study with controls; part was designed to examine students’ change in mental models of arithmetic before and after game play. All students took pre- and post mathematics tests. The 314 playing students used the game and taught their agents during regular math-classes for three months, whereas the control classes attended standard instruction and took the tests. A questionnaire was distributed in the end of the study to investigate students’ perceptions and performances of the agent-tutoring task. Results show that 1) there is a significant learning gain for playing student compared to controls, 2) the learning environment can engage children in advanced mathematical thinking in early education, 3) that young primary students can act as successful tutors. Thus, we conclude that teachable agents is a candidate for leveraging learning in educational games to deeper levels that transfer outside the game. This idea combines the motivational power of games with the reflective power of a teachable agent asking though-provoking, deep questions on the learning material during game play.

Pareto, L. (2014). A Teachable Agent Game Engaging Primary School Children to Learn Arithmetic Concepts and Reasoning. In International Journal of Artificial Intelligence in Education, Volume 24, Issue 3, pp 251-283. Springer New York.

Inom ramen av WGLN är Rutiga Familjen i USA denna vecka. Tre aktiva RF lärare och 2 rektorer från Uddevalla är tillsammans med Lena Pareto på studieresa i Palo Alto, San Francisco. Besöket innebär bl.a. en besöksdag på Stanford University samt besök av tre olika grundskolor och deras klasser.

Vi börjar med en grundlig rundtur på det vackra universitetsområdet. Under lunchen lär vi oss hur det amerikanska grundskolesystemet är uppbyggt för att kunna jämföra med den svenska modellen. Sedan träffar vi vår samarbetspartner Daniel Schwartz och hans kollegor som bl.a. presenterar sitt arbete med Rutiga Familjen.

Rancho Milpitas Middle School är den första skola som vi besöker. Vi bekantar oss med skolans filosofi, träffar Dr. Kimi Schmidt och hennes SO-elever, en årskurs 7-klass som arbetar med algebra samt skolans särskolegrupp.

Sedan besöker vi tre klasser med skolans yngsta elever (5-6 år) på Encinal Elementary School. Dessa klasser fokuserar på läs- och skrivkunnighet. Vi träffar lärarna och eleverna, vi observerar deras koncentrerat arbete på sina iPads, laptops eller med audiomaterial.

Slutligen träffar vi lärarna och eleverna på Ponderosa Elemenatry School. Vi bekantar oss med tre årskurs 3 matematikklasser, där de på olika sätt arbetar med algebra.