Advanced

Teaching without learning : Is it OK With weak AI?

Månsson, Kristian LU and Haake, Magnus LU (2018) 19th International Conference on Artificial Intelligence in Education, AIED 2018 In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 10948 LNAI. p.241-245
Abstract

Two different learning models for a teachable agent were tested with respect to perceived intelligence, the protégé effect, and learning in Swedish grade 5 and 6 students. A strong positive correlation was found between perceived intelligence and the protégé effect, but no significant differences were found between the two different implementations of the learning algorithm. The results suggest that while the perceived intelligence of the agent relates to the induced protégé effect, this perceived intelligence did not correspond to the implemented learning model. This, in turn, suggest that a simple learning model can be sufficient for a teachable agent system, but more research is needed.

Please use this url to cite or link to this publication:
author
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
keywords
Learning model, Learning outcome, Perceived intelligence, Protégé effect, Teachable agent
host publication
Artificial Intelligence in Education - 19th International Conference, AIED 2018, Proceedings
series title
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
volume
10948 LNAI
pages
5 pages
publisher
Springer Verlag
conference name
19th International Conference on Artificial Intelligence in Education, AIED 2018
conference location
London, United Kingdom
conference dates
2018-06-27 - 2018-06-30
external identifiers
  • scopus:85049369052
ISSN
1611-3349
0302-9743
ISBN
9783319938455
DOI
10.1007/978-3-319-93846-2_44
language
English
LU publication?
yes
id
2a19cb6e-a86e-4736-9c43-9e8d24f7475d
date added to LUP
2018-07-17 11:32:14
date last changed
2019-01-06 14:00:05
@inproceedings{2a19cb6e-a86e-4736-9c43-9e8d24f7475d,
  abstract     = {<p>Two different learning models for a teachable agent were tested with respect to perceived intelligence, the protégé effect, and learning in Swedish grade 5 and 6 students. A strong positive correlation was found between perceived intelligence and the protégé effect, but no significant differences were found between the two different implementations of the learning algorithm. The results suggest that while the perceived intelligence of the agent relates to the induced protégé effect, this perceived intelligence did not correspond to the implemented learning model. This, in turn, suggest that a simple learning model can be sufficient for a teachable agent system, but more research is needed.</p>},
  author       = {Månsson, Kristian and Haake, Magnus},
  booktitle    = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)},
  isbn         = {9783319938455},
  issn         = {1611-3349},
  keyword      = {Learning model,Learning outcome,Perceived intelligence,Protégé effect,Teachable agent},
  language     = {eng},
  location     = {London, United Kingdom},
  month        = {01},
  pages        = {241--245},
  publisher    = {Springer Verlag},
  title        = {Teaching without learning : Is it OK With weak AI?},
  url          = {http://dx.doi.org/10.1007/978-3-319-93846-2_44},
  volume       = {10948 LNAI},
  year         = {2018},
}