HELP DESIGN AND HELP USE IN A COMPUTER BASED

LEARNING ENVIRONMENT

J. C. Sakdavong, F. Amadieu, N. Huet and C. Dupeyrat

CLLE-LTC, UMR CNRS 5263, Université Toulouse le Mirail, 5 Allée A. Machado

31058 Toulouse Cedex 1, Toulouse, France

Keywords: Learning, Distance education, Computer supported education, ICT, AIED, Modelling, Help, Guidance,

Metacognition, Metacognitive guidance, Multiagent system.

Abstract: This paper presents the framework of elaboration of help in a text-processing computer-based learning

environment. A multiagent software system have been developed including help. Then it was tested in

participants who had to learn autonomously and for which we analyzed the behavioral pattern of using help.

Results showed a lack of regulation and highlight the need for adding a metacognitive guidance enabling

learner to efficiently use help.

1 INTRODUCTION

The following study has been sponsored by the

French National Agency for Research (ANR) for a

period of four years from 2006 to 2011 (ref.

CEAGMATIC).

For several years, the use of new technologies in

higher education and vocational training were

widely developed but did not inevitably entail an

improvement of learning. Several studies showed

that computer learning environments lead to poor

learning (e.g., Dillon & Gabbard, 1998; Shapiro &

Niederhauser, 2004).Furthermore, there is also a

very high dropout rate in distance training

environments (Lebel, 1995).

One of the reasons for this failure is the lack of a

fine analysis of the sequence of learning, the sub-

tasks required to acquire the contents of knowledge.

From an ergonomic point of view the learning

programs using new technologies were too often

elaborated without modelling the task of learning

that is without an analysis of the cognitive demands,

the types of potential errors associated with a

modelling of the activity of the learner (e.g.,

Boucheix, 2003 ; Leplat, 2002 ; Samurcay & Pastré,

2004).

In the present study, we will present how we

elaborated the learning sequence taking into account

the previous points.

Another reason which can contribute to this

failure lies in the fact that the educational and

technological tools require learners to take charge of

their own cognitive learning in autonomy for which

they are not all necessarily prepared (Azevedo &

Cromley 2004). This problem occurs particularly in

distance training (Lebel, 1995). To resolve these

difficulties of autonomous learning, it may be useful

to provide learner with helps (for example,

glossaries, work-out exercises) for learning in their

own (Baker, Puustinen & Lund, 2002). However,

learning outcomes partly depend on the quality of

the help provided (Puustinen, Volckaert-Legrier,

Coquin & Bernicot, 2009). If the help does not meet

the learners’ needs, it will not be used or will not

improve learning. Taking into account an analysis of

errors made on a task in help designing enables to

meet learner’s needs and increase the relevance of

the help provided. In the present study we will

describe how we elaborated help in this sense.

Research showed that help is often unsuited for

learners’ needs and rarely or inappropriately used

(e.g., Aleven et al., 2003; Puustinen et al., 2009;

Roll et al., 2005).

Furthermore, the appropriate use of help is

considered in the literature as a self-regulated

strategy (Puustinen, 1998). Self-regulation requires

metacognitive knowledge (e.g., Flavell, 1979), the

knowledge that individuals have of their own

knowledge, perceptions of difficulty, utility and cost

and the relative efficiency of strategies. According

304

C. Sakdavong J., Amadieu F., Huet N. and Dupeyrat C..

HELP DESIGN AND HELP USE IN A COMPUTER BASED LEARNING ENVIRONMENT.

DOI: 10.5220/0003345503040311

In Proceedings of the 3rd International Conference on Computer Supported Education (CSEDU-2011), pages 304-311

ISBN: 978-989-8425-49-2

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

to some authors (e.g., Nelson-Le-Gall, 1981), an

awareness of the need of help-seeking followed by

the decision of seeking help depends on an

individual’s metacognitive knowledge about the

utility and cost of help-seeking. According to

literature on metacognition, metacognitive

knowledge or perceptions of the utility and cost of

using strategies enabling a gain in performance is

positively associated to strategy use (e.g., Pressley et

al. 1985; Escribe & Huet, 2005). In an interactive

learning environment of statistics, we found that the

more students perceived the utility of on-line course,

of worked-out problem, the more they used it

(Noury, Huet, Escribe, & Narciss, 2008).

However, even when help is perceived as useful,

in some studies it was found that help is seldom used

(Hofer & al., 1996).

Furthermore, studies on help-seeking and

learning showed divergent results. Some authors

found that an increase of help used was not

associated with an increase in performance (e.g.,

Clarebout & Elen, 2009), others found a positive

relationship between the use of help and

performance (Jiang, Elen & Clarebout, 2009). These

divergent results reveal that the process of help-

seeking is complex. The help-seeking process can be

considered as a continuum from low to high self-

regulation (Puustinen & Rouet, 2009). Several

characteristics of self-regulated help-seekers or users

have been identified (Puustinen et al. In press). A

high self-regulated help-seeker is illustrated by: (1) a

high awareness of the need for help as indicated by a

high relation between confidence judgment in the

accuracy of the answer and performance (2) the high

rate of time before selecting help as an indicator of

the necessity of the need for help; (3) an

instrumental help selection (asking only for an

explanation or a general procedure or solving

principle and trying to solve the problem by oneself)

rather than an executive help (that is ready-made

answers); (4) the ability to re-invest the first use of

help to the similar exercises.

Within the context of Interactive Learning

Environment (ILE), the aim of the present study is

1) to show how we have elaborated help on the basis

of a previous analysis of errors made by the learners;

2) to explore the use of help in the new device and in

relation with perceptions of help utility and

performance.

2 METHOD

2.1 Materials

2.1.1 The Text-processing Computer-based

Learning Environment

The environment consisted in a software system for

learning how to use word processor formatting and

style-sheets. The learning material consisted of

screencasts dealing with word processor’s

procedures and application exercises. Within this

environment; the participants were not allowed to

transgress the sequence. This environment resulted

from a previous exploratory study carried out on 80

students (Sakdavong, Huet & Adreit, 2009). The

instructional screencasts were designed especially

for the experiment. Each screencast was composed

of a narration from the teacher and either an

explanation screen or a video of a word processor’s

procedure. These screencasts dealt with formatting

and using style-sheets. They were composed of

explanations of declarative and procedural

knowledge accompanied by examples. The

screencasts duration was about half an hour.

After viewing the screencasts, participants had to

deal with three exercises. These exercises were

provided by the computer-based learning

environment through two windows: one presented

the subject of the exercise and the other was a fully

functional generic word processor. All actions from

the participants in these two windows were recorded

by the computer system.

Each exercise was composed of two questions:

an initial question and a checking question.

Exercise 2 was more complex than exercise 1 in

that it required to master all of the procedures for

formatting and using style-sheets.

After reading the subject of a question, each

participant had to answer to it in the word processor

window.

For example, in the second exercise, the word

processor was presenting a small text from Molière

and the initial question was:

You have to modify the “default”

style in order that paragraphs are

in Times New Roman font, size of 10

points, in italics and with

justified paragraphs. You have to

modify the “title 3” style in order

that …

Figure 1 shows the appearance of the two

windows shown to each participant.

HELP DESIGN AND HELP USE IN A COMPUTER BASED LEARNING ENVIRONMENT

305

Figure 1: The Learning Environment.

Once a participant had validated her/his answers,

she or he had to assess how confident he or she was

in the accuracy of the answer and then, received a

feedback. If the answer was right, the participant had

to go on to the checking question. If wrong, the

participant was invited to accept or refuse to get

help. According to her/his choice, the participant

was getting help or not, then she/he had to go to the

checking question.

This checking question differed from the initial

question in the following aspect: it was not

associated with help or feedback. It had two roles: if

the participant had failed the initial question, the

checking question tested if the participant had

transferred the chosen help; if the participant had

been successful with the initial question, the

checking question tested if she or he really knew

how to solve the question or if she/he succeeded by

chance. A great care has been taken to ensure the

two questions were equivalent in difficulty.

2.1.2 The Provided Instrumental and

Executive Helps

Help was only provided after a failure on the initial

question of each exercise.

Figure 2 shows the general scheme of each

exercise.

If the participant accepted to receive help, she/he

was invited to select one of the four following

options: (1) “Go back to see the lesson in relation

with the error I did”; (2) “Read a text describing the

error I did”; (3) “Read a text describing the

Figure 2: Scheme of one exercise.

general procedure to solve the exercise”; (4) “Watch

a video showing the exercise’s solution”. Then

participant had to go to the checking question.

Help #1 to #3 were instrumental helps (Arbreton,

1998) and help #4 was an executive help.

Participant could only choose exactly one help

once she/he accepted to get help.

The instrumental help #1 was generally a link to

one of the first instructional screencasts. The

instrumental help #2 was a simple text identifying

the error but not how to solve it. The instrumental

elp #3 was the correct procedure to answer the

question, presented as an algorithm.

The executive help #4 was a screencast of the

correct solving procedure. This screencast was

without audio commentary.

The errors with their associated help have been

selected thanks to our previous exploratory study

carried out on 80 students (Sakdavong, Huet &

Adreit, 2009): we made a statistical analysis of the

most common and recurrent errors in a word

processor, and have selected the most difficult

exercises and interesting errors. Then we have

designed a general scheme of help in two steps: first

step to know if participant wanted help; second step

to know which type of help she or he would like.

The computer environment device was a

software system not for learning how to use word

processor formatting and style-sheets but to record

and analyse participant’s behavior in order to

provide data for the psychologist and automatic

feedback for the participant.

This device has been implemented as a multi-

agent system (Wooldridge, 2002). Thanks to this

choice, we had software elements (agents) able

Selection of one hel

Acce

tance Avoidance

Failure Success

Initial question

Proposition of getting help

Proposition of 4 types of help

Consulting of help

Checking question

Consultation

CSEDU 2011 - 3rd International Conference on Computer Supported Education

306

intrinsically to observe the activity and to produce a

behavior, also to communicate between them. The

system was very dynamic: creating agents during the

learner's behavior (for example, creating a new

helping agent when an exercise starts) or modifying

in real time the behavior of agents (for example, an

helping agent can change of behavior according to

an evolution of the learner’s profile) were very easy.

We have used the framework JADE

(Bellefemine, Cairo, Trucco and Rimassa, 2004) and

programmed the agents in the Java language.

One of the agents role was to record the learner’s

behavior as a sequence of actions (the activity

graph), we called it “Historical” agent. It records the

activity with respect to the full teaching scenario (for

example, it records which help the participant has

chosen in case of failure in solving a question; it

records the consultation time of help or the

processing time of one exercise), to the word

processor agent (for example, the learner selects a

paragraph then clicks on the shortcut button

“centering the paragraph”).

A second important agent role was the

“Scrutinizer” one, it made it possible to observe and

analyze the activity of learning. These agents are in

charge to identify characteristic behaviors. Such

agents are created specifically for each exercise.

They had a mechanism of subscription which

enabled them to receive from the “Historical” agent

the sequences of actions which they were in charge

to analyze. According to their analysis, they created

“Helper” agents or will communicate with the

existing “Helper” agents. They could also

communicate with the “Profile” agent in charge to

dynamically adapt the profile of the learner.

The “Helper” agents were the most important

ones: they provided the assistance by giving

feedback, proposing and displaying help. The

combination of “Scrutinizer” agents and “Helper”

agents while they gave a typical feedback were

acting as classical work on intelligent tutoring (e.g.,

Wood, Bruner& Ross, 1976) in which the tutor

decide which help providing on the basis of an

analysis of frequently made errors (Puustinen,

Volckaert-Legrier, Coquin & Bernicot, 2009).

In order to allow our team to analyze each

participant results, the device was generating XML

and CSV files containing all records.

2.2 Mesures

2.2.1 Prior Knowledge Test

In order to control previous experiences and

familiarity with text processing programs, each

participant was asked to report what were the text

processing programs he/she knew and their

frequency of use. The frequencies of use were very

extended among participants (from many times a

day to sometimes a year).

In addition, they were given a pre-test of seven

multiple choice questions to determine their level of

prior declarative knowledge in principles of text

processing programs (5 answer choices including the

choice “I don’t know” to avoid a random answer

bias). The questions dealt with shared procedures

and functions of text processing programs. They

focused on formatting and using style-sheets with

text processing programs. Each correct answer was

scored by 1 point. An example of a prior knowledge

question was as follows: Justifying text means

(Expected answer: Aligning text on both left and

right sides). Cronbach's Alpha for the prior

knowledge test was 0.60. The mean proportion of

correct answers of 49.23 % (SD = 26.00) highlighted

a large heterogeneity of knowledge within the

participants group.

2.2.2 Confidence Judgment and Perceived

Utility of Help

After finishing the question and before receiving a

feedback, the participant had to assess how

confident he or she was in the accuracy of the

answer on scale from 0 (not at all confident) to 100

(very confident).

For each consulted type of help, five questions

assessed the perceived utility of the concerned type

of help. The questions were designed on the basis of

existing measures of perceived utility (Davis, 1989;

Venkatesh, Morris, Davis, & Davis, 2003). The

questions were adapted to the learning and problem

solving aspects of the task. Participants rated their

answer on a scale from 0 “not at all” to 100

“totally”. A mean score was computed on the basis

of the five questions.

2.2.3 Actual use of Help and Solving

Behaviours

On-line recording tools were used to get data of the

exercise solving behaviors: accuracy of the answer,

time to decide to accept or refuse help, solve

exercise, solving procedure, chosen type of help

(executive or instrumental.), acceptance or refusal of

help. The recording tools were the agents presented

above. Each student action was recorded and/or

processed by the system. The system was reactive to

HELP DESIGN AND HELP USE IN A COMPUTER BASED LEARNING ENVIRONMENT

307

these actions and the system’s actions were recorded

too.

2.3 Procedure

Participants were tested in groups in a spacious

computer room, in order to prevent them seeing

what the other students were doing. The duration

was approximately 90 min. They performed

individually the tasks and they were not allowed to

use material other than the computer-based learning

environment.

Participants first filled out the template assessing

their past experiences with text processing program

use and their declarative knowledge about

formatting and using style-sheets. Next, the learning

step consisted in watching the instructional

screencasts about formatting and style sheets. The

main part of the experiment was the problem solving

step. Participants were instructed to perform three

types of exercises implying formatting and use of

style sheets. In order to familiarize participants with

the problem solving tasks, the exercises device and

the types of available help were presented through a

screencast. This exercise was designed to promote

participants’ familiarity with the exerciser

environment. The first exercise required formatting

procedure of a given text. The second exercise

required to use and change a style sheet to process a

given text. Each exercise was made of two

equivalent group of questions. Each participant was

first getting the subject of the first question, and then

was allowed to try to answer to it with the integrated

text processing program. When participants

indicated that they had finished, the system decided

if the exercise was correctly solved according to the

grid we have designed. In case of wrong solving, the

system proposed participants to use help.

Participants were allowed to reject help. Participants

accepting help had to make a choice among different

types of help. After consulting help, participants

dealt with the second question of the current

exercise (see figure 2).

At the final step, participants indicated their

perceived utility of the environment’s help.

3 RESULTS

3.1 Descriptive Statistics

Table 1 provides descriptive statistics for

performance and help use on the overall sample (n=

65).

Concerning performance, results showed that in

the initial question of exercise 1, a little more than

fifty percent of participants succeed. Among those

who failed, 44,83 % refused help (i.e., they

expressed that they do not need or do not want help);

55,17% accepted to select help. Concerning the

initial question of the second exercise, most of the

participants failed, and among them 76% accepted

help, 24 % refused help.

Table 1: Descriptive statistics for performance and help

use for the initial question of exercises.

N %

Exercise 1

Success

Failure

Help refusal

Help selection

55,4

44,6

24,6

Exercise 2

Success

Failure

Help refusal

Help selection

23,1

76,9

18,5

58,5

Concerning the kind of help selected, nearly all

of the participants selected an instrumental help: 14

out 16 for exercise 1 and 35 out 38 for exercise 2.

Concerning the kind of error made in the initial

question of exercise 1, most of the participants did

an error of forgetting a sub-goal (13 participants out

of 29); some participants (10 out 29) did an error not

identified by the help system (i.e., an error for which

no specific help had been elaborated, only general

help could be provided) and a minority did a wrong

choice of procedure (6 out 29).

Concerning the initial question of exercise 2,

most of participants (31 out 50) did a wrong choice

of procedure, the others forgot sub-goal (11 out 50)

or made an error not identified by the system (8 out

50).

3.2 Efficacy of Help used

In exercise 1, the acceptance or the refusal of help

was not related to performance; χ²= .144; p >.05.

Among those who refused help, 69.2 % failed again

in the checking question; 30.8% succeeded. Among

those who accepted help, only 37.5 % took benefit

from the help by succeeding in the checking

question. The other participants, although they

selected a help, they did not take benefit from it and

failed again in the checking question (62.5%).

CSEDU 2011 - 3rd International Conference on Computer Supported Education

308

Similarly to the previous exercise, in exercise 2,

the acceptance or the refusal of help was not related

to performance; χ²= .93; p >.05. Among those who

refused help, 92 % failed again in the checking

question; 8% succeeded. Among those who accepted

help, only 20.6% took benefit from the help by

succeeding in the checking question. The other

participants, although they selected an help, they did

not take benefit from it and failed again in the

checking question (79.4%).

Overall, these results showed that the help

selection did not lead the majority of learners to

succeed in a similar question.

A mean rate of the perceived utility was

computed for each learner who consulted help (N =

45 participants who used at least once a help in at

least one of the two exercises). The learners

generally perceived a positive utility of the consulted

help (M = 61.14; SD = 24.68).

For exercise 2 (checking question), those who

succeeded after consulting help displayed a higher

perception mean of the utility of the help (M=72.71;

SD=18.65) than those who failed (M=57.72;

SD=24.96), t(42)= 1.94, p<.05). Because, the score

of perceived utility was assessed after executing all

of the exercises we did not compute the relationship

between performance at exercise 1 (checking

question) and the utility perceived of the help used

in all of the exercises.

However, high perceived utility of help was

linked to a high confidence-judgment of success to

solve the checking exercise r(45) = .47, p < .01). In

other words, a positive confidence judgment of

performance was associated to a high perceived

utility of help consulted to perform the checking

question.

3.3 Individual Differences in

Self-regulated Learning

The analysis of individual differences in self-

regulation of behavior on the most complex exercise

(exercise 2) revealed that:

-15 participants succeed both the initial and the

checking question).

-4 participants succeeded in the initial question

but failed in the checking question

- Among the participants who failed in the initial

question, 12 participants refused the help and 11 out

of 12 failed in the checking question, considered as

bad regulated help-seekers.

- Among the participants who failed in the initial

question, 34 participants accepted help and only 7

succeeded in the checking question, showing that

they were self-regulated help-seekers. The other 27

were bad regulated help-seekers.

These behavioral patterns showed that the majority

of the participants lack of self-regulated skills and

consequently cannot appropriately use the helps.

Analysis of variance were computed between the

three groups who failed (see Table 2) (we excluded

one group who was composed of only one

participant who failed at the initial question, refused

help and succeeded at the checking question) on the

mean time duration of the decision to accept or

refuse help on the exercise 2. Results were not

significant, F(2,42)=1.30, p>.05. The self-regulated

group (N=7) did not differ from the other two

groups. The five behavioral pattern did not differ on

the confidence judgment in the answer at exercise 2

(initial question), F(4,59)= 1.27, p>.05 (see Table 2).

Table 2: Behavioral pattern and descriptive statistics of

confidence judgment in the answer at exercise 2.

Behavioral pattern N M SD

Failed initial q:

Help refusal-

failure checking q.

Help accepted-

failure checking q

Help accepted-

success checking q

Success initial q:

Success checking q

Failure checking q

56.18

54.63

56.43

75.33

63.25

36.42

31.33

26.64

22.17

33.60

Note. q: question

4 CONCLUSIONS

The aim of this study was to present the design of

the helps provided by a multi-agent system based on

a previous analysis of errors. Then, a second aim

was to explore how learners behaviour in a device

with help and especially are they self-regulated help-

seekers?

Results showed that taking charge of one’s learning,

even with a computer-based learning environment

with relevant helps, is not easy for most of the

learners. They lack of accurate monitoring:

judgment of confidence scores should be less high

when they failed. An important amount of them

refused help when needed, and consequently failed.

Some of them used help but inefficiently although

they selected an instrumental help, which is

HELP DESIGN AND HELP USE IN A COMPUTER BASED LEARNING ENVIRONMENT

309

considered as an adaptive help for learning. They

perceived the utility of the help but did not take

benefit from it.

These results are in the line of those found in the

literature and highlight the need for helping learners

to select help when needed and to be self-regulated

helping-seeker. In this perspective, the next step of

our study will be to conceive a metacognitive

guidance. If learners do not use correctly the helps, it

is because they have a lack of metacognitive

abilities: they do not regulate correctly their learning

behavior.

Two types of metacognitive guidance will be

proposed (double-blind, 2009):

• A suggested guidance that the learner can

accept or refuse

• An imposed guidance if the system identifies

a recurrent metacognitive mistake or lack.

This made the principal originality of our

approach from the point of view of psychology.

Then:

• The two types of metacognitive guidance will

be included in the multi-agent system as new

helping agents

• The online course including guidance will be

tested over many groups of learners and the

multi-agent system will again record all the

learners’ behavior into graphs

• Psychologists will analyze the graphs of

learners’ behavior in order to check if our

metacognitive helps are useful.

REFERENCES

Dillon, A., & Gabbard, R., 1998. Hypermedia as an

educational technology: A review of the quantitative

research literature on learner comprehension, control,

and style. In Review of Educational Research, 68(3),

322–349.

Shapiro, A., & Niederhauser, D.,2004. Learning from

hypertext: Research issues and findings. In D. H.

Jonassen (Ed.), Handbook of research on educational

communications and technology (pp. 605–620).

Mahwah, NJ: Lawrence Erlbaum Associates.

Lebel, C., 1995. Le tuteur et l’autonomie de l’étudiant à

distance. In Journal of Distance Education, vol. X, n°

1, pp. 5-24.Baker, M., Puustinen, M., & Lund, K.,

2002. Assessment of existing concepts: Deliverable

D.2.C.1. VIRTU@LIS. In Project Report, EU FP5

Contract IST-2000-28121. GRIC, UMR CNRS 5612,

University of Lyon 2, France.

Boucheix, J. M., 2003. Ergonomie et Formation :

Approche d’ergonomie cognitive des apprentissages

professionnels. In Psychologie Française, 48(2), 17-

34.

Leplat, J., 2002. Psychologie de la formation, jalons et

perspectives. In Toulouse: Octares 2002.

Samurçay, R. & Pastré, P., 2004. Recherches en

didactique professionnelle. In Toulouse : Octarès 2002

Azevedo, R. & Cromley, J.G., 2004. Does training on self-

regulated learning facilitate students'learning with

hypermedia? In Journal of Educational Psychology,

96, 523-535.

Puustinen, M., Volckaert-Legrier, O., Coquin, D. &

Bernicot, J. 2009. An analysis of students’

spontaneous computer-mediated help seeking: A step

toward the design of ecologically valid supporting

tools. In Computers & Education, 53(4), 1040-1047.

Aleven, V., Fischer, F., Schworm S., Stahl E., Wallace R.

2003. Help Seeking and Help Design in Interactive

Learning Environments. In Review of educational

Research, 73, 3, 277-320

Roll, I., Baker, R.S., Aleven, V., McLaren, B., &

Koedinger, K., 2005. Modeling students’

metacognitive errors in two intelligent tutoring

systems. In L.Ardissono, P. Brna and A. Mitrovic

(Eds.), Proceedings of the 10th International

Conference on User Modeling, UM'2005, 379-388.

Berlin: Springer-Verlag.

Puustinen, M. 1998. Help-seeking behavior in a problem-

solving situation: Development of self-regulation. In

European Journal of Psychology of Education, 13,

271-282.

Flavell, J., 1979. Metacognition and cognitive monitoring:

A new area of cognitive-developmental inquiry. In

American Psychologist, 34, 906-911.

Nelson-Le Gall, S., 1981. Help-seeking: An understudied

problem-solving skill in children. In Developmental

Review, 1, 224–246.

Pressley, M., Borkowski, J. G. and O’Sullivan, J. T., 1985.

Children’s Metamemory and the Teaching of Memory

Strategies. In Forrest-Pressley, D. L., MacKinnon, G.

E. and Waller, T. G.(eds.), Metacognition, Cognition

and Human Performance, pp. 111-153. New York:

Academic Press.

Escribe, C., Huet, N., 2005. Knowledge accessibility,

achievement goals and memory strategy maintenance.

British Journal of Educational Psychology, 75, 1-

19.

Noury, F., Huet, N., Escribe, C., & Narciss, S., 2008.

Metacognitive knowledge, achievement goals and help

seeking in learning on a web site of statistic problem.

In 3rd Biennial Meeting of the EARLI Special Interest

Group 16, Ioannina.

Hofer, M., Niegemann, H. M., Eckert. A., & Rinn, U.,

1996. Padagogische Hilfen fur interaktive

selbstgesteuerte Lernprozesse und Konstruktion eines

neuen Verfahrens zur Wissensdiagnose [Instructional

help for interactive self-directed learning processes

and construction of a new procedure for knowledge

diagnosis]. In Zeitschrift fur Berufs- und

Wirtschaftspddagogik Beiheft, 13, 53-67.

Clarebout, G. & Elen, J., 2009. The complexity of tool use

in computer-based learning environments. In

Instructional Science, 37, 475-486.

CSEDU 2011 - 3rd International Conference on Computer Supported Education

310

Jiang, L., Elen, J., Clarebout, G., 2009. The relationships

between learner variables, tool-usage behaviour and

performance, Computers. In Human Behavior, v.25

n.2, p.501-509, March, 2009

Puustinen, M., & Rouet, J.-F., 2009. Learning with new

technology, Help seeking and information searching

revisited. In Computers & Education, 53, 1014–1019.

Puustinen, M., Bernicot, J., & Bert-Erboul, A. (In press).

Written computer-mediated requests for help by

French-speaking students: An analysis of their forms

and functions. In Learning and Instruction.

http://dx.doi.org/10.1016/j.learninstruc.2010.07.005

Sakdavong, JC., Huet, N., Adreit, F.

Help design for the metacognitive guidance of the learner:

A proposition of computer-based system, In CSEDU

2009, Lisbonne, 23-26 mars 2009, ISBN 978-989-

8111-82-1, pp 210-216.

Arbreton, A., 1998. Student goal orientation and help-

seeking strategy use. In S.A. Karabenick (Ed.),

Strategic help seeking: Implications for learning and

teaching (p. 95–116). Mahwah, NJ.: Erlbaum.

Wooldridge M., 2002. An introduction to MultiAgent

Systems, In John Wiley & sons, Ltd, 2004.

Bellefemine, F., Caire, G., Trucco, T., Rimassa, G., 2004.

Jade programmer’s guide (jade 3.2). In Technical

Report, TILab S.p.A., 2004.

Wood, D., Bruner, J.S., Ross, G., 1976. The role of

tutoring in problem solving. In Journal of Child

Psychology and Child Psychiatry, 17, 89-100.

Davis, F. D., 1989. Perceived usefulness, perceived ease

of use, and user acceptance of information technology.

In MIS Quarterly, 13(3), 319–340.

Venkatesh, V., Morris, Davis, & Davis, 2003. User

acceptance of information technology: Toward a

unified view. In MIS Quarterly, 27(3), 425–47

HELP DESIGN AND HELP USE IN A COMPUTER BASED LEARNING ENVIRONMENT

311