International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B6. Istanbul 2004
ID name order directory
3 Aufnahmesysteme 3 inhalt neu/kap2/
36 Einführung 0 inhalt, neu/einführung
35 Physikalische Grundlagen 2 inhalt neu/pg
12 Photographische 1 inhalt neu/kap27
13 Prozess 1 inhalt neu/kap2/
14 Empfindlichkeit 2 inhalt neu/kap2/
15 Farben/Filter “ 3 inhalt_neu/kap2/
16 Farbphotographie 4. inhalt. neu/kap2/
17 Filme 5 inhalt neu/kap2/
Table 1.
4.4 Tests
Web-based learning, as well as conventional learning only is
successful, if the knowledge is applied correctly. Tests make
a fundamental contribution to the assessment of course
participants. During the tutorial students' knowledge is tested
in different stages.
After first authentication, existing knowledge is checked by
solving questions on a multiple choice basis. This gives the
system the ability to relate the user to one of three classes -
beginner, average student, advanced student. The more tests
the user has fulfilled, the more precise this classification will
be. In chapter 4.5 you will get more information about the
assessment's calculation.
Next, small tasks must be solved while going through the
lessons. If these questions are answered incorrectly, the
system gives support to find the correct solution. In the
expert's advice table the information where to look for
additional support is stored. Possibly the student has to repeat
the test. How the system tries to avoid putting same questions
again is explained in chapter 4.6. These pre-tests have no
influence on learners' assessment, but control the sequences
of learning modules.
After completing main chapters, new tests follow which
decide on participants marking. During a final test, covering
the entire area of remote sensing, the learner has to prove his
knowledge. Comparable to an examination situation, the
questions arc adjusted to the students' level of achievement.
question
question log ID metadata q
ID chapter ID ID
user ID question question ID
question ID level author
date keywords
time date
answer flag
state answer
advice ID
ID question ID
question ID correct, flag
directory answer
Figure 5. Layout of the question database.
chapter. ID name long
0 Aufnahmesysteme fuer Luft und Satellitenbilder
Oo à
0 Einführung
0 Physikalische Grundlagen der Fernerkundung
3 Photographische Aufnahmesysteme
12 Photographischer Prozess
Spektrale Empfindlichkeit photographischer Schichten
12 Farben und Filter
12 Farbphotographie
c3 c e © c © c
—
N
12 Filme zur Luftbildaufnahme
Example of a MySQL user table, representing a small part of chapter 2, data acquisition techniques.
Of course, this could only be achieved, if also test questions
are handled in a database. Questions are classified into three
levels. Figure 5 gives an impression of the tables used. Please
recognize the table for expert's advice, if a question was
answered incorrectly. The questions the students tried to
solve are registered in the protocol table as well as the time
needed and if the question was answered successfully.
4.5 Control of navigation
Administrating the content and tests in databases is a prior
condition for adapting the training course to the student's
skills. As mentioned before, the course allows an
achievement oriented, adapted navigation. The sequence of
actions is controlled by tests, which need to be analysed
carefully. As measure affecting the workflow the weighted
quotient of questions answered correctly in relation to all
answers is calculated. See Table 2 as example:
Number Level of Difficulty State Score Marat |
5 3 correct 15 15
2 3 false 0 6
0 2 correct 0 0
2 2 false 0 4
4 1 correct 4 4
4 1 false 0 4
17 Sum 19 33
Table 2. Performance ratio - example.
You can see, that 17 questions were answered in total and the
student solved 9 correctly, 8 answeres were wrong. Among
the 9 correct answeres, 5 were weighted with 3 and 4 with
factor 1. Taking no notice of the weights, the quotient is
calculated as 9/17=0,52. If the level of difficulty is
considered, a better result is allocated (19 / 33 = 0,58). This
weighted performance factor is calculated after every test and
is stored in the user database (see Figure 3).
According to this quotient, the student is guided through the
lecture. Appropriate to his knowledge, he is assigned to one
of three classes (beginner, average, advanced). Nevertheless
the system allows that the student also answers questions of a
different level of difficulty. Two indices, defined by the
course author, regulate the probability of how the questions
are distributed. As seen in Figure 6, experts could expect
20% questions to be easy, 30% to be medium and 50% to be
difficult, whereas a beginner will expect a ratio 50:40:10.
