CHAPTER 5
Knowledge Sharing Discussion Activity for Teacher Community using a Problem-Solving Strategy
This sub-study designs an online knowledge sharing discussion activity with a
problem solving strategy for the teacher community, and conduct empirical observations.
Quantitative content analysis, lag sequence analysis and original protocol analysis are used to explorethedepth ofteachers’knowledgeconstruction and patternsofdiscussion
behavior among members of the teacher community. Furthermore, the study explores
the influence and limitation of the activity regarding knowledge interaction.
5.1 Discussion Activity
The behavior of knowledge sharing relates to organizational properties (Yang,
2007; Yang & Chen, 2007; Bock, et al., 2005), and for different organizations, such
behavior may differ. In addition to its application in teaching, problem solving is used
for professional development training (Hsu, 2004). Since the previous two chapters
discussed learner communities, different from the scenarios of organizational culture
and knowledge sharing compared to teacher communities, behavioral pattern may
also be different. In view of this, although the theoretical background of the problem
solving strategy, the behaviors of discussion and the assistances towards knowledge
transfer are the same as in Table 7 in Chapter 4, the subject of this chapter targets the
teacher community instead of the learner community in Studies I and II. We designed
an online knowledge sharing activity for the teacher community that incorporates a
problem solving strategy as follows:
Step1: Introducing the concept of sharing teaching knowledge:
Since teachers lack the culture for initiative sharing, this study provided
practical workshops to introduce the importance and theory of knowledge sharing.
This established teachers' fundamental concepts about sharing teaching knowledge.
Step 2: Introducing the online discussion environment for knowledge sharing:
Teachers may be less familiar than students with the application and
operation of online knowledge sharing systems. By introducing the operation
process of the online knowledge sharing environment in practical workshops, teachers’resistanceto technology can beovercomeviaactualpractice,cultivating
theteachers’basicabilities in online discussion.
Step 3: Online Q&A conducted by teachers:
Teachers can send out teaching questions and answers via the system to solve
and discuss questions about teaching practice. The content and categories of
discussion are the same as those described in Table 7 in Chapter 4.
This study collected massive amounts of data over several months of discussion
records for content analysis, sequential analysis and original protocol analysis. By
doing this, the depth of knowledge construction for discussion and the pattern of
problem solving discussion behaviors can be explored.
5.2 Methods
5.2.1 Participants
To observeteachers’knowledgesharing behaviors,weconducted 18 sessions
of workshops in Taiwan from June to August, 2006. The topic of the workshops was
“how to apply ICT on teaching in an e-Learning context.”Thevoluntary participants
in this study were 495 teachers from locations around Taiwan who teach at
elementary or secondary schools.
5.2.2 Tools
5.2.2.1 WIDE-KM System
To observe teachers' online discussion behaviors, an online knowledge
sharing environment must be provided. Here, we added knowledge management
modules for teachers to the WIDE system. WIDE-KM extends the functions of
the original WIDE and adds many new interactive functionsforteachers’online
knowledge sharing to form an integrated environmentforteachers’knowledge
sharing. For a detailed introduction to WIDE-KM, refer to Appendix A.
Figure 8 An example of WIDE-KM teachers’problem solving forum.
Table 16 Coding schemeofteachers’problem solving discussion
Code Phase Example
P1 Propose or clarify questions
In the class, I need to talk to the students about the life and works of author A. Can you tell me what kind of teaching would be appropriate?
P2 Propose or further complete solutions
ThedataIgathered show:Theclimatein authorA’s homeland was so cold that it affected her writing style, filling it with a sense of isolation and helplessness. I often introduce my students to her works through this.
P3 Conduct
comparisons and discussions based on existing solutions
Ido notthink thatthepreviousteacher’scommentwas completesincethatauthor’sfamily background contributed to the sense of loneliness. The author was born to a single family which tried to survive in a war-torn era, thus she seldom felt any human warmth.
This is an important aspect.
P4 Propose summarizing conclusions.
By summarizing otherteachers’commentsand information, I believe that both climate and family background are capable of affecting how a person expresses his or her feelings. When teaching our students, we can describe both points at the same time orask thestudentsto look fortheauthor’sbiographical data, or we can encourage them to try to make
connectionsbetween theauthor’slifeand herworks.I think this is a good way to go.
P5 Other discussions those are irrelevant to the main topic.
One of my friends also lives in the city where author A lived in. He told me the hand-made cookies there are very delicateand delicious…
Weused theteachers’problem solving forum(see Figure 8) of WIDE-KM
as the tool for this observation. The problem solving forum of WIDE-KM is an
online discussion module that allows teachers to conduct Q&A of pedagogic
knowledge, post new question topics and discuss the questions with each other.
The topics are listed clearly so a teacher can click on a topic to access it and see
the content and relevant messages. The teacher can also click on “reply” to
provide responses such as clarifying questions, providing answers, comparing,
discussing, and drawing conclusions. The responses are listed based on the time
posted and are recorded in the database for future access.
5.2.2.2 Coding Schemes
Content analysis and sequential analysis requires coding of each posted
article. This study used the same types of code schemes as Study II, namely the
IAM model for knowledge construction and Table 8 for problem solving.
Between them, the codes and descriptions are the same as Study II. Due to the
differences in discussion scenario and field content, this chapter lists examples of
discussion content particularly related to problem solving for the teacher
community in the Table 16.
5.2.3 Design
This study analyzes the content of the knowledge sharing discussion with both
quantitative and qualitative approaches. Three analysis methods are adopted:
quantitative content analysis, lag sequence analysis and original protocol analysis.
Quantitative content analysis explores the level/depth of teachers’knowledge
construction, and lag sequential analysis analyzes sequential patterns of discussion
behaviors among members of the teacher community. The study also extracts some
original discussion text and processes it with original protocol analysis to trace and
discuss with the derived content/behavior patterns. Discussions, reviews and
comparisons are carried out with the results obtained from qualitative interpretation,
quantitative content analysis and sequential analysis.
5.2.4 Procedures
During the workshops, we not only held lectures about the concepts and
examples related to the workshop topic, but also provided an operation guidebook for
our research tool (the WIDE-KM platform) to the teachers. The researcher explained
how to use the online problem solving discussion forum and gave examples on how
to post and respond to questions, helping the teachers become familiar with the online
knowledge sharing environment. Each teacher was given an account in order to
access the website. Before the workshop concluded, the researcher encouraged the
teachers to actively participate in the Q&A activity of pedagogic knowledge. The
above research procedures were faithfully replicated in all the workshops. The
content of problem solving discussions by the teachers who participated from June to
September, 2006, was then analyzed. There was no intervention or participation from
teacher-trainers or other personnel during the entire discussion process in order to
ensure precise analysis of how teachers conduct online discussions in a free,
unsupervised, and unaffected environment.
5.2.5 Data Analysis
The coding method in our study was the same as Study II: each topic was
treated as a unit (each proposed question is a question-topic), and the messages in
each topic coded based on their temporal order (each topic can have multiple
responses). If a single message has two or more codes, the codes are listed based on
the temporal order (e.g., the first section of a message is coded as P1, the following
two sections as P2, thusthemessage’scodesareP1,P2).Afterallthemessageswere
coded based on the above method, each question-topic had a set of problem solving
coding data. Observation took place from 6/9-9/1, totaling three months, and 133
question-topics were posted. In order to verify the reliability of the coding content,
we randomly picked out 61 topics (near half the topics) for another rater’s analysis.
The inter-rater Kappa reliability of the coding of problem-solving discussion was
0.863 (p<0.01), and that for knowledge construction was 0.422 (p<0.01). Both sets
were statistically significant, and the coded data were put through sequential analysis
and knowledge construction content analysis.
5.3 Results and Discussion
5.3.1 Content Analysis
5.3.1.1 Content Analysis of Knowledge Construction
We gathered a total of 133 topics and 622 posted articles during the activity period and conducted content analysis. 677 codes were yielded after all the discussion content was coded; the distribution of the coded behaviors is shown in Figure 9.
Figure 9 Scale map of the coding of teachers’knowledge construction in the online discussions that integrate a problem solving strategy
From the distribution of codes, we find that knowledge construction in the
process of problem solving-based discussions concentrated mostly in C1
(sharing/comparing information, 89.38%). C3 (negotiation of
meaning/co-construction of knowledge, 6.23%), and C2 (discovery and
exploration of dissonance or inconsistency among participants, 3.68%) took up
about 10%, and C4 and C5 did not appear in the discussions. This shows that it is
hard to achieve the C4 and C5 phases of knowledge construction, as seen in
Studies I and II and other research conducted according to Gunawardena, Lowe &
Anderson’s(1997)coding system (Gunawardena, Lowe & Anderson, 1997; Jeong,
2003). In addition to information sharing and comparison (C1), this research also
found that C3 was the second most common, and C3 was higher than C2. This
reflects that the teachers focus more on further exploring knowledge during
problem solving, in addition to information sharing and comparison, and also
indicates that use of the knowledge sharing discussion activity inspires teachers to
further explore knowledge. We also found that C6 (discussions irrelevant to the
main topic) was rare and only took up 0.71% of messages. This shows that a
discussion with this strategy is less likely to have messages that are irrelevant to
the main topic, and that there is a high level of concentration in teachers’
discussions.
As for the phases of knowledge-construction except C1 (e.g., C2, C3, C4,
C5), however, the width of the discussions was limited (only took up 10%); C4
(testing and modification of proposed synthesis or co-construction) and C5
(agreement statement(s)/application of newly constructed meaning) are the
hardest phases to achieve, and the codes of C4 and C5 were not seen in this case.
5.3.1.2 Content Analysis of Problem-Solving
In order to understand the status of problem-solving from teachers’
discussions, a total of 133 topics and 706 codes were produced after coding by the
scheme in Table 8. The frequency and percentage of each code are shown in
Figure 10.
Figure 10 Scale map of the coding of teachers’problem solving in the online discussions that integrates a problem solving strategy
From the distribution of codes we found that in terms of problem-solving, most of the teachers’discussion behaviors were P2 (67.5%), followed by P1
(21.86%), and P3 (9.9%) was the rarest. P4 was not seen. This indicates that during discussions, most teachers proposed solutions or provided additional information (P2), followed by proposing questions (P1) and making further comparisons and analyses (P3), but they did not compile different solutions or form conclusions (P4). This shows that although the teachers solved or analyzed questions to a certain degree, they lacked integrated analyses and conclusions.
5.3.2 Sequential Analysis
5.3.2.1 Sequential Analysis for Knowledge Construction Behavior
A total of 133 topics contained 622 responses, and 706 C-codes were
obtained in accordance with the IAM coding scheme. We calculated the frequency
of each behavioral category immediately following another behavioral category,
as shown in Table 17. To co determine whether the sequential relationships were
statistically significant, we conducted a sequential analysis on the data. Results
were organized into Table 18, and a behavioral transfer diagram deduced (as
shown in Figure 11).
Table 17 Frequency transition table(C code, Study III)
C1 C2 C3 C6
C1 451 21 44 2
C2 14 4 0 0
C3 33 1 0 1
C6 0 0 0 2
Table 18 Adjusted residuals table (Z-scores) (C code, Study III)
C1 C2 C3 C6
C1 0.39 -5.38 6.98* -12.35 C2 -0.43 3.64* -1.21 -0.41
C3 0.50 -0.50 -1.75 1.34
C6 -1.32 -0.30 -0.39 15.06*
Figure11 Behavioral transfer diagram (C code, Study III)
Figure 11 presents all sequences in Table 18 that reached a level of
significance. Data shown in Table 17 and Figure 11 show the behavior patterns
that occurred during online discussions.
The sequences that reached significance during online problem-solving
discussions are C1->C3, C2->C2 and C6->C6. The sequences C1->C3 and
C2->C2 show that during the discussion based on problem-solving, the
continuation of teachers’knowledge-construction in C2 was significant. This
indicates that teachers tend to provide different perspectives and ideas, and that
when teachers internalize others’discussions, they sometimes provide opposing
views, questions, or different perspectives, and continue to switch between these
tendencies as they externalize the ideas. C1->C3 shows that although C3 was
limited, the teachers would often continue to engage in deeper analyses or
explorations of knowledge content after sharing information (C1->C3). C1->C3
and C2->C2 indicate that a step-by-step behavioral pattern is observable to a
certain extent in teachers’discussions. However, we also found that although C6
(discussion irrelevant to the main topic) was very rare, C6->C6 reached the level
of significance. This shows that once one off-topic message occurs in teachers’
discussions, it becomes more likely for such messages to continue.
5.3.2.2 Sequential Analysis for Problem Solving Behavior
A total of 133 topics contained 622 response messages, and 677 codes were
obtained after coding in accordance with the coding scheme in Table 8. The
frequency transition table is shown as Table 19, and the results are organized into
Table 20; we then deduced a behavioral transfer diagram (Figure 12).
Table 19 Frequency transition table (P code, Study III)
P1 P2 P3 P5
P1 5 132 3 0
P2 20 277 57 2
P3 0 47 7 1
P5 0 0 0 2
Table 20 Adjusted residuals table (Z-scores) (P code, Study III)
P1 P2 P3 P5
P1 -0.71 2.06* -4.54 -1.51 P2 2.73* -2.71 5.93* -1.91
P3 -1.75 0.27 0.14 0.79
P5 -0.30 -1.29 -0.49 14.79*
Figure12 Behavioral transfer diagram (P code, Study III)
Figure 12 presents all sequences in Table 20 that reached a level of
significance. Data shown in Table 19 and Figure 12 provide the pattern of
behaviors during online discussions.
The sequences that reached significance during online problem-solving
discussions are P1->P2, P2->P1, P2->P3, and P5->P5. This indicates that there is
a close sequential correlation between P1 (propose or clarify problem) and P2
(propose solutions), and that during the process of problem-solving and
discussing, the teachers tended to come up with more related questions, question
the solution-providers, or clarify the meaning of the questions after proposing
solutions to the main topic (P2->P1). Moreover, after proposing solutions, they
tended to compare and analyze different solutions and provide information
(P2->P3). These patterns indicate a step-by-step process of problem-solving in
teachers’discussions using problem-solving strategies, and show that teachers are
likely to question solution-providers, clarify the meanings of questions, or come
up with more questions (P2->P1). To some degree, this helps the teachers
repeatedly explore teaching-related knowledge, specifically externalize their
understanding and experience, and help solve problems that are tacit and difficult
to externalize, as mentioned by Carroll, et al. (2003). P5->P5 is similar to C6->C6,
showing that teachers are likely to engage in continued off-topic discussions.
5.3.3 Original Protocol Analysis
In order to further discuss the results of the content analysis and sequential
analysis in the above sections, we provide an excerpt of the questions and
discussions proposed by the teachers numbered n1536 and n1504, as shown in Table
21 and 22. The times when the messages were posted and the codes given by the
rater are also shown. Based on this case study, we can try to interpret and further
understand the processes of teacher knowledge-internalization and externalization
during problem-solving activities and their behavioral patterns inferred earlier:
Table 21 Extraction of an actual discussion example (Teacher no.: n1536, Study III)
Message No. Author Content P-code C-code Time
#7101 n1536 The textbook says that the three primary colors of light are red, green, and blue, but I have heard that these are not the three primary colors used in TV. For example, one of the primary colors used in TV is an orange-ish red. Does anyone know more about this?
P1 C1
2006/7/1 3
15:57:00
#7102 n1529 The three primary colors of light are red, green, and blue. This was tested on academic assessment exams.
P2 C1
2006/7/1 3
16:00:00
#7103 n1531 The colors we commonly see are mostly comprised of two or more colors: The three primary colors are red, green, and blue, and the original colors of the three pigments are magenta, yellow, and cyan. The mixture of the lights of the three primary colors yields white light, and the mixture of the three primary colors yields black. With appropriate proportions, we can use the three primary lights or colors to generate different colors.
P2 C1,C3
2006/7/1 3
16:01:00
Table 22 Extraction of an actual discussion example (Teacher no.: n1509, Study III)
Message No. Author Content P-code C-code Time
#9401 n1504 Is it true that organic vegetables do not need pesticides or net supports and still promise a good harvest?
P1 C1
2006/7/13 16:08:00
#9402 n1209 Air, temperature, moisture, and soil fertility all should be considered. Bugs are also a problem that needs to be solved!
P2 C1
2006/7/14 14:30:00
#9403 n1507 The concept of “organic”is not just limited to vegetables. Fish and meats can also be “organic”
since being organic is a way for us to show that we care about the environment. If you want to know more about organic vegetables, you could join the Fu-chi Teachers’Growth Camp over summer or winter vacation.
It’s a great “organic”camp, and the participants are required to eat only vegetables – organic ones, too! You also get to learn good concepts about education. I recommend it.
P2 C1
2006/7/14 16:08:00
#9409 n1630 Do you want to know how to grow organic vegetables or are you questioning how organic vegetables are achieved? I also want to know how to grow them.
P1 C1
2006/7/17 16:06:00
Let us look at the case in Table 21. After teacher n1536 proposed a question,
n1529 provided his/her initial understanding and information (code P2, C1) (see
#7102 message), but it lacked in-depth analysis and exploration. The message
proposed by n1536 regarding the three primary colors on TV was not even answered.
Afterwards, teacher n1531 found more information regarding this topic (C1, P2) and
conducted a more meaningful exploration and analysis (C3) (see the message of
#7103 on the analysis and additional information on colors, lights, and pigments).
Although this message was not able to completely solve the question of teacher
n1536 about TV’s primary colors or form a conclusive answer (P4), it nonetheless
helped the participants move from knowledge-sharing to in-depth internal
exploration and even to externalize comprehensive answers (C1->C3). However, a
lack of P4 limited teachers’discussions.
The example in Table 22 shows that the question proposed by teacher n1504
was not clearly externalized (message #9401 did not explain to others whether this
teacher wanted to know more about how organic vegetables are grown or the
effectiveness of plantation given certain methods), and this prevented participants
from focusing on a certain topic or analyzing more in-depth the topic of “organic
vegetables”(C3). A more complete answer was thus rendered impossible, which
may have led to teachers n1209 and n1507 only providing related comments and
supplementary information (C1, P2) rather than focusing and more deeply analyzing
the topic. Teacher n1630 then asked a question in order to clarify the original
question (P1) (message #9404). This P2->P1 behavioral pattern refocused the
discussion and increased its depth. It also shows that during a discussion that
involves complicated internalization, the teachers were capable of meta-cognition,
were able to detect the loss of focus or blind spots in a discussion, and could
specifically explain these issues. The two examples have mostly confirmed our
findings about content and sequential analyses.
By summarizing the above case studies, we can see that as for problem-solving
and knowledge-construction, teachers’discussions had the same characteristics as
those of students in Sub-study II, as both groups showed step-by-step knowledge
construction and the process of problem-solving with a certain depth; the teachers
were better in terms of meta-cognition and sensitivity in the discussion process.
However, they were limited in terms of providing integrative solutions (P4), perhaps
due to the fact that the questions asked by teachers were wider and more difficult, so
it was difficult to find the most appropriate solutions.
5.4 Summarizations
In this sub-study, we used problem-solving discussions in the teacher
community; the results of knowledge-construction show that except for C1, the width
of knowledge-construction was limited (C2 and C3 only totaled 9.9% and C4 was not
found). Off-topic messages were rare (C6:0.7%), and the patterns tell us that the
continuity of discussions and knowledge-analysis were high. The C2->C2 pattern
shows that the teachers continued to propose different perspectives in order to deepen
discussions. The percentages of solutions and in-depth analyses were high (P2:67.5%,
P3:9.9, total 77.4%), and the behavioral pattern of proposing answers (P1->P2,
P2->P1, P2->P3) showed some level of rationality and depth, with the teachers
clarifying answers or asking new questions after proposing solutions (P2->P1).
However, integrated solutions (P4) did not occur.