peter shea's community of inquiry ii: new research on the future of online learning
DESCRIPTION
SLN SOLsummit 2009 presentation - by Peter Shea http://slnsolsummit2009.edublogs.org New Approaches in Online Learning Research This presentation will share new research towards a theory of online teaching and learning. This ongoing investigation employs novel research methods including quantitative content analysis, social network analysis, and cluster analysis to further our understanding of pedagogical, social, and cognitive processes essential for quality online education.TRANSCRIPT
Community of Inquiry II:New Research on the
Future of Online Learning
SLN SOLsummit MeetingFebruary, 2009
Peter SheaSLN Senior ResearcherEducational Theoryand Practice & CCIUniversity at Albany
Overview
• New research towards a theory of onlineteaching and learning.
• Research methods including– quantitative content analysis, social network
analysis, and cluster analysis• Goal: further our understanding of
pedagogical, social, and cognitiveprocesses essential for quality onlineeducation.
Overview: Community of InquiryFramework
> 1000 papers refer toCoI foundational studies
Community of Inquiry Framework
Social PresenceThe ability ofparticipantsto identify with thecommunity (e.g., courseof study), communicatepurposefully in a trustingenvironment, anddevelop inter-personalrelationships by way ofprojecting theirindividual personalities.
Cognitive PresenceThe extent to whichlearners are able toconstruct and confirmmeaning throughsustained reflectionand discourse in acritical community ofinquiry.
Teaching PresenceThe design, facilitation and direction of cognitive and social processes for the purpose ofrealizing personally meaningful and educationally worthwhile learning outcomes
Elements,Categories &Indicators
ELEMENTS CATEGORIES INDICATORS
(examples only)
Social Presence Open Communication Learning climate/r isk-free expression
Group Cohesion Group identity/ collaboration
Personal/Affective Self projection/e xpressing emotions
Cognitive Presence Triggering Event Sense of puzzlement
Exploration Information exchange
Integration Connecting ideas
Resolution Appl ying new ideas
Teaching Presence Design & Organization Setting curriculum & methods
Facilitating Discourse Shaping constructive exchange
Direct Instruction Clarifying, explaining, demonstrating
Questions
• Do instructor skills in teaching presencefoster students’ social presence?
• Do student experiences of teaching andsocial presence “predict” their experienceof cognitive presence?
Teaching Presence IndicatorsWhen the online instructor:
• Communicates course topics, goals, and due dates• Provides clear instructions• Helps students clarify their thinking• Keeps students on task, engaged, and participating• Encourages students to explore new ideas• Focuses discussion on issues that aided student learning• Establishes a sense of course community• Provides explanations or demonstrations to help learners better
understand the content• Gives helpful feedback• Asks for feedback on how the course could be improved
It leads to…
Social PresenceStudents:• gain a sense of belonging in the course• form distinct impressions of course participants• find online communication an excellent medium for social
interaction• are able to identify with the thoughts and feelings of other
students• feel comfortable conversing online• feel comfortable participating in discussions• feel comfortable interacting with course participants• feel comfortable disagreeing with others• online discussions helped students develop a sense of
collaboration• feel their points of view are acknowledged by others
Which in turn leads to…
Cognitive Presence• Course activities pique curiosity• Problems posed increase interest in course issues• Students feel motivated to explore content related topics• Students brainstorm & find relevant information to aid them in
resolving questions• Online discussions help students appreciate different perspectives• Combining new information helps students answer questions• Learning activities help students create solutions• Reflection on course content & discussions help students
understand fundamental concepts• Students can describe ways to test & apply their new knowledge• Students develop solutions to course problems that can be
applied in practice• Students can apply knowledge created in their courses to work or
other non-class related activities
Structural Equation Modeling
TeachingPresence
q12.38
q11.28
q10.24
q9.27
q8.18
q7.23
q6.19
q5.24
q4.53
q3.43
q2.40
q1.37
.89
1
.85
1
.87
1
.85
1
.90
1
.881
.901
.87
1.69
1
1
.78
1
.80
1
q13.46
.74
1
CognitivePresence
q34 .44
q33 .39
q32 .40
q31 .27
q30 .27
q29 .29
q28 .49
q27 .44
q26 .47
q25 .34
q24 ..36
q23 .48
.75
1
.78
1
.781
.851
.86
1.84
1
.72
1
.75
1
.721
.81
1
.80
1
.72
1
SocialPresence
q22
.40
q21
.38
q20
.41
q19
.18
q18
.29
q17
.27
q16
.62
q15
.67
q14
.65
.77
1
.78
1
.77
1
.91
1
.86
1
.85
1
.62
1
.58
1
.59
1
.52(.49)**
Gender Age AcademicLevel
.06(.04)*
.02(.08)**
.00(.01)
.06*
.00
.22**
.75
.52(.52)**
.49(.47)**
EffectsTable 2: Unstandardized Path Coefficient and Total Effects
Path Unstanderdized
Coefficient
Standardized
Coefficient
Direct Effects
Gender to Teaching Presence .06* .04*
Age to Teaching Presence .02** .08**
Academic Level to Teaching Presence .00 .01
Teaching Presence to Social Presence .52** .52**
Teaching Presence to Cognitive Presence .49** .47**
Social Presence to Cognitive Presence .52** .49**
Total Effects
Gender to Social Presence .03 .00
Gender to Cognitive Presence .05 .00
Age to Social Presence .01 .00
Age to Cognitive Presence .02 .00
Academic Level to Social Presence .00 .01
Academic Level to Cognitive Presence .00 .01
Teaching Presence to Cognitive Presence .77** .72**
Note. *p<.05, **p<.001
Weaknesses of Model
• Research methods limited/problematic• Based primarily in discussion• Limited to one form of presence at a time• Other areas of courses not considered• Model categories biased toward what can
be discovered in discussions
Solutions
• Need to look at other areas of courses• Use multiple methods
– QCA, SNA, Cluster Analysis• Goal is to code entire courses for
– TP SP CP simultaneously• Revise and expand these categories• Develop more comprehensive model• Build theory w/ dialogue between methods
Current study• The current study goes beyond prediction of
student reports of “presence”.• Quantitative content analysis of two complete
online courses with…• Low v Higher Instructor Presence• Goal: find TP, SP, and CP
patterns/associations in discussions andother courses areas.
• Determine if previous research results can beverified, refined, or extended.
Re-examining Affective Indicatorsfor Measuring Social Presence inthe COI Framework
Three Categories of Social PresenceIndicators (Garrison et al. 2000)
Activities that build andsustain a sense of groupcommitment; use ofgreetings, names, “us” or“we,” social sharing
Group Cohesion
Responses that build andsustain relationships; tacitexpressions of support,encouragement, andacceptance
Open Communications(Interactive Responses)
Expression of emotions,feelings, mood
Affective Expression
Courses and CodingProcess
Coding Process• Two coders• Coded random module for practice• Met to negotiate• Identified problems• Established and documented guidelines• Recoded original module• Message = unit of analysis• Coded for presence/absence of SP indicators:
AF, OC, CH or NC
Reflection on the course itselfSP-CH5Course Reflection
Sharing information unrelated to the courseSP-CH-4Social Sharing
Communication that serves a purely social function; greetings orclosures
SP-CH-3Phatics, salutations & greetings
Addresses the group as we., us, our, groupSP-CH-2Addresses or refers to the group
using inclusive pronouns
Addressing or referring to the participants by nameSP-CH-IVocativesGroup Cohesion (CH)
Offering specific advice to classmatesSP-OC-8Personal Advice
Expresses disagreement with other or contents of others messagesSP-OC7Expressing Disagreement
Expressing agreement with others or contents of others messagesSP-OC-6Expressing agreement
Complimenting others or contents of others' messagesSP-OC-5Complimenting. expressingappreciation
Students ask questions of other students or the moderatorSP•OC4Asking questions
Direct references to contents of others' postsSP-OC-3Referring explicitly to others'messages
Using software features to quote others' entire message or cut andpassing selections of others' messages
SP-OC-2Quoting from others' messages
Using reply feature of software, rather than starting a new threadSP-OC-1Continuing a threadOpen Communication(OC)
Expressing personal values, beliefs and attitudesSP-AF5Expressing values
Unconventional expressions of emotion. includes repetitiouspunctuation, conspicuous capitalization, emoticons
SP-AF4Use of unconventional expressionsto express emotion
Presents details of life outside of class, or expresses vulnerabilitySP-AF3Self-disclosure
Teasing, cajoling, irony, understatements, sarcasmSP-AF2Use of humor
Conventional expressions of emotionSP-AFIExpressing emotionsAffective (AF)
DefinitionCodeIndicatorsSP Categories
Original Codes from Rourke et al. 1999. Swan et al. 2001 in blue
Instructor JYInstructor KS
2 to 3 postings a weekMinimum Number ofRequired Postings
1617Number of StudentsCompleting the Term
2019Number of StudentsAt Start of the Term
454490Total Number ofDiscussion Postings
One discussion from Modules 1-5Number ofDiscussions Coded
Approximately 2 weeksDuration ofDiscussions
Two Identical 15-week Courses
Initial and Negotiated Inter-rater Reliability using Cohen’s Kappa
Course A (KS) Course B (JY)Initial k Negotiated
kInitial k Negotiated
kModule 1 075 0.95 0.75 0.97Module 2 0.76 0.94 0.78 0.99Module 3 0.70 0.93 0.66 0.91Module 4 0.71 0.99 0.71 0.96Module 5 0.75 1 0.82 1
Initial and Negotiated Holsti’s Coefficient of Reliability
Course A (KS) Course B (JY)Initial Negotiated Initial Negotiated
Module 1 0.91 0.982857 0.90 0.981Module 2 0.91 0.980695 0.91 0.99Module 3 0.89 0.977778 0.87 0.97Module 4 0.89 0.998273 0.88 0.98Module 5 0.99 0.90411 0.93 1.00
.92.85With AF IndicatorsRemoved
.74.73With All SPCategories/Indicators
InstructorJY
InstructorKS
Pre-negotiatedaverage IRR* for 5Discussions
Number of Individual Student Indicators perModule
# ofstudents
Avgindicator per
studentJY AF OC CH indicators
Module 1 70 103 41 214 19 11.26Module 2 36 73 28 137 19 7.21Module 3 20 71 24 115 18 6.39Module 4 40 63 29 132 17 7.76Module 5 7 23 8 38 17 2.24Total 173 333 130 636
Number of Individual Student Indicators perModule
# ofstudents
Avgindicator
per studentKS AF OC CH indicators
Module 1 59 65 24 148 18 8.22Module 2 43 46 23 112 16 7.00Module 3 39 72 22 133 16 8.31Module 4 73 122 68 263 16 16.44Module 5 32 63 43 138 15 9.20Total 246 368 180 794
Findings•Overall increase in SP indicators in CourseKS•Overall decrease in SP indicators in CourseJY
Note: All numbers reflect negotiated occurrences ofindicator.
Previous Research• Swan (2003) found that Affective (AF) and
Open Communication (OC) increase overtime while Cohesive (C) indicatorsdecrease.
• Vaughan (2005) found that CH increasesover time while AF and OC decrease.
• Garrison & Arbaugh (2007) indicate that SPshould become more transparent over time.
Findings• All 3 indicators decreased in Course JY
• AF slowly declined, and OC and CHincreased in course KS.
• Preliminary findings on SP support all threeaforementioned patterns.
JY # of stud.Module 1 19Module 2 19Module 3 18Module 4 17Module 5 17
KS # of stud.Module 1 18Module 2 16Module 3 16Module 4 16Module 5 15
Note: All numbers reflect negotiated occurrences ofindicator.
Research Questions
How does:• instructor teaching presence correlate with
student social presence?• instructor social presence correlate with
student social presence?• instructor presence correlate with student
presence?
Instructor TP – Student SP
Instructor SP – Student SP
Total Posts: Teacher - Student
Conclusions
• Student presence tracks instructorpresence.
• Is this a better predictor of totalinteraction?
• Instructor SP is a better predictor ofstudent SP
• New direction for survey research?• Ask students to assess instructor SP as
well as or instead of TP?
Cluster Analysis
• Segment survey respondents into• Low, medium, high• TP, SP, CP• Look for patterns and associations
Cluster AnalysisClusters Item
Low Medium High
The instructor communicated course topics -.90 -.05 .75
The instructor communicated course goals -.89 -.08 .74
The instructor provided clear instructions -.89 -.04 .75
The instructor communicated due dates -.84 -.06 .67
The instructor helped students learn -1.05 -.07 .85
The instructor helped students clarify their thinking -1.05 -.05 .85
The instruct or kept students engaged & participating -1.04 -.06 .85
The instructor kept students on task -1.04 -.07 .87
The instructor encouraged students to explore new ideas -1.08 -.08 .83
The instructor established a sense of course community -1.01 -.07 .86
The instructor helped focus discussion on issues that aided student
learning
-.97 -.08 .80
The instructor gave feedback that helped students -1.03 -.07 .82
My instructor provided explanations or demonstrations to help me
better understand the content of the course.
-1.06 -.06 .84
My instructor provided feedback to the class during the
discussions or other activities to help us learn.
-1.04 -.06 .81
My instructor asked for feedback on how this course could be
improved.
-.87 -.05 .76
Total Number of Student s 520 1422 1682
TP and SP Correlate with CP
• Evidence of an equivalence model• Increase in SP correlates with higher CP• But better TP compensates for lower SP
CP X TP X SP and Type of Instruction
Significant Interaction
• Student who report low or medium TP inonline courses report lower CP than likestudents in hybrid courses.
• Students who report high TP in full onlinecourses report higher CP than studentwho report high TP in hybrid courses
Interaction
Social Network Analysis
• Method to determine the location ofparticipants within their network ofinteractions
• Measures of centrality, prestige, densityetc
• Can reveal who is contributing todiscourse and learning in significant ways
SNA: JY Module 1
SNA: Measures
Density
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
M01 M02 M03 M04 M05
Modules
De
ns
ity
(m
atr
ix a
ve
rag
e)
KS
JY
Other equivalencies
• Is SNA density roughly equivalent tostudent social presence measure?
• If so, an incredible time savingdevice…
• Can be computed automatically versus• Weeks of coding in QCA
Density
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
M01 M02 M03 M04 M05
Modules
De
ns
ity
(m
atr
ix a
ve
rag
e)
KS
JY
Social Network Visualizations
Next steps
• Coding TP and SP outside of discussions• Coding CP• Further dialogue between survey research
and QCA