<%BANNER%>

Effects of Detailed Customization of Student Avatars on Teacher Expectations of Students

Permanent Link: http://ufdc.ufl.edu/UFE0023574/00001

Material Information

Title: Effects of Detailed Customization of Student Avatars on Teacher Expectations of Students
Physical Description: 1 online resource (178 p.)
Language: english
Creator: Beck, Dennis
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: ethnicity, gender, learning, life, pedagogy, second, virtual
Teaching and Learning -- Dissertations, Academic -- UF
Genre: Curriculum and Instruction (ISC) thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The gender and ethnicity of students have been shown to specifically affect teacher expectations of students in real-life classrooms. As part of a Multi User Virtual environment (MUVE), people socially interact via avatars which have the capability to be customized to minute details of ethnicity and gender. Currently, teachers in MUVEs instruct students with little to no knowledge of their own potential biases and prejudices toward students of different genders and ethnicities. Our purpose was to examine the impact of the gender and ethnicity of student avatars on teachers? expectations and perceptions. Our study surveyed 453 teachers in Second Life, a popular MUVE. Teachers were asked to review a transcript, image and video of student avatar in Second Life and then respond to a questionnaire about the student?s intellectual and relational abilities. Teachers also responded to a demographic questionnaire. Results indicate that choice of avatar gender and ethnicity does influence teachers? expectations and perceptions in a MUVE. Results may aid teachers in MUVEs to discover potential biases and prejudices toward some student avatars, as well as level the playing field for student avatars of all ranges of gender and ethnicity.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Dennis Beck.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Ferdig, Richard E.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0023574:00001

Permanent Link: http://ufdc.ufl.edu/UFE0023574/00001

Material Information

Title: Effects of Detailed Customization of Student Avatars on Teacher Expectations of Students
Physical Description: 1 online resource (178 p.)
Language: english
Creator: Beck, Dennis
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: ethnicity, gender, learning, life, pedagogy, second, virtual
Teaching and Learning -- Dissertations, Academic -- UF
Genre: Curriculum and Instruction (ISC) thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The gender and ethnicity of students have been shown to specifically affect teacher expectations of students in real-life classrooms. As part of a Multi User Virtual environment (MUVE), people socially interact via avatars which have the capability to be customized to minute details of ethnicity and gender. Currently, teachers in MUVEs instruct students with little to no knowledge of their own potential biases and prejudices toward students of different genders and ethnicities. Our purpose was to examine the impact of the gender and ethnicity of student avatars on teachers? expectations and perceptions. Our study surveyed 453 teachers in Second Life, a popular MUVE. Teachers were asked to review a transcript, image and video of student avatar in Second Life and then respond to a questionnaire about the student?s intellectual and relational abilities. Teachers also responded to a demographic questionnaire. Results indicate that choice of avatar gender and ethnicity does influence teachers? expectations and perceptions in a MUVE. Results may aid teachers in MUVEs to discover potential biases and prejudices toward some student avatars, as well as level the playing field for student avatars of all ranges of gender and ethnicity.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Dennis Beck.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Ferdig, Richard E.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0023574:00001


This item has the following downloads:


Full Text

PAGE 1

1 EFFECTS OF DETAILED CUSTOMIZATIO N OF STUDENT AVATARS ON TEACHER EXPECTATIONS OF STUDENTS By DENNIS E. BECK A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Dennis E. Beck

PAGE 3

3 To my sweet Jennifer: You encourage and inspire me. Soli Deo Gloria

PAGE 4

4 ACKNOWLEDGMENTS So m any people need to be thanked for their contribution to this di ssertation. First and foremost, I thank my Lord Jesus Christ, who ha s rescued me from my sin and given me eternal life. His unconditional love for me has captured my heart and inspir es my service. I also thank my wife Jennifer, without whom I would have given up long ago and neve r finished. Her selfsacrifice on behalf of my academic success amazes me and just makes me want to love her more. I also thank my children, Hannah and Caleb. Their joy in the little things of life has helped me to see what is really important. I am very thankful for an exceptional docto ral committee. Dr. Rick Ferdig, Dr. Kara Dawson, Dr. Sevan Terzian, Dr. Jame s Algina and Dr. Paul Fishwick have been a great help and inspiration to finish with excel lence. I would particularly like to thank Dr. Ferdig for his generous time and commitment. Throughout my doctoral work he encouraged me to develop independent thinking and research skills and became a good friend as well. He continually stimulated my analytical thinking and greatly assisted me with sc ientific and narrative writing. The many hours spent by Dr. Algina on the design and analysis of my data are much appreciated, as well as the timely assistance of Dr Dawson, Dr. Terzian, and Dr. Fishwick. I also thank my church family at Abundant Grace Community Church. Many times they have listened to my dissertation ideas and re sponded with patience and encouragement while providing a loving environment for our family to be a part of. I also thank my employer, Campus Crusade for Christ International. Th eir passion for Christ a nd professional excellence has helped mold me into the professional that I am today. Finally, I than k my parents, in-laws, and the many relatives who have been an encouragement to press on with my dissertation and finish.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ...........................................................................................................................8ABSTRACT ...................................................................................................................... .............10 CHAP TER 1 INTRODUCTION .................................................................................................................. 11Purpose Statement ............................................................................................................. .....11Face-to-Face Expectations and Perceptions in Education ...................................................... 12Multi User Virtual Environments in Education ...................................................................... 13Problem Statement ............................................................................................................. .....14Research Questions ......................................................................................................... 15Significance .................................................................................................................. ...17Definition of Terms .........................................................................................................192 LITERATURE REVIEW .......................................................................................................21Introduction .................................................................................................................. ...........21An Overview of Multi User Virtual Environments ................................................................ 22Multi User Virtual Environments in Education ...................................................................... 24Second Life in Education ........................................................................................................26Face to Face Interaction ...................................................................................................... ....28Computer Mediated Comm unication Interaction ...................................................................30Avatar-based Interaction in a Mu lti User Virtual Environment ............................................. 33Identity experimentation with Avatars ............................................................................ 34Social Interaction with Avatars ....................................................................................... 35Condition Awareness and Gestures ................................................................................. 37Issues of Gender and Ethn icity in the Classroom ................................................................... 39History of Gender and Ethnicity Issues ...........................................................................40Research on Gender and Ethnicity Issues .......................................................................42Ethnic Minorities Biases ..........................................................................................44Gender Biases ...........................................................................................................46Ethnicity and Gender Issues Conclusion ..................................................................47Conclusion .................................................................................................................... ..........473 METHODOLOGY ................................................................................................................. 49Introduction .................................................................................................................. ...........49Research Design .....................................................................................................................49Population and Sample ......................................................................................................... ..50

PAGE 6

6 Privacy and Confidentiality ................................................................................................... .52Data Integrity and Security .....................................................................................................53Sampling Procedure ................................................................................................................53Instrumentation ............................................................................................................... ........54Data Collection .......................................................................................................................55Rationale for Study Format ....................................................................................................56Avatar Creation .......................................................................................................................58Study Progression ...................................................................................................................58Model Equation ......................................................................................................................60Data Analysis Techniques ......................................................................................................60Limitations ................................................................................................................... ...........644 RESULTS ....................................................................................................................... ........69Introduction .................................................................................................................. ...........69Participant Demographics ...................................................................................................... .69Descriptive Statistics ........................................................................................................ ......70Discussion of Outliers for IQ variable .................................................................................... 70CrossTabs Table Analysis ......................................................................................................71Analyses of Research Questions ............................................................................................. 73Ethnicity Related Questions ............................................................................................73Summary: Ethnicity Related Questions .........................................................................74Gender Related Questions ...............................................................................................75Summary: Gender Related Questions ...........................................................................76Interactions Between Gender and Ethnicity .................................................................... 76Summary: Interactions Be tween Gender and Ethnicity ................................................81Conclusions .............................................................................................................................825 IMPLICATIONS ..................................................................................................................102Research Questions and Discussion ..................................................................................... 103Ethnicity Related Questions ..........................................................................................104Gender Related Questions .............................................................................................108Interactions Between Gender and Ethnicity .................................................................. 114Summary: Direct Implications ............................................................................................. 118Indirect Implications: Detailed Customization .................................................................... 123Ambient Implications .......................................................................................................... .125Cumulative folder method ............................................................................................. 125Demographics Implications ...........................................................................................127Implications of Virtual Representations of Self ............................................................ 129Other Future Study Ideas ...................................................................................................... 132Conclusion .................................................................................................................... ........134

PAGE 7

7 APPPENDIX A IRB ........................................................................................................................... .............137B PRE-SURVEY NOTI FICATION LETTER ......................................................................... 142C INFORMED CONSENT FORM .......................................................................................... 143D AVATAR PHOTOGRAPHS ................................................................................................145E LETTER TO TEACHERS ...................................................................................................147F STUDENT HIGH SCHOOL TRANSCRIPT .......................................................................148G TRANSCRIPT COMMENTS AND DE MOGRAPHIC S URVEY REGARDING CONTROL VARIABLES .................................................................................................... 149H DEBRIEF EMAIL ................................................................................................................ 152REFERENCE LIST .....................................................................................................................154BIOGRAPHICAL SKETCH .......................................................................................................178

PAGE 8

8 LIST OF TABLES Table page 1-1Various selected educationa l uses of Second Life ............................................................. 20 3-1 Overview of instruments/materials us ed and what they are used to measure ....................66 3-2List of variables ............................................................................................................. .....67 4-2Education Completed .........................................................................................................83 4-3Age. ....................................................................................................................... .84 4-4Gender ...................................................................................................................... ..84 4-5Number of children ............................................................................................................84 4-6Marital status ................................................................................................................ ......85 4-7Ethnicity ..................................................................................................................... ........85 4-8Religious affiliation ......................................................................................................... ..85 4-9Continent of origin ........................................................................................................... ..86 4-10 Annual family income .................................................................................................... ....86 4-11 Daily Internet use ...............................................................................................................86 4-12 Time spent in Second Life each week ............................................................................... 87 4-13 Residency length in Second Life .......................................................................................87 4-14 Primary motivation for coming to SL ................................................................................ 88 4-15 Dependent Variables Descriptive Statistics: without IQ Deletions ..................................88 4-16 Dependent Variables Descriptive Statistics: with IQ Deletions .......................................89 4-17 Independent variables Descriptive Statistics: without IQ Deletions .................................89 4-18 Independent variables Descriptiv e Statistics: with IQ Deletions ......................................90 4-19 Gender of the Avatar as Designated by the Researcher by Perceived Gender of the Avatar Crosstabs ................................................................................................................91

PAGE 9

9 4-20 Gender of the Avatar as Designated by th e Researcher by Perceived Ethnicity of the Avatar Crosstabs ................................................................................................................92 4-21 Ethnicity of the Avatar as Designated by the Researcher by Perceived Gender of the Avatar Crosstabs ................................................................................................................93 4-22 Ethnicity of the Avatar as Designated by the Researcher Perceived Ethnicity of the Avatar Crosstabs ................................................................................................................94 4-24 Continuous Data Set Hypotheses Results .......................................................................... 96 4-25 Participant Ethnicity with Students Relationships with Instructors Estimated Marginal Means .................................................................................................................97 4-26 Perceived Gender with Attitude toward School Estimated Marginal Means .................... 97 4-27 Perceived Gender Participant Gender Interaction for Attitude toward School Estimated Marginal Means ................................................................................................97 4-28 Perceived Gender Participant Ethnic ity Interaction for Student IQ Estimated Marginal Means .................................................................................................................98 4-29 Perceived Gender Participant Ethnic ity Interaction for Attitude toward School Estimated Marginal Means ................................................................................................98 4-30 Participant Ethnicity Perceived Gender Perceive d Ethnicity Interaction for Student IQ Estimated Marginal Means .............................................................................. 99 4-31 Participant Ethnicity Participant Gender Perceive d Ethnicity Interaction for Student IQ Estimated Marginal Means ............................................................................ 100 4-32 Participant Ethnicity Participant Gender Perceived Gender Perceived Ethnicity Interaction for Student IQ Estimated Marginal Means .................................................... 101 5-1 Similarities and Differences in demogra phics of Real Life vs. Second Life Teachers ... 136

PAGE 10

10 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECTS OF DETAILED CUSTOMIZATIO N OF STUDENT AVATARS ON TEACHER EXPECTATIONS OF STUDENTS By Dennis E. Beck December 2008 Chair: Richard E. Ferdig, Ph.D. Major: Curriculum and Instruction The gender and ethnicity of students have b een shown to specifically affect teacher expectations of students in real-life classrooms. As part of a Multi User Virtual environment (MUVE), people socially interact via avatars whic h have the capability to be customized to minute details of ethnicity and gender. Currently, teachers in MUVEs instruct students with little to no knowledge of their own potential biases and prejudices toward student s of different genders and ethnicities. Our purpose was to examine the impact of the gender and ethnicity of student avatars on teachers expectations and percepti ons. Our study surveyed 453 teachers in Second Life, a popular MUVE. Teachers were asked to re view a transcript, image and video of student avatar in Second Life and then respond to a questionnaire about the students intellectual and relational abili ties. Teachers also res ponded to a demographic questi onnaire. Results indicate that choice of avatar gender and ethnicity does influence teachers expect ations and perceptions in a MUVE. Results may aid teachers in MUVEs to discover potential biases and prejudices toward some student avatars, as well as level the play ing field for student avatars of all ranges of gender and ethnicity.

PAGE 11

11 CHAPTER 1 INTRODUCTION Purpose Statement Social interaction is im porta nt to learning (Vygotsky, 1978) and can occur in technology mediated forms (e.g., 5th dimension project Cole, 1995; Nicolopolou & Cole, 1993; CSILE Scardamalia & Bereiter, 1996; Scardamalia, et al., 1994; LaFerriere, 2002). One form of technology mediated interaction th at is increasingly being used in education is the Multi-User Virtual Environment (MUVE; New Media C onsortium, 2007; Castronova, 2001). These environments allow users to experience a graphically engaging world while interacting with others as an avatar (Lastowka & Hunter, 2004). An avatar is a persons graphical representation of himself or herself (Meadows, 2008; Wood, et al 2005; Dieterle & Clarke, in press). Users can choose the physical appearance of their avatar by customizing mi nute details of shape, skin, hair, eyes, ethnicity and gender (Damer, 1998; Rehak, 2003; Lastowka & Hunter, 2004). Unfortunately, we know little about the effect of this avatar ch oice on student success in these environments (Lastowka & Hunter, 2004). We also have an inadequate understanding about the effect of avatar choice on teacher expectations and perceptions of student success. The gender and ethnicity of students has been shown to specifically effect teacher perceptions and expectations of students in face-to-face classrooms (Clifford & Walster, 1973; Guttm ann & Bar-Tal, 1982; Walther & Tidwell, 1996; Ferguson, 2003; Frawley, 2005; Van Duzer, 2006; Au warter & Aruguete, 2008). If we were to discover the influence of avatar choice (i.e. ge nder and ethnicity) on teacher expectations and perceptions of student success, then we would be able to help teachers and students to avoid the problems of unintentional disadvantage and unfair advantage. These problems threaten the possibility of bias against student avatars of pa rticular genders and ethnicities. Furthermore,

PAGE 12

12 teacher bias in their perceptions of students often results in lower academic performance by students (Auwarter & Aruguete, 2008; Rist, 1970). As a result of these problems, we need to study the influence of avatar choice (i.e. gender and ethnicity ) on teacher perceptions and expectations of student success. Face-to-Face Expectations and Perceptions in Education Interactions between teachers and s tudents are important in the face-to-face classroom because these interactions form most teachers expectations and perceptions of the student (Levy et al, 2003; Irvine, 1986). As a result, learning to interact with teachers is an important social interaction skill for students to learn in the classr oom. Teacher-student interaction is a two-way process: Each participant influences the othe r's behaviors (Levy et al, 2003; Irvine, 1986). Brophy & Evertson (1981) and Muijs & Reynolds (2005) describe this process as a two-sided relationship between teachers a nd students who are active, initi ating, and relevant in their interactions. Teacher perceptions and expectations regarding student ethnicity are critical to students academic accomplishments (Dee, 2004; Fergus on, 2003; Roscigno, 1998; Alexander, Entwisle, & Thompson, 1987; Irvine, 1986; Rist, 1970). Socio-economic trends translate into hardship for ethnic minority students, consideri ng that the majority of lower socioeconomic students are from an ethnic minority (Ferguson, 2003; Rosciglio, 1998) In general, instructors expect students from a poor background to do worse than middleand upper-class students, not considering aptitude (Dee, 2004; Ferguson, 2003; Alexander, Entwisle, & Thompson, 1987; Beady & Hansell, 1981; Gay, 1975; Rist, 1970; Simpson & Erickson, 1983). Teacher gender biases also influence their expectations and perceptions of student success (B eaman, et al., 2006). Oyserman et al. (2001) contend that females are socialized to be more involved with

PAGE 13

13 relationships and the approval of others than ar e males. Due to this socialization, females academic self-worth may be hurt more when they experience negative interactions from teachers. These characteristics may well shape teachers expectations and perceptions of their female students. Perceptions and expectations ar e indeed a powerful force in education. Multi User Virtual Environments in Education Teachers are increasing ly expressing their pe rceptions and expectations of students through the use of technology (New Media Consortium, 2007). One type of technological environment used for learning is known as the Multi-U ser Virtual Environment (MUVE) (New Media Consortium, 2007; Castronova, 2001). MUVEs are technology-based simulations that engage people in a different, yet cohesive reality. Users navigate the worl d, view and control items using technology varying from a keyboard and mouse to a head-mounted di splay with gloves (Laferriere, et al., 2002). This world is usually seen as a two or three-dimensional graphical depiction of other people (avatars) who can be represented as animals, humans, or whatever they want. Not considering subject matter and intend ed audience, all MUVEs allow numerous users to concurrently (a) log on to a virtual world, (b ) manipulate virtual items, (c) customize their own person through an avatar, (d) in teract with other users and artificial intelligences, and (e) experience modeling and mentoring in simulated environments that help them to think about a difficulty in their physical context (Dede et al., 2004). Experts predict that 80 percent of active Internet users will regularly participate in an MUVE by the end of 2011 (Gartner, 2007). Today s array of virtual world environments has evolved through the synthesis of social networking, simulation and online role playing games. The MUVE Second Life ( http://www.secondlife.com ) currently boasts m ore than 6.6 million residents worldwide and over 7 million U.S. dollars are spent in a given month. The top 80

PAGE 14

14 percent of residents hail from more developed nations, but a signi ficant portion also log on from places like Africa, Southeast Asia, and Latin America. The typical resident is male, between 2534 years old, and among the top 5% of the worlds wealthiest individuals. A significant number of universit ies, colleges, schools, orga nizations, and businesses are exploring the educationa l potential of Second Life. Accord ing to the Chronicle of Higher Education (Foster, 2007), as of September 21, 2 007, more than 150 colleges in the United States and in 13 other countries have a presence in SL. Additionally, according to the Second Life Wiki (SimTeach, 2007), 17 educational organizati ons; 4 libraries; and 4 museums are currently active in SL. Second Life education relate d websites number more than two hundred. The purposes of these groups involvement in SL ar e as varied as their creators. Foreman (2003) envisages virtual worlds as the learning environments of the futur e (p. 14), but it is safe to say that MUVEs have significan t educational potential. Students have the ability to choose the mi nute details of their avatars in MUVE environments (Damer, 1998; Re hak, 2003; Lastowka & Hunter, 2004). For example, in Second Life, an avatar can be changed on a 0 to 100 scale for shape (Body, Head, Eyes, Ears, Nose, Mouth, Chin, Torso, Legs, Male/Female, etc.), skin (skin Color, Face Detail, Makeup, Body Detail, etc.), hair (Color, Style, Eyebrows, Facial), and eyes (Color, texture, etc.). Specifically, gender and ethnicity can be clearl y expressed in a variety of wa ys. The influence of student avatar choice on teacher expectations and perceptions is currently unknown. Problem Statement Unfortunately, little research has been done regarding the influence of avatar choice (i.e. gender and ethnicity) on teacher pe rceptions and expectations of student success. Student gender and ethnicity does influence teacher expectations and perceptions of student success in face-to-

PAGE 15

15 face classrooms (Clifford & Walster, 1973; Guttmann & Bar-Tal, 1982; Ferguson, 2003; Frawley, 2005; Van Duzer, 2006; Auwarter & Aruguete, 2008; Beaman, et al., 2006; Roscigno, 1998; Alexander, Entwisle, & Thompson, 1987; Irvine, 1986; Rist, 1970; Dee, 2005). This research needs to also be done in MUVEs because students in these environments can become anyone by customizing their avat ar. Student avatar customiz ation and teacher gender and ethnicity biases may combine to create the problems of unintentional disadvantage and unfair advantage. For example, lets say that a male teacher in SL was biased toward their own gender in their rating of IQ for student avatars. Stude nts who didnt know this information could end up choosing a female avatar when taught by a male teacher, which could result in unintentional disadvantage to this student. However, a st udent who did know this information could end up choosing a male avatar when taught by a male teacher, resulting in unfair advantage. Both possibilities are troubling because the academic playing field in MUVEs may not be level it could be biased toward student avatars of part icular genders and ethnicities. Furthermore, teacher bias in their perceptions of students often results in lower academic performance by students (Auwarter & Aruguete, 2008; Rist, 1970). Thus, this is more than just a concern about a few students who may or may not have knowledge of the results of th is study. The results of this study concern all teachers and students because all students can be affected in their academic performance by teacher expectations and perceptions. As a result, we need to study the influence of avatar choice on teacher perceptions and expectations of student success. Research Questions The overarching question in th is stu dy was whether avatar choice affected teacher expectations and perceptions of student academic success. Many references have been given that have convincingly shown that teacher biases ba sed on the gender and ethnicity of students do

PAGE 16

16 exist in face-to-face cl assrooms. However, this study was done in a MUVE. The sample selected mirrors the overall population of Sec ond Life, which is mostly Caucasian and from western nations. Thus, we appear to have a sample that accurately represents the ove rall population of SL. The question then becomes: How does the world of the physical compare to the virtual? Does SL mirror the real world in terms of biases, prej udices, and discrimination? Each of the next 15 research questions is examined with that intent. When you are in SL, does that capture the biases that are inherent in the real world? For ease of perspective, questions are broken down into those that are ethnicity related, gender relate d, and gender/ethnicity interactions. Ethnicity related questions o Does the ethnicity of the avatar affect teacher perceptions and expectations? o Does the ethnicity of the participant aff ect teacher perceptions and expectations? o Is there an interaction between the ethnicity of the avatar and the ethnicity of the participant? Gender related questions o Does the gender of the avatar affect teacher perceptions and expectations? o Does the gender of the participant affect teacher perceptions and expectations? o Is there an interaction between the gende r of the avatar and the gender of the participant? Interactions between gender and ethnicity o Is there an interaction between the ethnic ity and gender of the avatar and teacher perceptions and expectations? o Is there an interaction between the gender of the avatar and the ethnicity of the participant?

PAGE 17

17 o Is there an interaction between the ethnic ity of the avatar and the gender of the participant? o Is there an interaction between the ethnic ity of the participant and the gender of the participant? o Is there an interaction between the ethnicity of the avatar, the gender of the avatar, and the ethnicity of the participant? o Is there an interaction between the ethnicity of the avatar, the gender of the avatar, and the gender of the participant? o Is there an interaction between the ethnic ity of the avatar, the ethnicity of the participant, and the gend er of the participant? o Is there an interaction between the gende r of the avatar, the ethnicity of the participant, and the gend er of the participant? o Is there an interaction between the ethnicity of the avatar, the gender of the avatar, the ethnicity of the participant, and the gender of the participant? Significance MUVEs have significant and la rgely untapped educational pote ntial (Table 1-1). Forem an (2003) envisages virtual worlds as the learning environments of the future (p. 14). MUVEs hold considerable potential for the development of complex social practices such as leadership, collaboration, and relationships. These worlds are complex rational groups distinguished by their social practices (Stei nkuehler, 2004). Virtual worlds are compelling because social relations, collaboration, and info rmation sharing are essential ingredients, and encourage collaboration both within and beyond game parameters (Delwiche, 2006). According to Yee (2006), more than half of those involved in virtua l worlds have gained proficiency in mediation and leadership, such as solving conflict in groups. The educationa l potential of MUVEs becomes clear with these kinds of potential learning outcomes.

PAGE 18

18 As a well-known and widely distribute d MUVE, Second Life possesses significant potential for innovations in le arning and research. For exam ple, Dave Taylor, knowledge transfer leader at the Nationa l Physical Laboratory, says that the use of SL opens up, new opportunities for collaboration across disciplines and geographies that would not otherwise occur (Edwards, 2006, p. 32). These online environm ents are also being e xplored as potential sites for research in the social sciences and th erapy due to the apparent transference of social norms such as gender, interpersonal space and eye gaze (Yee, et al., 2007; Blascovich et al., 2002) as well as the affordances of recorded interactions and customizable environments. Unfortunately, instructional practices in MUVE s remain adaptations of those used in webbased and face-to-face classrooms (Delwiche, 2006; Keesey, 2007). With the similarity in appearance and movement in MUVEs to real life appearance and movement, there is reason to believe that the impact of student avatar physical characteristics could be as important as their impact in traditional schooling. Unfortunately, research only addresses face-to-face and two dimensional online student populations. Theref ore, the study of student avatar physical characteristics and their effect on expected curre nt and future performance of the student by the teacher could help teachers in MUVEs to discover potential biases and prejudices toward some students, as well as level the playing field for student avat ars of all ranges of detailed customization. Once we understand the effect of detailed customization of student avatars on expected current and future performance of the student by the teacher, we will begin to understand the implications of how these characteris tics could impact social interaction in work, learning, and other environments. Additionally, w ith the benefits to using MUVEs in education mentioned above, any hindrance to a teachers ex pectations and perceptions about a students

PAGE 19

19 ability to socially interact, participate, and conc eptually and spatially grasp content needs to be addressed. Gender and ethnicity bi ases are two such hindrances th at this study will address. Definition of Terms This study determ ines the effects of detailed customization of student avatars on grades and treatment given by teacher avatars. The follow ing are definitions of terminology used in this study: 1. An Avatar is an Internet user's repr esentation of himself or herself, whether in the form of a three-dimensional model used in computer games, a two-dimensional icon (picture) used on Internet forums and other communities, or a text construct found on early systems such as MUDs (Wood, et al, 2005; Meadows, 2008, Wikipedia, 2008, Retrieved October 22, 2007 from http://en.wikipedia.org/wiki/A vatar_%28virtual_reality%29 paragraph 1). 2. A Multi Use r Virtual World Environment is a computer-based simulated environment intended for its users to inhabit and interact via avatars. This habitation usually is represented in the form of two or three-dimensional gr aphical representations of humanoids (or other graphical or text-based avatars). MUVEs enable multiple simultaneous participants to (a) access virtual contexts, (b) interact with digital artifacts, (c) represent themselves through avatars, (d ) communicate with other participants, and (e) take part in experiences incorporating modeling and mentoring about problems similar to those in real world contex ts. From: Dieterle, E., & Clar ke, J. (in press). Multi-user virtual environments for teaching and learning. In M. Pagani (Ed.), Encyclopedia of multimedia technology and networking (2nd ed). Hershey, PA: Idea Group, Inc. 3. Computer Mediated Communicati on is any form of data exchange across two or more networked computers. More frequently, the term is narrowed to include only those communications that occur via computer-media ted formats (i.e., instant messages, e-mails, chat rooms) between two or more individua ls (Thurlow, et al, 2004; Wikipedia, 2008; Retrieved January 25, 2008 from http://en.wikipedia.org/wiki/Computerm ediated_communication ). 4. Multi-User Dungeon (MUD) a virtual environment that supports the simultaneous participation of multiple users in a text-based game. From Dieterle, E., & Clarke, J. (in press). Multi-user virtual environments fo r teaching and learning. In M. Pagani (Ed.), Encyclopedia of multimedia technology and networking (2nd ed). Hershey, PA: Idea Group, Inc.

PAGE 20

20 Table 1-1. Various selected educ ational uses of Second Life Institution Purpose Location Idaho State University Bioterrorism preparedness Program http://irhbt.typepad.com/play2train Dartmouth College Simulation for distribution of medical supplie s in crisis. http://iml.dartmouth.edu/index.html National Oceanic and At mospheric Administration Interactive educational simulations about the ocean and weather http://www.esrl.noaa.gov/ Global Kids Island Place for teen residen ts to learn about social and world issues. http://holymeatballs.org/ Kids Connect Island Youth collaborate via perform ing, storytelling and collaboration http://zoomlab.org/kc/ The Social Sim ulation Research Lab Library with papers, websites, homepages, and references of interest to social scientists. http://tinyurl.com/y3wlat BrainTalk Communities Inc. A place for autis tic and cerebral palsy patients to interact socially http://braintalk.blogs.com/brigadoon http://braintalk.blogs.com/live2give/ Am erican Cancer Society Walkathon raised more $ in a short time than what they would make in real life over many months. http://www.cancer.org /docroot/GI/conte nt/GI_1_8_Second_Life_Relay.asp Seattle Univ ersity Propert y law course applies issues of law to virtual environments. http://fizzysec ondlife.blogspot.com University o f Houston Design Economics course students try their entrepreneurial skills http://www.arch.uh.edu New Media Consortium Virtual laboratory constructed to provide dozens of settings for social interaction experim ents. http://sl.nmc.org/wiki/Main_Page

PAGE 21

21 CHAPTER 2 LITERATURE REVIEW Introduction Social interaction is an impor tant component of student le arning in traditional schooling environm ents (Vygotsky, 1978). Several researchers have suggested that interaction is also of importance to Multi-User Virtua l Environments (MUVEs) (Taylor, 1999; Qvortrup, 2001, 2002; Siggard Jensen, 2007). As part of a MUVE, peop le socially interact via avatars (Lastowka & Hunter, 2004). These avatars have the capability to be customized to minute details (Damer, 1998; Rehak, 2003; Lastowka & Hunter, 2004). Unfortunately, we know little about the influence of avatar choice on teacher expecta tions and perceptions of student success. In particular, the gender and ethnicity of students have been shown to specifically effect teacher expectations of students in f ace-to-face classrooms (Clifford & Walster, 1973; Guttmann & BarTal, 1982; Ferguson, 2003; Frawley, 2005; Van Duzer, 2006; Auwarter & Aruguete, 2008). If we were to discover the effect of the choice of avatar (i.e. ge nder and ethnicity) on teacher expectations and perceptions of student success, we would be able to help teachers in MUVEs to discover potential biases and prej udices toward some student av atars, as well as handle the problems of unfair disadvantage and unfair advantage mentioned earlier. This study explores the effect of the gender and ethnicity of student av atars on teacher expectations and perceptions of student success. The goal of this chapter is to explore th e current state of research in MUVEs by establishing it within the context of existing rese arch that explores soci al interaction and its impact on teachers expected student performance. This chapter delves into the literature surrounding Multi-User Virtual Environments, face-to-face interaction, computer-mediated interaction, avatar-bas ed interaction, and the hist ory of gender and ethnicity bias in schools. This

PAGE 22

22 literature review a) distinguishes what has been done in the fields in question from what needs to be done, b) places the topic or problem in the broa der scholarly literature, and c) synthesizes and gains a new perspective on the literature (Boote & Beile, 2005). This review of literature will function to establish a context for multi-user virtual environments and provide a basis for discussing the effect of detailed customization of student avatars on expected current and future performance of the student by the teacher. An Overview of Multi User Virtual Environments In m any ways, the growth of MUVEs parall els the growth of th e Internet, as new technologies shift from small scale products that are the first of their kind and created at great expense toward large scale, typical products av ailable to all (Bartle, 1999). These technologies are both used and co-constructed by the general public. With the Web, this occurred through the use of hypertext, wikis, blogs, mashups and other similar technol ogies. In the case of MUVEs users can create 3D artifacts, buildings, and soci al spaces where people interact. The social nature of these MUVEs is a critical component of understanding their fundamental nature and utility (Qvortrup, 2001, 2002). While the magnitude of MUVEs is very different than what has come before, the idea of a persistent space that offers users the ability to soci ally interact with others and take part in the co-construction of the world with which they inte ract is not. Trubshaw an d Bartle authored the first multi-user dungeon (MUD) in 1978 (Bartle, 1999) This was a game that allowed multiple players to engage in role-playing adventures. M UDs were multi-player and persistent and users often found the social interacti on experienced within MUDs the most compelling aspect. In 1989, the MOO (MUD, Object Oriented) was devel oped to allow users to create and modify

PAGE 23

23 places within a MUD space, enabling the construction of various types for social, educational, and other uses (Bartle, 1999; Castronova, 2006). Even though the development of three dimensi onal rendering technologies, such as Virtual Reality Modeling Language (VRML) paralleled MOO development, it lacked the capability to create multi-user environments (Durbridge, 2004). Companies such as ActiveWorlds ( http://www.activeworlds.com ) created persisten t, multi-user spaces that provided many of the same characteristics that MOOs had, but in thr ee dimensions. At the same time, graphic role playing games, the successor to the MUD, we re transitioning to the Web. These MMORPGs (Massively multi-player online role playing game s) moved to full online-based networks in 1997 with the launch of Ultima Online (Bartle, 1999). Membership in massively multi-player online role-playing games (MMORPGs) expanded quickly, and by mid-2006 exceeded 16 million subscriptions to various MMOG environments (Castranova, 2006; http://www.mmogchart.com ). Multi-User Virtual Environm en ts are deeply immersive and highly scalable threedimensional systems (Dieterle & Clark, in press) Several interesting examples of MUVEs are currently available: 1. Second Life ( http://www.secondlife.com ) possesses a sm all amount of the total number of participants in MMOGs worldwide, but it is one of the largest and most rapidly developing of the new wave of persistent MUVEs. 2. Linden Labs created Teen Second Life ( http://teen.secondlife.com ) that r equires adults aged 18 and over to complete a b ackground check before they enter. There ( http://www.there.com ) offers sim ilar defining characterist ics, but restrains user activities in order to limit immoral activity. 3. Multiverse ( http://www.multiverse.net) is creating an open plat form for the development of both MUVEs and MMOGs through a commo n set of standards and clients.

PAGE 24

24 4. A consortium of academic institutions is developing the Croquet project ( http://www.opencroquet.org ) as an open-source environm ent for building collaborative virtual learning environments. 5. The Acceleration Studies Foundation has es tablished a Metaverse Roadmap Project ( http://www.metaverseroadmap.org ) to forecas t and shape the development of the new 3D web. Multi User Virtual Environments in Education MUVEs have significant and largely untapped e ducational potential. Foreman predicts that shared graphical worlds ar e the learning environments of the future (2003, p. 14). In many MUVEs, people enter the vi rtual world through th e use of an avatar, a character that embodies their presence and intent (Taylor, 1999; Qvortrup, 2001, 2002; Siggard Jensen, 2007). Many popular MUVEs allow for multiple users to be in the same virtual space and interrelate with each other at the same time (New Media Consortium, 2007). Even in their nascent state, virtual worlds allow for the development of r eal life cultures through the use of individuated dialects, political configurations, multifaceted social customs, social networks, social capital and common history (Steinkuehler, 2004; Jakobsson and Taylor, 2003). MUVEs combine social networking, seamless sharing of rich media, and a feeling of presen ce in a generalized, persistent non-contextual environment that is applicable to almost all disciplines (Castronova, 2001). MUVEs offer the prospect for users to inte rrelate in a way that delivers a sense of presence that is wanting in other technology (Castronova, 2001; New Media Consortium, 2007). This aspect lends itself to role-playing and s ituation construction, freeing up learners to assume the responsibilities of a physicist, artist, physician, or architect wit hout the real-wor ld training (or the real-world consequences (S teinkuehler, 2004). The eff ect is two-fold, providing an environment free of the limited thinking that of ten accompany deep single-disciplinary training, while allowing for risk-free experimentation and thinking outside the box (Delwiche, 2006).

PAGE 25

25 This can lead to expanded learning of cultural and societal experiences as well as broad experimentation with new forms of human expression and endeavor (Delwiche, 2006). MUVEs hold considerable potential for deve lopment and training in complex social practices such as leadership, co llaboration, and relationships. Th ese worlds are complex rational groups distinguished by their soci al practices (Steinkuehler, 2004). When novices enter a virtual world, they can be progressively initiated into intricate soci al scaffolding by means of the support of other group members (Steinkuehler, 20 04). Virtual worlds are compelling because social relations, collaboration, and information sharing are essential ingr edients, and encourage collaboration both within and beyond the envi ronments parameters (Delwiche, 2006). According to Yee et al., (2007), more than half of those involved in virtual worlds have gained proficiency in mediation and lead ership, such as solving confli ct in groups. Although scholarly evidence is lacking on whether thes e skills transfer to real world situations, they have been recognized by employers (Brown and Thomas, 2006). Multi-User Virtual Environmen ts can also facilitate enhanced exploration of and experimentation with various social roles (Tur kle, 1995). The student can explore new social roles in an authentic situation while interacting with other individuals, a situation that has been shown to have significant psychol ogical and learning advantages (Turkle, 1995). Virtual worlds have been shown to promote role-playing beha viors (Delwiche, 2006), which have been shown to help students break away from the control of modern customs and beliefs (Luff, 2000), affect attitudes and behavior (Bell, 2001), and can have significant th erapeutic benefits (Douse & McManus, 1993; Hughes, 1998). For example, Peter Yellowlees (2006) documents the use of Second Life to help students experience the ro le of the patient. St udents learn about the subjective experience of psychosis as they naviga te through a virtual psychi atric ward. In this

PAGE 26

26 environment, users can literally see and hear hallucinations as a patient might, as they walk through the halls of the virtual hospital (Yellowlees, 2006, p. 441). This application of Second Life enables students to explor e and experiment with the role of the patient, thus gaining important insights into particular psychoses a nd developing a deeper em pathy than through more traditional means. Second Life in Education As a well-known and widely distribute d MUVE, Second Life possesses significant potential for innovations in education (Edwards 2006). For exam ple, Dave Taylor, knowledge transfer leader at the Nationa l Physical Laboratory, says that the use of SL opens up, new opportunities for collaboration across disciplines and geographies that would not otherwise occur (Edwards, 2006, p. 32). The development of complex social practices such as collaboration is enhanced by two major benefits: social context and vi sual context (Harris & Lowendahl, 2007). In a learning experience, studen ts are more likely to develop as leaders, collaborators, and relaters, experience flow, and d eeply experience alterna tive roles when there is a social and visual contex t (Harris & Lowendahl, 2007). Second Life and other virtual wo rlds also possess great poten tial for research. According to Yee, et al. (2007) and Blasc ovich et al. (2002), soci al customs of gender, interpersonal space and eye gaze transfer into virtual environments even though the method of movement is entirely different (i.e., via keyboard and mouse as opposed to eyes and legs). As a result, these online environments are also being inves tigated as unique research sites for clinical therapy, the social sciences, and other academic fields. The New Media Consortium (2007) is attempting to work toward interdisciplinary research in the social sciences, but the inclus ion of other academic disciplines could provide great benefit. Much could be learned by in cluding biology, chemistry,

PAGE 27

27 physics, and other disciplines not typically incl uded in social scienc e research. These environments can be customized through the use of in-world 3D construction tools. Locations can be produced to be relevant to any academic discipline or content area, places and artifacts can be as true to life and customized, or as ge neral and vague as desired. Even objects of large or micro scale can be easily portrayed. The combined effect of recorded interactions and customizable environments provide the ideal conditions for future research. For all of its potential and current uses, ther e are still major educa tional aspects that are underdeveloped in MUVEs such as SL in par ticular, scalable, replicable, objective, empirical investigations, as well as development of best practices (Oliverio & Beck, 2009; Delwiche, 2006; Keesey, 2007). Whereas the Second Life Be st Practices in Education Conferences ( http://slbestpractices2007.wikispaces.com ) are a m ajor step forward in developing best practices in education, they also served to highlight th e need to move beyond simp le re-creation of the classroom experience to more of an emphasis on creative practices when using MUVEs (Keesey, 2007). However, many emergent educational aspects of MUVEs still need to be researched. One such aspect is the influence of avatar choice (i .e. gender and ethnicity) on teachers expectations and perceptions of student success. This is important because teacher expectations and perceptions regarding student gender and ethnicity are cri tical to students academic accomplishments (Dee, 2004; Ferguson, 2003; Roscigno, 1998; Alexander, Entwisle, & Thompson, 1987; Irvine, 1986; Rist, 1970). As a result it is important to study the influence of avatar choice (i.e. gender and ethnicity) on teacher perceptions and expectations of student success. Differing kinds of interaction that ar e pertinent to teacher-student interactions are discussed below.

PAGE 28

28 Face-to-face Interaction In face-to -face interaction, physical proximity is a given. Interac tion and communication are inseparable from the physical body. In essen ce, interaction is on a basis of what-you-see-iswhat-you-get. Because physical presence gives immediate represen tation, identity is more tied to that physical presence (Goodwin, 1986). Prev ious research on faceto-face interaction has shown that participants use audiovisual cues to coordinate their activities. For example, Goodwin (1986) found that participants use gaze direction to coordinate turn taking and that gestures and hand motions may be used to indicate story structure and transitions from one conversational activity to another. Also, Jefferson (1989) found that the placement of laughter particles may indicate how to in terpret and respond to a speakers utterance. The accountability of real-life social activities en ables participants to achieve tig ht coordination in interaction (Moore et al, 2007). Individuals employ a variou s types of observational information regarding the actions of others so as to understand their actions and to closely s ynchronize their actions with those of others. This information consists of: (1) live unfolding of turns-at-talk; (2) the understanding of embodied activities; and (3) th e focus of eye gaze for gesturing purposes (Moore et al., 2007). As a result, gaps and overlaps between turns are minimal. Turn-taking in real-life conversation tends to be tightly coor dinated and fast, which is made possible by the predictability of turns-at-talk (Moore et al., 2007). In face-to-face interaction a person can easily display their on-going activities through their visual conduct and demeanor (Hin dmarsh et al., 1998). People us e gestures in at least four distinct ways: (1) to perform an action through the gesture alone, such as a greeting with a wave or an agreement with a nod; (2) to refer to objects by pointing; (3) to empha size particular words; and (4) to describe objects by simulating their shape, spatial relationships or motions (Kendon,

PAGE 29

29 1980; McNeill, 1992). A key factor in the production of gestures, regardless of the type, is that they must be coordinated with the eye gaze of the intended recipient (Hindmarsh et al., 1998; Moore et al, 2007). Because gestur e is a visual mode of communi cation, it doesnt count if the recipient does not see it. A basic resource in this kind of c oordination is holding (Kendon, 1980; Hindmarsh et al., 1998) the gesture until you can see that the recipient has seen and understood it (Moore et al., 2007). In face-to-f ace interaction, participants can read and sense a situation, respond to it, and adjust communica tion and action according to this reading. In general, in these interactions participants monito r each others utterances-in-progr ess to determine when and how to respond to each other (Moore et al., 2007). In oral conversation without face-to-face contact (such as telephone calls), partic ipants can rely on verbal information and paralanguage to coordinate turn exchange (Sacks et al., 1974). One of the advantages of face-to -face interaction is that interpersonal effects develop faster than with technology based methods (Walther and Tidwell, 1996, McQuillen, 2003). Face-toface interaction occurs in a co llaborative environment conti nuously kept in sync by joint modification and improvement and includes the c ooperative dedication of i ndividuals in the coconstruction of the message (Galimberti, 1994: Goodwin & Heritage, 1990). It permits criticism that allows the social significance of the message to be processed directly. This type of feedback has shown to lead to superiority in distinguishing between relevant and irrele vant ideas, ability to achieve consensus in discussion, and an increas ed conformity and conve rgence of ideas (Kerr and Murthy, 2004). The limitations of face-to-face communication are inherent in its properties. Because interaction is inseparable from the physical body, many-to-many communi cation is not possible (Warschaur, 1997). Also, less reflec tion is encouraged in face-toface discussion because of the

PAGE 30

30 real time immediacy of conversation (Zornoza, Ripoll, & Peir, 2002). Fewer ideas are generated in a problem solving process due to ha ving to face the others non-verbal objections and interruptions, and a high frequency of interp ersonal conflict occurs in face-to-face discussion (Zornoza, Ripoll, & Peir, 2002). As stated above, participants know more about each other in face -to-face communication because of the presence of audio-visual cues an d identity markers (Moore et al, 2007). This leads to a more realisti c perception of self (Walther & Tidwel l, 1996). Relationships are initially anchored by the everyday social and cultural co nstruction of gender, age, race, or by cultural conventions that cover other human relationships (Carter, 2004). Communication is unequal in participation due to these same factors. As we turn to examine computer-mediated interaction, it will be important to keep in mind the characte ristics of face-to-face interaction. Seeing the differences and similarities betw een different forms of interacti on will help us to compare them with interactions in MUVEs. Computer Mediated Communication Interaction Berger (2005) posits that new technologies do not sim ply provide communicators with more channels, but potentially alter the way they conduct face-to-face interactions. Through computer mediated communicati on, users and information hosts are interconnected within the existing telecommunication infrastructure, enabli ng computers to process messages and serve as a mass medium (Berger, 2005). The responsiven ess, the interconnect ed, point-to-point configuration of particip ants, and the rising popularity of CMC ha ve led some to talk about it as an interpersonal and many-to-ma ny mass media cluster (Rafaeli, 1986, 1988). This is in contrast to the more traditional one-to-m any description of broadcast and print media formats that have defined mass media since the Gutenberg press. CMC has been shown to be a type of

PAGE 31

31 communication that is significantly differe nt than non-digital written communication (Warschauer, 1997). The primary cause for this di fference is that interact ion with a traditional computer interface takes a longer amount of time than regular face-to-face interaction (Warschauer, 1997). Additionally, the lack of qual ities like posture, facial appearance, and voice tone also contribute to this difference. Another difference is the time and place independent nature of communication which a llows participants to compose and collect mail at any time in any location (Warschauer, 1997). CMC tools have been proven to increase soci al interaction in th e online environment (Repman, Zinskie, & Carlson, 2005). Prior rese arch on CMC support systems has revealed two robust effects on the group inter action process (Zornoza, Ripoll, & Peir, 2002). First, CMC promotes more equal participa tion and influence among group me mbers. Second, by restricting the exchange of interpersonal cues, CMC seem s to exert a depersona lizing, task-orientating effect on group interaction compared to FTF meetings (Kiesler et al., 1984; Siegel et al., 1986). A positive outgrowth of this effect seems to be that the frequency of conflict that commonly occurs in face-to-face interaction is reduced by the leaner nature of CMC (Harmon, 1998). This seems to occur by focusing on ideas and issues rath er than on personalities (Poole et al., 1991). A number of studies also show that a higher proportion of group communications deals with instrumental versus expressive functions in CMC interactions as compared to face-to-face discussions (Hiltz, Johnson, & Turoff, 1986; Siegel et al., 1986). CMC studies that have examined the role of gender and ethnicity have attempted to extend theories of face-to-face communi cation patterns (Eakins & Eaki ns, 1978; Lakoff, 1975; Tannen, 1990) to CMC modes. For example, women's on line interaction is more similar to a rapport based style, compared to a report based style that men seem to support (Kaplan and Farrell,

PAGE 32

32 1994). Similarly, Kaplan and Farrell (1994) sh ow that messages composed by females are shorter and their involvement is motivated by th eir intent to continue the conversation, as compared to the intent to reach agreement on an issue. Historically, face-to-face communication situations that recognize the individuals involv ed have put women in a less empowered position in contrast to men (Tannen, 1990). The absence of gesture and socio-cult ural information that identifies gender acts to lessen conversational gender-based limitations in a CMC environment (Matheson & Zanna, 1992; We, 1993). Thus, men ma y sense less of a requirement to present an appearance of social independence (Lee & Mi chael, 1999). This may result in men using interactional approaches that are normally s een in women (Lee & Michael, 1999). Conversely, women in a CMC environment may exhibit inter actional traits that are traditionally more masculine (Lee & Michael, 1999). Face-to-face communication takes place in a collaborative envir onment that is continually kept in balance through mutual modification and adjustment (Galimberti, 1994: Goodwin & Heritage, 1990). However, CMC takes place in a very different environment that is much less collaborative because of the special affordan ces of the technology (Brennan, 1991). Two normal characteristics of face-to-face communicati on are absent in most CMC environments: The cooperative dedication of i ndividuals, and the co-constructi on of the message. Second, the constructive criticism that pe rmits a true understanding of th e message to be processed (Mantovani, 1996). CMC's characteristic of enabli ng communication between large groups of people is similar to a group verbal face-to-face co mmunication. However, two diffe rences exist. First, CMC communication allows a group to tie reflection and interaction together by collaboratively developing knowledge. Second, CMC has proven to vary from face-to-face conversation

PAGE 33

33 concerning interruption, balance, turn-taking, decision making, equality, and consensus. Research has shown that CMC based communication tends to allow those who are regularly left out of conversation in a face-to -face environment to benefit from increased participation (Riva & Galimberti, 1998; Sproull and Kiesler, 1991; McGu ire, Kiesler, and Siegel, 1987; Huff and King, 1988). Why does equal opportunity occur more in a CMC environment than in a face-to-face environment? CMC lessens social cues such as a ssociated with status, gender, race, accent, and handicap (Sproull & Kiesler, 1991). It also lessens gestures or facial appearance like scowling and uncertainty (Finholt, Sproull & Kiesler, 1989). Finally, CMC seems to permit people to participate when and how fast th ey want (Sproull & Kiesler, 1991). As stated above, participants know less about each other in CMC interactions because of the absence of audio-visual cues and ident ity markers (Dubrovsky, Kiesler, & Sethna, 1991; Kiesler, Siegel, & McGuire, 1984; Kiesler, Zu brow, Moses, & Geller, 1985; Siegel, Dubrovsky, Kiesler, & McGuire, 1986). This leads to a more idealistic perception, and a selective selfpresentation of self (Walther & Tidwell, 1996). Also, relationshi ps are not based in the everyday social and cultural construction of gender, age, ra ce, or by cultural conven tions that cover other human relationships (Carter, 2004). As a resu lt, communication seems to be more equal in participation due to these same factors (Carter, 2004). As we turn to examine MUVE-based interactions, it will be important to keep in mind the characteristics of CMC and face-to-face interaction. Seeing the differences and similarities between these different forms of interaction will help us to compare them with interactions in MUVEs. Avatar-based Interaction in a Multi User Virtual Environment The word avatar b ecame popular in the modern technology based sense when it was used by Neal Stephenson in his 1992 novel Snow Crash. Generally defined, an avatar is not only a

PAGE 34

34 representation produced for use by an individual in a MUVE, but can be any image used to represent a person online (Lee & Shin, 2004). T ypes of avatars may include: animal, cartoon, celebrity, evil, real face, idios yncratic, positional, power, seductiv e, and many other options (Lee & Shin, 2004). As early as 1985, Chip Morningstar co ined the usage of the term avatar to denote the image of a person in virtual reality: a m ovable three-dimensional image that can be used to represent somebody in cyberspace, for exampl e, an Internet user (Castranova, 2001, p. 6). Seen from the perspective of virtual worlds, av atars may be described as the citizens of these worlds. These avatars, unlike those in video games, can be thoroughly customized and are intended for communication (Lastowka & Hunter, 2004). The normal participant dedicates much time and money to create and customize the avatar (Lastowka & Hunter, 2004). The player can freely within the confines of th e worlds reality select sex, a ppearance, profession, and physical features (Damer, 1998). This possibility of rich customization is important to our study because it enables avatar choice (i.e. gende r and ethnicity) to be expressed in understandable ways. Identity experimentation with Avatars In 1995, Sherry Turkle pointed out that playi ng with identity m ight become one of the most important aspects of avatars. Like the anthropological concept of masking people use avatars to experiment with being another pers on or trying out different roles and functions (Turkle, 1995). Avatars allow users to constr uct their identity by choosing a name, changing physical characteristics and apparel, and thr ough talking, discussing, a nd negotiating about the identity that users want to show in real time (Talamo & Ligorio, 2001). When people use avatars, they choose to use certain relevant characte ristics of their identities as strategic resources to enhance their participation and the overall effectiveness of the community (Widdicombe, 1998). Some researchers consider the characteristics of gender and ethnicity as socially

PAGE 35

35 constructed within communities (W aters, 1990; Nagel, 1994). Eviden ce of this type of identity experimentation can be seen in the wide range of gender and ethnicities that populate current MMORPGs and MUVEs. Social Interaction with Avatars In face-to -face interaction, knowing the identity of the people with whom we interact is crucial to comprehending and assessing the inte raction (Matusitz, 2005). However, in MUVEs identity is often unknown (Turkle, 1995). Regardless, MUVEs create a sense of presence, immersion and engagement that seems to enhance social interaction (Schiano, 1999). The social interactions that players have a nd the social networks they build up are what make virtual worlds able to retain players long-term (Bartle, 2003). Some scholars argue that as the avatar begins to resemble human communication behaviors, the interactions will also become more effective (Taylor, 1999). Some research on avatar-based communicati on does exist. In 2002, Lars Qvortrup published comprehensive findings from a Danish research proj ect on virtual reality, also covering some early studies of avatars social interaction (Q vortrup, 2001, 2002). Also, Ralph Schroeder (2002) and Sigga rd Jensen (2007) have publis hed the findings from several international projects on avatar interaction. Fo r example, for the purpose of project focused meetings: 1. It is challenging to arrange meetings between avatars, manage attention and focus, and construct a group focal point (Siggaard Jensen, 2007) 2. Meetings between two avatars seem to occur w ithout difficulties (Sigga ard Jensen, 2007). 3. Large group meetings between avatars often experience problems managing, interacting, and concentrating. This is especially the cas e if the content of th eir meeting goes beyond mere exploration and play (Siggaard Jensen, 2007). 4. It is challenging to maintain a conversation between more than two avatars in text chat while exploring the virtual envir onment (Siggaard Jensen, 2007).

PAGE 36

36 5. The avatars movements do not accurately reflect a physical bodys movements and gestures that are regularly utilized to manage interact ion (Siggaard Jensen, 2007). Conversation analysis has shown th at a generic feature of real-life social interaction is that gaps and overlaps between turns are minimal (Sack s et al., 1974). In re al life, one second is considered the maximum sta ndard length of a pause with in conversational sequences (Jefferson, 1989; Schegloff and Sacks, 1973). Turn-t aking in real-life conve rsation tends to be tightly coordinated and fast, which is made possible by the projectabil ity of turns-at-talk (Jefferson, 1989). By hearing a turn unfold in real-time, a hearer ca n anticipate where the turn is likely to end and can start his or her turn prom ptly at the next transition relevance place (Sacks et al., 1974). In MUVEs that onl y employ CMC communication this kind of projectability and tight coordination among turnsat-chat is usually not possible. Most MUVEs employ a text-chat system very similar to IM in which you cannot see a pl ayers turn unfold in real time (Garcia and Jacobs, 1999; Herring, 1999; and also Cher ny, 1999). In IM the construction of chat messages is private and on ly becomes public all at once when the player presses the ENTER key (Garcia an d Jacobs, 1999). This simple feature makes the organization of turns and sequences dramatical ly different from voice interacti on as Garcia and Jacobs (1999) demonstrate: you cannot achieve one speaker at a time, pairs of actions (e.g., questionanswer, greeting-greeting) become interspersed with other pairs, lengthy pauses develop between pairs, turns are not repaired by others mid-course, and more (C urtis, 1996). Because they employ IM-style messaging systems, chat in todays virt ual worlds exhibits all of these organizational characteristics (Curtis, 1996). Voice-enabled chat has somewhat changed th is situation. Turns are now heard unfolding in real time, which appear to eliminate many of the problems discussed above. The results of voice enabled chat are still inconc lusive as no research has yet been done on them, but they may

PAGE 37

37 be similar to research done on oral conversati on without face-to-face c ontact (such as telephone calls). This research shows that people can re ly on whether utterances are possibly complete (Sacks et al., 1974) and intonation that indicates a speaker is done to coordinate turn exchange. However, users are still not able to see and process the visual cues present in a face-to-face conversation. One feature of human bodies, whic h has an enormous social impact, is that you can see what a person is doing, to varying degr ees, simply by looking at them (Moore et al, 2007). More research is needed on voice enabled chat to determine the full effect of this technology on social interaction. Condition Awareness and Gestures Despite the fact th at MUVEs provide avatars, they actually reveal very little information about a users condition than the physical body does (Moore, et al., 2007). In face-to-face communication, individuals utilize various types of information about the actions of others so that they can understand these actions and then pr oceed in synching their own actions with them (Moore et al., 2007). Avatars can appr oach each other, face each othe r, gesture to each other, in some cases exchange facial expressions, and more Social interaction in MUVEs is neither only an instance of computer-mediated communicati on nor voice enabled ch at (Oliverio & Beck, 2009). In both cases, communication involves two different kinds of awareness information: (1) real-world, or what a player is currently doing in the physical world and (2) in-game, or what a player is currently doing in the virtual world (Moore, et al., 2007). For example, when initiating a conversation with a nother avatar, one may be interr upting a real-world conversation the person is having with a family member or an in-game conversation the avatar is having with a third avatar. Ideally it would be useful for a player who is about to initiate a conversation to know if the recipient is currently busy in the real world or in the game world. However, the

PAGE 38

38 strategies for providing these two types of awaren ess information are very different. Real-world awareness information, because it is external to th e system, requires the use of sensors to monitor the players physical environment (Schroeder 2002). On the other hand, in-game awareness information is internal to the system itself and so sensing is not an issue, only how to present player-activity information to other players (S chroeder, 2002). In todays MUVEs, like in earlier CMC (Bowers et al., 1996), li ttle of either type of awarene ss information is usually made available to players (Qvortrup, 2001, 2002). Relaying in-game awareness information about a fellow players current on-going activities in the game world is vital for managing player-toplayer interaction. Acco untability and tight coordination depe nd on participants access to these kinds of observational info rmation (Qvortrup, 2001, 2002). In face-to-face interaction, people communicate not only with their mouths but also with their hands and bodies (Moore et al., 2007). They can use their bodies to create an infinite number of unique gestures, just as they can put words together in an infinite number of unique combinations (Moore et al., 2007). People use gestures in at least four distinct ways: (1) to perform an action through the gest ure alone, such as a greeting with a wave or an agreement with a nod; (2) to refer to objects by pointing; (3) to emphasize particul ar words; and (4) to describe objects by simulating their shape, spatial re lationships or motions (Kendon, 1980; McNeill, 1992). In most MUVEs, gestures cannot be tightly coordinated w ith words in the chat because gesture and chat must be done as separa te commands (Qvortrup, 2001, 2002). And in most systems, a gesture consists of two parts, an an imation in the avatar and a text emote describing the gesture (Schroeder, 2002). Solving the te chnical challenges in simulating face-to-face interaction will become increasingly important as MUVEs are used more and more as

PAGE 39

39 simulations for training in real-world skills. Th is is also one of the reasons why the study of social interaction by avat ars is so important. Understanding MUVEs and interactions within MUVEs helps us to grasp the importance of this study. Because social interaction is important to learning (Vygotsky, 1978), and people interact via avatars in MUVEs (Lastowka & Hunt er, 2004), then it is im portant to study social interaction in MUVEs. Additionally because avatar attributes can be customized to fine detail, it is important to study the impact of these attrib utes on avatar based soci al interaction (Lastowka & Hunter, 2004). Gender and ethnicity are proba bly the most obvious among the many possible avatar attributes that are available. The gende r and ethnicity of students have been shown to specifically influence teacher expectations and perceptions of student success in real life classrooms (Clifford & Walster, 1973; Guttm ann & Bar-Tal, 1982; Walther & Tidwell, 1996; Ferguson, 2003; Auwarter & Aruguete, 2008). Therefore, it is important to explore the influence of avatar choice (i.e. gender and ethnicity) on te acher expectations and perceptions of student success. However, in order to do that we need to develop a historic al and research based perspective on issues of gender a nd ethnicity in the classroom. Issues of Gender and Ethnicity in the Classroom The purpose of this study is to understand the influence of avatar choice (i.e. gender and ethnicity ) on teacher perceptions and expectations of student success. The gender and ethnicity of students has been shown to impact teacher ex pectations of students in face-to-face classrooms (Clifford & Walster, 1973; Guttmann & Ba r-Tal, 1982; Ferguson, 2003; Frawley, 2005; Van Duzer, 2006; Auwarter & Aruguete, 2008). However, no research has been found on the impact of gender and ethnicity biases in MUVEs.

PAGE 40

40 History of Gender and Ethnicity Issues An understanding of historical precedent in the education of wom en and ethnic minorities is needed in order to set the stage for this study. Having a historical perspective will help us to avoid the pitfalls of those who have gone before us and to create gender and ethnicity equitable classrooms in MUVEs. Current educational pr actice is founded on educational history. How teachers teach, interact with, and evaluate their students is rooted in historical precedent. Therefore, the examination of that history will help us to continue ou r ongoing improvement of instructor practices, inter actions, and evaluations. Women and minorities have struggled to achie ve equity in the classroom for over 300 years. For many years, women were raised to be no more than wives and mothers and to keep an orderly house (Beecher, 1842). Ethnic minorities we re taught to be content with positions in manual labor that favored their inferior intellect s (Anderson, 1988). Historical records show that educational classrooms were segregated accord ing to gender and ethnicity (Anderson, 1988). Curriculum was also very different, despite effo rts by many in touting the importance of women receiving an equal education with men (Wolls tonecraft, 1792; Beecher, 1842). Even in the present, many researchers have shown that gender and ethnic equity has no t yet been achieved in the classroom (Campbell, 1990; Fennema & Pe terson, 1985; LaFrance, 1981; Masland, 1994; Noddings, 1992; Sadker & Sadker, 1994; Fer guson, 2003; Auwarter & Aruguete, 2008). Much of what has been done on behalf of gender equity for women has had ulterior motives. For example, Horace Manns teacher professionalization opened the door to the teaching profession for women, although it was pr imarily financially motivated and not an assertion of the equality of women. In The I ndispensable Teacher, Mann wrote that women can teach just as good as a man, will cost only 2/3 th e expense, and will infuse moral purity that a

PAGE 41

41 man cannot (Cremin, 1957). His reasoning went a long way toward seeing women enter the teaching workforce. This in turn led to the creation of teaching colleges for women. However, even then the quality of the cu rriculum was inferior to those offered to men (Cremin, 1957). Throughout the course of American educat ional history, established, wealthy groups continually sought to marginaliz e and control the influence a nd education of minority groups (Anderson, 1988). Foremost among the minority groups affected were African-Americans. Blacks withstood a barrage of attacks from Southern planters, poor white farmers, and northern philanthropists that were designed to elimin ate or limit their access to equal education (Anderson, 1988). These efforts led to little or poor quality e ducation for African-Americans (Anderson, 1988). The efforts of the above men tioned groups to margin alize and control the education of African-Americans hi ghlight the educational goals of these interest groups. Far from academic excellence and intellectual developm ent, they desired African-Americans to learn the skills that would keep them in low income rural and urban employment (Anderson, 1988). In short, they wanted to use education to contro l the socio-economic makeup of the United States and particularly the South, in order to promote their own capitalistic goals of more wealth and power (Anderson, 1988). Lieutenant Henry Pratts appro ach to the education of Native American youth epitomized common educational approaches toward ethnic minorities (Reyhner & Eder, 1989). In essence, he wanted to kill the Indian and save the man. He completely removed Native American youth from their homes and geographic surroundings a nd immersed them into a way of life that emphasized the English language, religion and cultu re (Reyhner & Eder, 1989). This approach was also adopted by the public schools in th e 1800s and early 1900s (Tyack, 1974). They combined this approach of immersion into Ameri can culture with a Social Darwinist justification

PAGE 42

42 of unfair segregation of minorities into special educational tracks focused on manual labor and women into home economics tracks that prepared them for life as a homemaker, wife, and mother (Tyack, 1974). One example of this wa s the development of standardized tests to determine what track a student should pursue in high school. Tyack (1974) showed that poor and minority students were overwhelmingly placed in the vocational or industrial tracks, while students from middle and high income families we re regularly placed in the college track. Women were regularly shifted into a home economics track. Although scientific management resulted in more courses, the courses were st ill limited by track, and the tracks seemed to be limited according to race, gender, and class (Tyack, 1974). Research on Gender and Ethnicity Issues Research co ntinues to level the playing field for women and minorities. Researchers such as the Sadkers (1994), Greenberg (1994), Ca mpbell (1991), Noddings (1992), and Singer (1995) have worked tirelessly to raise the consciousness of public school educators and institutions on the issue of gender equity. Similarly, research ers such as Irvine (1986), Simpson and Erickson (1983), and Chavous, et al., (2008) have labored to increase awaren ess of ethnic inequities for teachers and their schools. However, many believe that gender and ethnicity biases still exist, even at the college level. Wo men and minorities still have to endure humiliation at the hands of insensitive instructors (Chavous, et al., 2008). Decision makers mu st attend to the plight of women and ethnic minorities. The equitable education of women a nd minorities is an absolute necessity in order to overcome our worlds ma jor problems (Sadker & Sadker, 1994; Chavous, et al., 2008). Gender and ethnicity stereotypes have a bi g impact on first impressions and for wider personal evaluations. These stereo types affect juror decisions (S igall & Ostrove, 1975), helping

PAGE 43

43 behavior (Benson, Karabenick, & Lerner, 1976), employment opportunities, and job evaluations (Dipboye, Arvey, & Terpstra, 1977). Stereotypical bias base d on gender and ethnicity is pervasive in many aspects of so ciety, including the educational system. Education provides, at least in theory, equal opportunities to all i ndividuals. In practice, however, differential expectations about students are typically present. Many physical characteristics are capable of evoking initial expectations and impressions, including ethnicity and gender (Braun, 1976; Brophy, 1983; Brophy & Good, 1974; Dusek, 1985; Fi nn, 1972; Ferguson, 2003; Auwarter & Aruguete, 2008). Many studies done in educational contexts choose head shot imag es and put them in a file that also contains the students school record. Teachers tend to utilize a students image and record to form impressions of the student. Th e file is used to study the expectations and perceptions of teachers toward students of part icular features. For example, teachers might receive a folder containing an academic profile of a student or a statement of a student's transgression along with a photograph (e.g., Clifford & Walster, 1973; Marwit, Marwit, & Walker, 1978). The academic description or the tran sgression is held constant while the attached photo is varied to portray different physical characteristics of the student. Patzer (1985) felt that the cumulative folder technique was an accepta ble method for studying physical characteristics because, when teachers review students, they co mmonly receive an academic description and a photo in the student's record. The cumulative folder procedure, however, is not without limitations. Relying on photographs for evaluating physical characteris tics may be problematic because photographs provide a static cue for basing attributions and evaluations. Providing only a static cue may oversimplify the conceptualization of physical characteristics by viewing them as a one-

PAGE 44

44 dimensional construct. Argyle and McHenry (1971) argued that photographs and brief exposure time do not simulate a real world situation. In that study subjects viewed an individual with or without glasses for either a brief durati on (15 seconds) or an extended interval (5 minutes). Individuals wearing glasses were evaluated as other cues such as dynamic expressive style, body characteristics, and attractiveness of dress co ntribute to forming overall impressions (Riggio et al, 1991). These cues are typically neglected in this type of res earch. On the other hand, Zajonc (1980) claimed that clear affective reactions to another person can develop in just a fraction of a second. Of course, it is possibl e that both positions are correct That is, judgments may be formed in a few seconds or less, but, neverthe less, those judgments may well change when the target person is presented for a sustained period of time. The effects of physical characteristics have been studied with respect to a wide vari ety of judgments, including: (a) intelligence and academic potential, (b) grades and achievements, (c) various social skills, and (d) miscellaneous attributions. Ethnic Minorities Biases Research on stereotypes has of ten scr utinized the degree to wh ich different signs, such as ethnic origin, social class, ge nder, or physical appearance tr igger stereotypi c reactions. For example, studies that have explored stereotypic perception of people due to their ethnicity have revealed that Israeli Jews of Asian-African an cestry are perceived as less intelligent, less disciplined, less motivated to achieve, less polite, less sociable, less rational less considerate, but more emotional, more generous, and more peacef ul than Israeli Jews of European-American origin (e.g., Peres, 1971; Rim & Aloni, 1969; Yinon, Abend, & Chirer, 1975). Many of the attributes that changed as a function of the signs revealing stereotypic perceptions are applicable to teachers' evaluatio ns and expectations of academic achievement in

PAGE 45

45 classroom situations (Braun, 1976; Brophy & Good, 1974). As a result, it is not surprising that teachers perceive African-Amer ican children as being less acc ountable for achievement and ascribe their triumphs mostly to luck or the ha rd work of the school system. However, teachers ascribe their failures to environmental limitati ons (Wiley & Eskilson, 1978). Studies found that teachers perceive students of different ethnicities more negatively and expect them to have lower grades than students of sim ilar ethnicities (e.g., Cooper et al., 1975; Miller, McLaughlin, Haddon, & Chansky, 1968; Ferguson, 2003). Ethnicity is a potent source of input into t eachers' impressions. The influence of ethnic stereotypes in teacher expectancies was first suggested by Kenneth Clark in 1963. Later research found that teachers (a) rated Bl ack students less favorably, (b ) treated Black students less favorably in the classroom, and (c) held lower academic expectations for Black students than they did for White students (e.g., Cooper, Baron, & Lowe, 1975; DeMeis & Turner, 1978; Rubovits & Maehr, 1973; Ferguson, 2003). Furthermor e, in a study of the perception of facial beauty, Cross and Cross (1971) found that Blacks were rated le ss positively than Whites by both Black and White teachers. Adams (1978) interviewed 112 African Americ an and 128 Caucasian Head Start teachers with regard to initial teacher expectancies base d on physical characteristics, gender, and race. Caucasian students and girls were rated as more intelligent and as higher achievers than were African American students and boys. The aut hor concluded that, although the physical characteristics of an individual influenced the initial expectation of the preschool teachers, ethnicity exerted the strongest influence. Kehl e, Bramble, and Mason (1974) had previously found similar effects for Caucasian females compar ed to African American females and males.

PAGE 46

46 Gender biases Gender is an other factor that may affect teach er judgments. Compared to females, males are called on more frequently and given more attention (e.g., Hall & Sandler, 1984; Sadker & Sadker, 1994). Girls, however, tend to be favored by their teachers, perhaps because of the stereotype of the quiet and passive female (Worrall, Worrall, & Meldrum, 1988). Females are seen as more sensitive to the feelings of others more tactful, more dependent, more motivated, more passive, more subjective, neater, and quieter than males (Broverman et al., 1972; Feldman & Kiesler, 1974; Richmond-Abbott, 1979; Rim & Aloni, 1969; Ruble & Ruble, 1980). Males appear to view school as less impor tant in their lives as females (Clark et al., 2006). In a national survey, male students consisten tly reported that they did not listen in class, complete assignments, and did not do their best work (NCES, 2005). Females viewed their academic work as interesting and meaningful compared to males who did not appear to see the critical nature of their academic performance as it relates to future occupation (NCES, 2005). Also, current data on academic performance clearly shows males falling behind females (Burns and Bracey 2001; Clark et al. 2006; Kafer 2004). The evidence of the influence of student gende r on teacher perceptions and expectations of intelligence, academic potential and academic performance is mixed. Kehle, Bramble, and Mason (1974) gave teachers photographs of students and had the teachers rate the essay performance and personality characteristics of th e students. Teachers held higher expectations for Caucasian females than they did for Caucasian males. Other studies also found that the effect was stronger for females than for males (Adams, 1978; Hore, 1971). However, other studies on physical characteristics and teacher expectancies have not found significant gender differences in

PAGE 47

47 teacher expectations and perceptions (Clifton & Baksh, 1978; Felson, 1980; Salvia, Algozzine, & Sheare, 1977). Ethnicity and Gender Issues Conclusion The stereoty pic perception is alive. In face-to-face classrooms, student gender and ethnicity influences teacher percep tions and expectations of student success. With the increasing prevalence of MUVEs in society and in education, we need to determine the influence of avatar choice (i.e. gender and ethnicity) on teacher perceptions and expectations of student success. Conclusion New technologies are not created in a vacuum Instead they are alrea dy rooted deeply into an established culture (Rogers, 2003). MUVEs may be relatively newly born, but they are already filled with cultural ideas about gender and et hnicity. As a result, it is important that we do not assume a blank slate mentality about concepts like these in the virtual world. We need to remember that they have roots in the physical world that suppor t and extend into the virtual and then study how the virtual world changes the expressions of these concepts. Understanding social interaction within face-to-face, CMC-based and MUVE environments along with the history and research of issues of gender and ethnicity in the classroom help us to see the need for this study. The importance of social interaction in technological and nontechnological learning environments highlights the n eed to study these kinds of interactions in all environments, especially MUVEs. Additionall y, because students in MUVEs can choose their avatar attributes to fine deta il, it is important to study their impact on avatar based social interaction. Among the many possible choices of avatar attributes are gender and ethnicity. Gender and ethnicity discrimination against students has been a perennial concern for researchers and practitioners in educati on (Sadker & Sadker, 1994; Ferg uson, 2003; Auwarter & Aruguete,

PAGE 48

48 2008). Specifically, the gender and ethnicity of students have been shown to specifically effect teacher expectations of student s in real life classrooms (Clifford & Walster, 1973; Guttmann & Bar-Tal, 1982; Walther & Tidwell, 1996; Fe rguson, 2003; Frawley, 2005; Van Duzer, 2006; Auwarter & Aruguete, 2008). Unfort unately, the current lack of research on the influence of avatar choice in MUVEs lead to the problem s of unfair disadvantage and unfair advantage mentioned earlier. As a result, it is important to study the influe nce of avatar choice (i.e. gender and ethnicity) on teacher expectations a nd perceptions of student success. In this chapter, a review of literature was introduced that established the foundation for an understanding of Multi-User Virtual Environments and their use in education. The use of a particular MUVE called Second Life was also di scussed as well as its uses in Education. A context was also developed for the studys impact on social interaction through a discussion of face-to-face, computer-mediated and avatar-b ased interaction in a multi-user virtual environment. Finally, a discussi on of the issues of gender and et hnicity biases in education was presented. This literature review sets the stage for the ar gument that an understand ing of the impact of avatar choice (i.e. gender and ethnicity) on t eacher expectations and perceptions of student success will help teachers in MUVEs to discover potential biases and prejudices toward some students, as well as address the problems of unfair disadvantage and advantage for student avatars of all ranges of detailed customization.

PAGE 49

49 CHAPTER 3 METHODOLOGY Introduction The purpose of this study was to exam ine the influence of avatar choice (i.e. gender and ethnicity) on teachers perceptions and expectatio ns of student success. This chapter describes the plan for answering the research questions de scribed in chapter 1. It also explains why the procedures for studying this topic were used, and divides the research process into three major steps: 1) conducting an exhaustive search of the sources of teachers in Second Life; 2) from these sources, extracting a sample, administering th e survey, and analyzing the data; and 3) accumulating and summarizing the findings. This is then followed by a conclusion. Research Design Mayer (200 3) calls for a return to evidence-b ased rather than dogma based research in education, asking for the identification and formul ation of specific studie s that involve rigorous testing. Other researchers have joined th is movement, requiring higher standards of experimental and non-experiment al methods. Schrum, et al. (2005) renewed Mayers (2003) call, stating that educati onal faculty need to endorse a platinum standard for "...scientific research that involves rigorous resear ch" (Schrum, et al., 2005, p. 204). Dawson & Ferdig (2006) clarified that this standard needs to be very sp ecific, containing strong definitions, heuristics, and rubrics; as well as be inclusive of all research methodologies. Schrum, et al. (2005) suggest that the future research agenda s hould address a) the connection between teacher beliefs about technology, b) teacher practices with technology, and c) student learning outcomes, while Dawson & Ferdig (2006) broaden the scope of th e agenda to include all educational technology subfields.

PAGE 50

50 A quantitative research design was used to asse ss the impact of the gender and ethnicity of the student avatar on the teachers expectati ons and perceptions. This outcomes focused, positivistic methodology was selected to provide for a scalable, replicable, objective, empirical investigation. The adoption of a positivistic pers pective made it possible to test theory through analysis of data collected from research subjects, develop best possible estimations and an indication of the accuracy of outcomes (Doole y, 2001; Duncan, 1975). Additionally, through the use of quantitative data collection we were able to sample a large number of individuals and project these results onto a population. Quantitative data generated in this study was analyzed using statistical procedures. The purpose of a quantitative study is to unc over the connection between an independent variable and a dependent variab le in a population. The quantitativ e research design chosen for this study was descriptive partic ipants are only measured one tim e. Descriptive studies describe observations in a systematic manner in order to show patterns and c onnections that might otherwise go unnoticed. A descriptive study needs a large sample in orde r to obtain a precise estimate of the relationship between two variables. There is a smaller chance that an estimate of relationship between two variables will be biased if the participation rate is high and if the sample has been randomly chosen from a population. Additionally, random assignment of participants to treatment conditions also lower the chance of bias. Population and Sample The proposed research study sought to survey teachers in Second Life regarding how avatar choice influences teachers p ercepti ons and expectations of student success. A partnership with the Soci al Research Foundation ( http://www.socialr esearchfoundation.org ) was established to facilitate the co llection of data. The Social Re search foundation was a non-profit

PAGE 51

51 organization dedicated to changi ng lives through online, interactiv e education programs. This organization operated the First Opinions Panel, wh ich was the largest consumer research panel in Second Life. Through this organization panelist s were identified who were teachers in Second Life or who planned to teach in Second Life. If not enough teachers were identified through the First Opinions Panel, other sour ces of data were to be the Second Life for Researchers and Second Life for Educators listservs, and the follo wing Second Life groups that operate in Second Life: Real Life Education in Second Life Second Life Grad Student Colony, Second Life Research, K-12 Educators, and Community Colleges in Second Life. The Social Research foundation and the First Opinions Panel have already collected data on 33 different real life attributes of their panelists ranging from gender, age, marital status, # of children in household, family income, fam ily net worth, home ownership, car ownership, education completed, business ownership, person al investing, televi sion viewership, news source, Internet daily use, research topic pref erence, time in second lif e each week, residency (overall time in Second Life), SL land ownership, spending in SL, primary motivation for coming to SL, primary business in SL, SL business size, significance of SL business to real life business, types of events most attended in SL and groups led in SL. The Social Research Foundation describes itself as being initially focused on the re search of online educational solutions for vulnerable groups in society, e.g. ch ildren/youth, immigrants, seniors. It is a nonprofit organization established in 2003 that is dedicated to changing lives through online, interactive education programs. The organization has created the First Opinions Panel, which is the largest standing panel in Second Life with thousands of members. Teachers and prospective teachers in Sec ond Life come from a wide variety of backgrounds and teach students ranging from 13 to 83 in courses and informal seminars that vary

PAGE 52

52 from the biological sciences to English as a second langua ge. A significant number of universities, colleges, schools, organizations and businesses are expl oring the educational potential of Second Life. S econd Life education related websites number more than two hundred. The purposes of these groups involvement in SL are as varied as their creators. Individuals were recruited from the First Opin ions Panelists who indi cated that they were teachers in Second Life. The Social Research Foundation recruited participants by sending an instant message to all panelists who fit their profile of being a teacher or prospective in Second Life. Interested participants then clicked on a li nk to be taken to the study survey. The top 80% of Second Life residents hail from more develo ped nations, but a significant portion also log on from places like Africa, Southeast Asia, and Lati n America. The typical resident is male, between 25-34 years old, and among the top 5% of the worlds wealthiest i ndividuals. This was a concern because it was not completely represen tative of teacher demographics worldwide, but is typical of SL teachers. Privacy and Confidentiality This research study fell under the s crutiny of th e University of Florida Institutional Review Board (IRB). Behavioral and Non-Medical IRB, designated IRB02 within the University of Florida system, is responsible for reviewing and monitoring research with human subjects conducted at the University of Florida. The board reviews research studies that involved behavioral observations and r ecordings, non-invasive physiological recordings, analysis of documents that were previously gathered for non-research purposes, evaluation of behavioral/social interventions or manipulations, educational a ssessments, interviews, surveys, cognitive tests, and taste/food evaluations of wholesome food within FDA guidelines. The University of Florida IRB (2007) accepts as basi c principles those expressed in the Nuremburg

PAGE 53

53 Code (1947), the Declaration of Helsinki (revise d 1975), and the Belmont Report (1979). Before engaging study participants, a research proposal was prepared, presented, and approved by the University of Florida IRB (Appendix A). Data Integrity and Security The dawn and progression of the Internet has m ade data privacy and protection issues of prime importance for government entities, ed ucational institutions, private industry and individuals. It is the priority of the researcher to ensure the privacy and protection of the data collected from the participants. Therefore, a structured policy has been adopted for data storage, transformation, and reporting. The Social Res earch Foundation offers a secure website and database from which panelists could log in and take surveys and in which data was securely stored. All data was de-identif ied by the Social Research Fo undation prior to the researcher receiving it and was then password protected on the re searchers hard drive. Sampling Procedure A large sam ple of at least 383 participants was selected for this study. This was important because a large sample is usually needed for an accurate estimate of the relationship between variables. However, the sample for this study wa s not random in nature. This is due to the fact that the number of teachers availa ble to participate in the study were only a little more than the sample needed. This is a concern because random sampling procedures are utilized to ensure an unbiased cross section of participants (Dillman, 2007) and control for selection threats to validity (Dooley, 2001). A selection threat is any factor other than the pr ogram that leads to differences in results. Whenever we suspect that outcome s differ in a group because of prior sample differences we are suspec ting a selection bias.

PAGE 54

54 Instrumentation This study employed an online survey adapted from research by Clifford & Walster (1973) and Guttmann, J., & Bar-Tal, D. (1982). The instruments developed in these studies were originally created in the context of the impact of gende r, ethnicity and physical attractiveness on teacher expectations and percep tions. The procedure and instrument developed by Clifford & Walster (1973) were part of a study that was pl anned to discover the impact of the physical attractiveness of students on teachers expectations and perceptions of their students intellectual and social behavior. Results of Clifford & Walsters 1973 study were: 1. Teachers did perceive attractive students to have higher educational potential than unattractive children (Cliffo rd & Walster, 1973, p. 251). 2. Attractive students were expected to have highe r IQ's, to have parents who were interested in academic achievement, and to have a hi gher future education prediction than less attractive students (Clifford & Walster, 1973, p. 254). 3. Attractive students were expected to related better with fellow classmates than unattractive students (Clifford & Walster, 1973, p. 254). The procedure developed by Guttmann & Bar-Tal (1982) was part of a study that was designed to determine the stereotypic impre ssion of academic performance triggered by providing information only on the ethni c origin and sex of the student. On the basis of previous literature, it was assumed that given informati on about a students ethnicity and gender, the teachers' judgments would be based on their st ereotypic impression of the corresponding ethnic groups and gender. The study successfully descri bed the stereotypic impression of academic performance triggered by providing information only on the ethnic orig in and sex of the student. Multivariate analysis yielded a significant main effect only for ethnic origin (F(8,90) = 3.33; p < .001). Univariate analysis yielded main effects for ethnic origin on ability (F(1,97) = 14.51,~< .001 ( M = 4.1 1 vs. X = 3.62)); interest (F(1,97) = 18 43 ~< .001 ( M = 4.16 vs. M =

PAGE 55

55 3.55); and conditions at ho me (F(1,97) = 6.30, p < .001 ( M = 4.21 vs. M = 3.62)). The results of the analysis on diligence only approached si gnificance (F(1,97) = 3.40, p < .07). Thus, for example, subjects expected a student of Asian-Afri can origin to have lesser intellectual aptitude, less interest, and lower economic home conditions than a student of European-American origin. Finally, the analysis of variance of the rating of expectations of student's future achievement yielded a main effect only for ethnic origin (F(1,97) = 6.30, p < .001). That is, for example, a student of Asian-African origin ( M = 4.02) was expected to have lower academic achievement than a student of European-American origin ( M = 3.71). This study shows that when teachers are presented with information regarding a student s ethnicity they evaluate that student based on their own stereotypes. In this case, they expect Asian-African students to have lesser intellectual aptitude, lower interest in academics, less d iligence, and lower economic background than students of European-American origin. These results are in line with findings that show that Israeli Jews of Asian-African origin are per ceived as less intelligent and less motivated than Israeli Jews of European-American origin (see e.g., Rim & Aloni, 1969; Yinon, Abend & Chirer, 1975). The results of these studies reveal the presence of teacher stereotypic perceptions. They also show that these stereotypes impact teacher s expectations and perceptions. As typically conducted, there is no real contact with the people being judged. This type of research shows what people believe in the absen ce of any physical contact. These results provide eviden ce to support the ut ilization of these instruments and procedures with the study population. Data Collection The survey was adm inistered through the secure website of The Social Research Foundations First Opinions Panel. Participants were contacted vi a an email and in SL with a

PAGE 56

56 pre-survey notification letter (see Appendix B) and instant message. This email briefly explained the study and notified th em that an email was to be sent containing a link to a webbased survey. The email offered participants the opportunity to remove themselves, through the use of an unsubscribe link, from the sample popula tion and to contact the re search team for more information about the study. Approximately seven days after the pre-surv ey notification letter was sent to participants an email was sent directing them to the survey website. Nonrespondents received two subsequent follow-up em ails soliciting their pa rticipation, the first follow up was sent four days afte r the first solicit ation. After another four days, solicitation 3 was sent. Before sending solicitations ema il addresses were removed by those who had requested nonparticipation and those who have already respo nded. At the beginning of the survey, participants were asked to sign an informed consent form (see Appendix C) that explained the study and requested their involvement. After informed consent was given, participants were given phases one and two of the surveys explained below. Rationale for Study Format Careful consideration w as given to the format of the survey. According to Reeves and Nass (2002), putting information on different screens can be used to effectively chunk visual segments into meaningful pieces. This defines the important breaks, the equivalent of placing periods and commas in text. The placement of a br eak to another screen is a strong signal that one unit is finished and another will begin. The information th at surrounds the cut, just like information that surrounds any unit of actual behavi or, will be the most memorable. At the same time, the sequences of computer screens studie d by Reeves and Nass (2002) showed that those that had many cuts were given le ss attention than those that had none. In ot her words, media can easily exceed human processing capabilities. When this occurs people t une out. According to

PAGE 57

57 Mayer and Moreno (2003), humans possess separate systems for processing visual and audio content, each of these systems is restricted in terms of the quantit y that can be dealt with at one time, and significant learning engages cognitive processing that build s connections between visual and audio content. As a result, cognitive overload happens when the anticipated cognitive processing of the learner surpasses their available cognitive capacity. Consideration was given to both Reeves and Nass (2002) and Mayer and Morenos (2003) research. The following is an explanation of each page: 1. Page 1: The informed consent form was placed on a separate page due to the requirement that the user must indicate agreement in order to continue with the survey. 2. Pages 2 and 3: The letter to the teacher a nd the high school transcript were included on separate pages in order to provide a visual orienting response by the participant (Nass and Reeves, 2002). 3. Pages 4 and 5: Page 4 in the survey contai ned the image of the avatar and the machinima video of the avatar. Page 5 contained the opinion survey questions from Clifford and Walster (1972). 4. The video, image and transcript were separated from the opinion survey questions (Clifford and Walster, 1972, next page) due re search that has shown that putting questions on a different screen lessens th e potential bias of the participant (Reeves and Nass, 2002). Cognitive overload was also a considerati on because participants had to process a transcript, image, video, and answer the questions all on separate screens (Mayer and Moreno, 2003). That was four separate activit ies (transcript, image, video, questions) plus the activity of representational holding (recalling information from the transcript, image, video and questions on the previous pages after participant has moved on to another page). On one hand, placing all of the activities on one page could have led to a cognitive overload in essential processing. Essential processing consists of the cognitive processes that are utilized to understand the presen ted material (Mayer and Moreno, 2003). One solution to this difficulty was to allocate a certain amount of time between parts of the survey. In this way, the survey would be divide d into smaller, more easily processed parts. The individual could choose releva nt words and pictures in each part and as a result has the necessary time and ability to manage synthesize this information. When the participant is ready, he or she can then continue onward to the next part by clicking a button. This approach is also supported by research that ha s found that cuts between segments that are semantically related will be less intrusive than cuts between segments that are semantically

PAGE 58

58 unrelated (Nass and Reeves, 2002). However, th e problem of represen tational holding still exists. The participant must now hold a mental representation of the transcript, image and video in working memory while answering the opinion survey questions. Mayer and Moreno (2003) consider a method for reducing cogn itive load is to be sure that the learners possess skill in holding mental representati ons in memory. Because our sample is composed of teachers in Second Life, it is assu med that they would possess skill in holding mental representations in memory with a mini mum of mental effort. Thus the approach was selected to keep the transcript, image, video, and questions on separate pages. 5. Page 6: The next section of the survey was presented on one page listing several demographics questions (17). This format was decided on because it was desired for participants to be freshly visually oriented at the beginning of each part of the survey (Reeves and Nass, 2002). Cuts to a new scr een demand attention, a nd research has shown that brain wave response changes where cuts to a new screen appeared (Reeves and Nass, 2002). 6. Pages 7: The last section of the survey consisted of a page with the image and the video that was identical to the page mentioned above followed by three questions asking the three most salient characteristics of the avatar, the ethnicity of the avat ar, and the gender of the avatar. These questions were separated fr om the image and video due to the fact that research has show that putting questions on a di fferent screen lessens the potential bias of the participant (Reeves and Nass, 2002). Segm entation was also used in order to avoid cognitive overload (Mayer and Moreno, 2003). Avatar Creation Every effort was m ade to create avatars that would be correctly id entified by the teachers in the study. Groups of 4 (2 male, 2 female) from each ethnicity (Cau casian, African American, Hispanic/Latino, and Asian) were gathered to participat e in the avatar customization process. These groups took hours to tailor th e avatars from their own ethnicities to the point where they clearly represented an individual from their own ethnicity. This was done in order to make sure that each avatar accurately repr esented its ethnicity and gender. Study Progression During the first phase of the study, teachers were given a letter explaining the study, a student transcript, a photograph and m achinima video of the st udent avatar and an opinion

PAGE 59

59 survey (Table 3-1; Clifford & Walster, 1973). Each teacher was randomly assigned to one of 8 possible conditions. Conditions were based on the ethnicity and gender of the avatar photographs (see Appendix D). The letter explaining the study appeared on th e page after the informed consent (see Appendix E). The letter was used primarily to explain the studys purpose and to seek the teachers cooperation. It began by discussing the value and rationale of school records and the purpose of permanent record files, a nd transcripts. The letter then proceeded to explain that in an attempt to answer these questions, we were ex amining a variety of transcript forms used by different universities in Second Li fe. Of course, this was not the true purpose of the study. This letter was meant to deceive teachers into believ ing that this study was about the evaluation of transcripts so they would express any gender and ethnicity biases they might have. The high school transcript appeared on the next page. A high school student transcript was used in place of a student summary report sin ce the Clifford & Walster study was used with 5th grade teachers. High school transcripts best approximated the information given regarding achievement on the 5th grade student summary re port (Appendix F). The next page contained the avatar photograph and machinima video of the student avatar. The photograph and image was one of four male a nd four female avatars of the following ethnic categories: (U.S. Census Bureau, 2008): Asian, Hispanic or Lati no, Black or African American, White or Caucasian American (see Appendix D). The next page of the survey consis ted of the following five items: 1. I would estimate that the student has an IQ of In the original study (Clifford & Walster, 1973), possible answers ranged from 1 (96-100) to 7 (126-130). Our study rescaled the IQ variable as a ratio variable in order to achieve greater accuracy. 2. I would speculate that the student's social rela tionships with classmates are ---. Range of possible answers: from 5 (v ery good) to 1 (very bad).

PAGE 60

60 3. I would speculate that the student 's relationships with instruct ors are ---. Range of possible answers: from 5 (very good) to 1 (very bad). 4. I would guess that the student's attitude toward school is one of -. Range: from 6 (strong interest) to 1 (strong indifference). 5. I would predict that the stude nt would continue their edu cation through ---. Range: from 1 (1 year of college ) to 10 (Ph.D.). Space was also provided for the teachers to comment on their reactions to the transcript format and the type of information it provided. On the next page of the survey, teachers were asked to complete some demographic information (See Appendix G). Model Equation The m odel for this study is as follows: Dep variable (Q1-Q5) = F(Ethnic ity, Gender, Ethnicity*Gender) + (3-1) After this model was run and analyzed, the fo llowing interacting variables were added, one by one, and separately analyzed: 1. Gender of participant a. Dep variable (Q1-Q5) = F(Ethnic ity, Gender, Ethnicity*Gender, ParticipantGender, ParticipantGender* Ethnicity, ParticipantGender*Gender, ParticipantGender*Ethnicity*Gender) + (3-2) 2. Ethnicity of participant a. Dep variable (Q1-Q5) = F(Ethnic ity, Gender, Ethnicity*Gender, ParticipantEthnicity, ParticipantEthnicity*Ethnicity, ParticipantEthnicity*Gender, ParticipantEthnicit y*Ethnicity*Gender) + (3-3) Data Analysis Techniques The overarching question in th is stu dy was whether avatar choice affected teacher expectations and perceptions of student academic success. Many references have been given that have convincingly shown that teacher biases ba sed on the gender and ethnicity of students do exist in face-to-face clas srooms. But how does the world of the physical compare to the virtual?

PAGE 61

61 Does SL mirror the real world in terms of biases prejudices, and discrimination? Each of the next 15 research questions was ex amined with that intent. For ease of perspective, questions were broken down into those that are ethnicity related, gender relate d, and gender/ethnicity interactions. An Analysis of Variance (ANOVA) was used with data on teacher expectations and perceptions as dependent variables atte mpting to answer the following questions: Ethnicity Related o Research question 1: Does the perceived ethnicity of the avatar affect teacher perceptions and expectations? H0 : The ethnicity of the student avatar does not affect expected current and future performance of the st udent by the teacher. Independent Variable(s): Perceived ethnicity of the avatar o Research question 2 : Does the participant ethnicity of the avatar affect teacher perceptions and expectations? H0 : The ethnicity of the participant does not affect teacher perceptions and expectations. Independent Variable(s): Participant ethnicity o Research question 3: Is there an interaction betwee n the perceived ethnicity of the avatar and the ethnicity of the participant? H0 : There is no interaction between the perc eived ethnicity of the avatar and the ethnicity of the participant. Independent Variable(s): Perceived Ethnicity by Participant Ethnicity Gender Related o Research question 4 : Does the perceived gender of the avatar affect teacher perceptions and expectations? H0 : The perceived gender of the avatar does not affect expected current and future performance of the student by the teacher. Independent Variable(s): Perceived gender of the avatar

PAGE 62

62 o Research question 5: Does the gender of the partic ipant affect teacher perceptions and expectations? H0 : The gender of the participant does not affect teacher expectations and perceptions. Independent Variable(s): Participant gender o Research question 6: Is there an interaction betw een the perceived gender of the avatar and the gender of the participant? H0 : There is no interaction between the pe rceived gender of the avatar and the gender of the participant. Independent Variable(s): Perceived gender by Participant gender Interactions Between Gender and Ethnicity o Research question 7: Is there an interaction betw een the perceived ethnicity and gender of the avatar and teacher perceptions and expectations? H0 : There is no interaction betw een ethnicity and gender of the avatar and teacher expectations and expectations. Independent Variable(s): Perceived ethnicity by Perceived gender o Research question 8: Is there an interaction betw een the perceived gender of the avatar and the ethnicity of the participant? H0 : There is no interaction between the pe rceived gender of the avatar and the ethnicity of the participant. Independent Variable(s): Perceived gender by Participant ethnicity o Research question 9: Is there an interaction between the perceived ethnicity of the avatar and the gender of the participant? H0 : There is no interaction between the perc eived ethnicity of the avatar and the gender of the participant. Independent Variable(s): Perceived ethnicity by Participant gender o Research question 10: Is there an interaction between the ethnicity of the participant and the gender of the participant?

PAGE 63

63 H0 : There is no interaction between the ethni city of the participant and the gender of the participant. Independent Variable(s): Participant ethnicity by Participant gender o Research question 11: Is there an interaction between the perceived ethnicity of the avatar, the perceived gender of the avatar and the ethnicity of the participant? H0 : There is no interaction between the pe rceived ethnicity of the avatar, the perceived gender of the avatar, and the ethnicity of the participant. Independent Variable(s): Perceived ethnicity by Perceived gender by Participant ethnicity o Research question 12: Is there an interaction between the perceived ethnicity of the avatar, the perceived gender of the avat ar, and the gender of the participant? H0 : There is no interaction between the pe rceived ethnicity of the avatar, the perceived gender of the avatar, a nd the gender of the participant. Independent Variable(s): Perceived ethnicity by Perceived gender by Participant gender o Research question 13: Is there an interaction between the perceived ethnicity of the avatar, the ethnicity of the participan t, and the gender of the participant? H0 : There is no interaction between the pe rceived ethnicity of the avatar, the ethnicity of the participant and the gender of the participant. Independent Variable(s): Perceived ethnicity by Participant ethnicity by Participant gender o Research question 14: Is there an interaction between the perceived gender of the avatar, the ethnicity of the participan t, and the gender of the participant? H0 : There is no interaction between the perceived gender of the avatar, the ethnicity of the participant, a nd the gender of th e participant. Independent Variable(s): Perceived gender by Participant ethnicity by Participant gender o Research question 15: Is there an interaction between the perceived ethnicity of the avatar, the perceived gender of the avatar, the ethnicity of the participant, and the gender of the participant?

PAGE 64

64 H0 : There is no interaction between the pe rceived ethnicity of the avatar, the perceived gender of the avatar, the ethnicity of the participant, and the gender of the participant. Independent Variable(s): Perceived ethnicity by Perceived gender by Participant ethnicity by Participant gender After the study was closed and data collect ed, a debrief email was sent to inform participants of the true purpose of the study and assured them of co nfidentiality (see Appendix H). Limitations Lim itations of this study include: 1. The length of time that the survey was avai lable may be a limitati on. Individuals who respond to Internet-based surveys tend to re spond much quicker than traditional paperbased surveys. As a result, the potential exis ts that the survey was biased toward early responders and left out late responders. Future studies should consider making the survey available for a longer period of time in orde r to obtain responses from all types of participants. 2. Issues of external validity exist due to the fact that the study was done in one Multi User Virtual Environment, Second Life. This may potentially prevent th e generalizability of this study to other virtual environments. Addi tionally, the surveys was only provided in an English format, thus non or limited English speaking individuals may be underrepresented in the sample population. Finally, threats to external validity may exist because the typical resident of Second Life is male, between 25-34 years old, and among the top 5% of the worlds wealthiest individuals. Future studies should consider replicating this experiment in multiple virtual environments, languages, and targeting a wider demographic. 3. The possibility for socially desirable responses by the participants exis ts as the nature of topical matter covered by the su rvey may lead respondents to sk ew responses to paint their attitudes toward issues of et hnicity and gender in a more fa vorable light. Future efforts should focus on strengthening the deception por tion of this study in order to lessen the probability of socially desirable responses. 4. The sample for the study was non-random in nature. This is due to the fact that the number of teachers available to participate in the study was only a little more than the sample needed. As a result, the study may not be gene ralizable to the larger population. This will

PAGE 65

65 no longer be a problem in the future as educa tional institutions incr ease their involvement in MUVEs and the population of teacher s in MUVEs increases significantly. 5. This was an experimental study in which teachers were given only limited information about hypothetical students. Natu ralistic studies (those taking place in real classrooms) in which teachers have access to large amounts of information and feedback may show less expectancy effect than expe rimental ones do. Future studi es should consider observing teacher-student interaction in a MUVE and comp are these results with the results of this study. 6. This was a deception study in which teachers were led to believe that they were attempting to evaluate different forms of transcripts used in MUVEs. As a result, some of the teachers may have guessed the researchers motives for this study from the content of the survey. This is important because many times study participants think about what the researcher wants them to do. Future efforts should focus on strengthening the deception portion of this study in order to lessen the probability of participants guessing the true purpose, as well as providing a debrief survey to determ ine the effectiveness of the deception. 7. Avatars were created by groups of Caucasia n, African American, Hispanic/Latino, and Asian college students. Results would probably change if the characteristics of the avatars were changed. Future studies should consider re plicating this study with different avatars. 8. The question asked in order to qualify a person to be a participant of this study could have been interpreted in many different ways. A re you teaching in Second Life? could have been interpreted by a person who uses SL as a tool, lecture platform, or a variety of different teaching methods. This study doe s not touch on the impact of the teaching method used on teacher perspectives and exp ectations. However, future studies should consider this as a poten tial control variable.

PAGE 66

66 Table 3-1. Overview of instruments/material s used and what they are used to measure Instrument/Materials used What does it measure?/How is it used? Presurvey Notification letter Notified potential part icipants that a survey was available to take and how much they would be paid for taking the survey. Informed consent form Used to obtain informed consent from participants Introductory Lette r (Clifford & Walster, 1973) Introduces teachers to the st udy and explains the studys purpose. Student Transcript (Clifford & Walster, 1973) Used as a deception to make te achers think that the studys purpose is to examine differe nt teacher transcripts and give comments. Opinion survey (Clifford & Walster, 1973) Measures teachers estimate of students IQ, relationships with classmates, relationships with instructors, attitude toward school, and prediction of students future education level Student image Information about the students sex and ethnicity (Clifford & Walster, 1973; Guttmann & Bartal, 1982) Used to measure teacher bias for or against specific ethnicities and genders Machinima video of student information about the students sex and ethnicity (Clifford & Walster, 1973; Guttmann & Bartal, 1982) Used to measure teacher bias for or against specific ethnicities and ge nders. A video was added to what the original study used in orde r to contextualize the study to the MUVE environment of Second Life. Demographic questionnaire (Pohan & Aguilar, Social Research Foundation) Used to measure potential covariates. List of demographics obtained from suggested list by Pohan & Aguilar (2001) and by attributes already collected by Social Research Foundation. Debrief email All participants received an email which debriefed them on the true purpose of the study and assured them of it being confidential.

PAGE 67

67 Table 3-2 List of variables Variable Name Variable Description Variable Type ParticipantID Participant identification # Numeric Session ID Session ID # Numeric Submit Time Time/Date the survey was submitted Date/Time AvatarGENDER Gender of the Av atar as designated by the Researcher Categorical AvatarETHNICITY Ethnicity of th e Avatar as designated by the Researcher Categorical InformedConsent Whether or not they agreed to Informed Consent Dichotomous StudentIQ Estimation of students IQ Numeric RELwithClassmates I would specul ate that the student's social relationships with classmates are ______. Likert (1-5) RELwithINSTRUCTORS I would sp eculate that the student's relationships with instructors are ________. Likert (1-5) ATTITUDEtowardschool I would guess that the student's attitude toward school is one of _______. Likert (1-5) EDUCATIONPrediction I would pred ict that the student would continue their education through _______. Likert (1-10) TRANSCRIPTComments Please provid e your comments and reactions to the transcript format. Be as specific as possible in your criticism Alphanumeric ParticipantGENDER What is your gender? Categorical ParticipantAGE What is your age? Categorical NumberofCHILDREN How many ch ildren do you have in your household? Categorical MARITALSTATUS What is your marital status? Categorical ParticipantEDUCATION What le vel of Education have you completed? Categorical ParticipantETHNICITY What is your race/ethnicity? Categorical ParticipantEthnicityOther Space to en ter other race/ethnicity Alphanumeric Participantorigin What is your continent of origin? Categorical ParticipantOriginOther Space to enter other origin Alphanumeric ParticipantRELIGION What is your religious affiliation? Categorical ParticipantReligionOther Space to enter other religion Alphanumeric ParticipantFAMILYINCOME What is your annual family Income? Categorical ParticipantINTERNETuse What is your Internet daily use? (NOT including time spent in Second Life) Categorical TimeInSLeachweek How much time do you spend in Second Life each week? Categorical SLResidencyLength How long ha ve you been a resident of Second Life? Categorical PrimarymotivationforComingt oSL What is your primary motivation or reason for coming to SL? Categorical

PAGE 68

68 Table 3-2. Continued Variable Name Variable Description Variable Type ParticipantsStudentsTaught You are a teacher of what kind of students? Categorical ParticipantsStudentsTaughtOth Er Space to enter other type of students Alphanumeric AvatarSalientCharacteristics1 Wh at are the three most salient characteristics of the student in this video and picture? (Please use 1 or 2 word description) part 1 Alphanumeric AvatarSalientCharacteristics2 Wh at are the three most salient characteristics of the student in this video and picture? (Please use 1 or 2 word description) part 2 Alphanumeric AvatarSalientCharacteristics3 Wh at are the three most salient characteristics of the student in this video and picture? (Please use 1 or 2 word description) part 3 Alphanumeric PerceivedAvatarGENDER What gender is th e student in the picture? Categorical PerceivedAvatarETHNICITY What race/ ethnicity is the student in the picture? Categorical

PAGE 69

69 CHAPTER 4 RESULTS Introduction As stated in Chapter 1, the study reported he re exam ined in detail the relationship of perceived student avatar gender and ethnicity to teacher expectations and perceptions. This chapter is organized in terms of a) participant demographics, b) preliminary analyses, and c) main analyses. The preliminary analyses section first checks for ceiling, floor, and non-normality issues for all dependent and independent variables. Se condly, it examines the pr oportion of participants who misidentified the gender a nd ethnicity of the avatars (a ccording to the researchers categories). The main analyses section is or ganized in terms of the ethnic ity related, gender related, and gender/ethnicity interaction relate d research questions posed in Chapter 1. It examines the impact of student avatar gender and ethnicity on teacher perceptions of: student IQ, relationships with classmates, relationships w ith instructors, attitudes towa rd school, and prediction of the level of education the st udent will complete. Participant Demographics The data for survey participants indicated that 489 surveys w ere attempted. Of those attempted, 455 were completed. Those who did not complete the survey appeared to cease taking the survey on the first few questions. A su mmary of the descriptive statistics can be found in Tables 4-1 to 4-14 which summarize basic dem ographic information related to the participants in the study. The participants were primarily teachers who taught either adult ( 37.8%) or K-12 students (29%; Table 4-1) and possessed an undergraduat e (26.6%) or graduate degree (30.1%; Table 4-

PAGE 70

70 2). Gender (Table 4-4), religious affiliation (Table 4-8) and age (Table 4-3) appeared to vary widely, although a majority of participants were under 30 years of age (51.5%; Table 4-3). They were mostly from Europe (34.9%) or North Amer ica (48.6%; Table 4-9), single (52.1%; Table 45), Caucasian/White (77.4%; Table 4-7), with no children (66%; Table 4-5), and possessing an annual income below 75,000 U.S. dollars (53%; Tabl e 4-10). The greater part of participants spent more than 3 hours daily on the Internet (60.8%; Table 4-11) and more than 6 hours in Second Life each week (63.2%; Table 4-12). They had also resided in SL for more than 12 months (56.9%; Table 4-13) for non-work relate d motivations (68.2%; Table 4-14). Descriptive Statistics The f irst step in the preliminary analyses was to run descriptive statistics for all independent and dependent variables checking for ceiling, floor, and non-normality issues. The mean and standard deviation for the depe ndent (Table 4-15) and independent (Table 417) variables reflected the inclusion of extreme outlier points in the data. These were corrected through the trimming of these data points as desc ribed below in the section entitled Discussion of Outliers for IQ variable. The mean and standa rd deviation for the dependent and independent variables with these deletions are displayed in Table 4-16 and Table 4-18. Discussion of Outliers for IQ variable Upon visual inspection, the IQ data contained several data points that initially appeared to be far outside of what would be a norm al IQ estimate. For example, several participants estimated IQ's below 10 or above 300. These da ta points became a problem when attempting to analyze the IQ data for statistical relationships As a result, 19 outli ers were deleted by the following method. A different outlier analysis was run for each equation used in order to determine which IQ scores to keep. The outlier analysis relied only on stan dardized residuals as

PAGE 71

71 the outlier detection strategy (C ohen, Cohen, West, and Aiken, 2003). Any standardized residual higher than 3 and less than -3 wa s considered an outlier and elim inated. This was an iterative process: the analysis was run, cas es were found that were outlie rs, these were eliminated, the analysis was rerun, more cases were found that were outliers, the analysis was rerun, etc., until there were no more outliers (i.e. all the data points had a standardized residual between 3 and -3). This took as many as eight iterations before th ere were no more outliers. The form of outlier detection run eliminated participants whose IQ estimations were much farther away from what would be expected relative to everyone else (Cohen, Cohen, We st, and Aiken, 2003). CrossTabs Table Analysis Due to the studys focus on avatar choice, it was important to iden tify the proportion of participan ts who misidentifie d the gender and ethnicity of the avatars (according to the researchers categories). The analysis of the gende r of the avatar as designated by the researcher against the perceived gender of the avatar by the participants showed that of the 248 total participants who were assigned a male avatar, 19 9 or 80.2% of those part icipants perceived the avatar as male. The analysis also showed that of the 241 total particip ants who were assigned a female avatar, 221 or 91.7% of those participants perceived the avatar as female (Table 4-19). Another analysis necessary to the study was to identify whether certain genders designated by the researcher are more likely to have their ethnicity misidentified by the participants. The ethnic percentages for male avatars vs. female av atars appeared to be very close, so avatar gender did not influence the ethnic identification process (Table 4-20). It was also critical to this research to identi fy whether certain avatar ethnicities designated by the researcher are more likely to have their gender misidentified by the participants. Avatars designated as Caucasian and Hispanic/Latino we re most likely to be perceived to correspond

PAGE 72

72 with the researcher designated gender category. However, the analysis shows that avatars designated as African American and Asian by the researcher were more likely to have misidentifications of the researcher designate d gender category. In both the African American and Asian researcher designated ethnicity categories, males were being perceived as females (Table 4-21). The analysis of the ethnicity of the avatar as designated by the researcher against the perceived ethnicity of the avatar revealed that: 1. Of the 129 total participants who were assigned an African-American avatar, 115 or 89.1% perceived the avatar as Afri can-American (Table 4-22). 2. Of the 116 who were assigned an Asian avatar, 78 or 67.2% perceived the avatar as Asian (Table 4-22). 3. Of the 125 who were assigned a Caucasian avat ar, 106 or 84.8% perceived the avatar as Caucasian (Table 4-22). 4. Of the 119 who were assigned a Hispanic/Latin o avatar, 77 or 64.7% pe rceived the avatar as Hispanic/Latino (Table 4-22). Participants who misidentified either the gender or ethnicity of the avatar according to the researchers categories were assigned to the group that they perceived it to be. For example, if they received an Asian male avat ar but perceived it as a Hispan ic/Latino female avatar, they were assigned to the Hispanic/Latino group. As a result of this crosstabs analysis, the sa mple for this study was not randomly assigned to the eight different ethnic cate gories and analyzed based on the re searcher categories. This was due to the fact that several participants indicated that the avatar they we re assigned to was of a different ethnicity or gender th an what the researcher original ly intended. See Table 4-23 for a detailed description of these differences. The Asian male and Hispanic/Latino female were the most difficult for participants to identify. As a result, because the purpose of this study was to

PAGE 73

73 determine the impact of student avatars gend er and ethnicity on teachers expectations and perceptions, it was logical to use the teachers perceptions to self -assign them to one of the eight categories. Analyses of Research Questions This section contains the result s of th e analyses of the resear ch questions. Results of these analyses were considered significant if p<.05. It is important to note that there were a low number of Black/African-America n, Hispanic/Latino, Asian, and Ot her participants. As such, there were also empty cells in the research desi gn because of the low num ber of participants in certain categories. As a result, the sample may be unrepresentative of those ethnic groups in SL because of the low number of part icipants in some categories, which could in turn cause the results to be skewed in the direction of those fe w participants responses, rather than those of a larger group (Miller, 1991). The overarching question in th is study was whether avatar choice affected teacher expectations and perceptions of student academic success. Many studies have been cited that have convincingly shown that teacher biases ba sed on the gender and ethnicity of students do exist in face-to-face clas srooms. However, this study was done in a MUVE. The results below are discussed in terms of the ethnicity related, gender related, and interactions between gender and ethnicity questions. Ethnicity Related Questions The m ain question addressed in this section of questions was Does ethnicity (of the student or teacher) influence te acher perceptions and expectations of student success? Each question is described an d addressed separately. Research question 1: Does perceived ethnicity of the avatar predict teacher perceptions and expectations?

PAGE 74

74 Research question 1 was analyzed using ANOV As to examine the relationships between teacher expectations and percepti ons and the perceived ethnicity of the avatar. No significance was found for any of these relations hips at p < .05 (Table 4-24). Research question 2: Does the ethnicity of the participant predict teacher perceptions and expectations? Research question 2 was analyzed using ANOV As to examine the relationships between teacher expectations and percep tions and the ethnicity of the participant. The relationship between participant ethnicity and relationships with instructors wa s significant at p < .05 (F(4) = 2.65, p = .03; Table 4-24 ). Other ethnicity teachers in SL estimated students (regardless of ethnicity or gender) as having th e highest levels of relationshi ps with instructors (M = 3.12), compared to Hispanic or Latino teachers in SL (M = 3.04), Caucasian/White teachers in SL (M =2.99), Asian teachers in SL (M =2.71), and African American/Black teachers in SL (M = 2.26; Table 4-25). Research question 3: Is there an interaction between th e ethnicity of the avatar and the ethnicity of the participant? Research question 3 was analyzed using ANOV As to examine the relationships between teacher expectations and perceptions and the inte raction between the ethnicity of the avatar and the ethnicity of the participant. No significan ce was found for any of these relationships at p < .05 (Table 4-24). Summary: Ethnicity Related Questions The ethnicity-related questions asked the ove rarching question, Does ethnicity (of the student or teacher) influence te acher perceptions and expectations of student success? The results of this study found that under certain ci rcumstances ethnicity does influence teacher perceptions and expectations of student success. For example, Other ethnicity teachers in SL

PAGE 75

75 estimated students (regardless of ethnicity or gender) as having the highest levels of relationships with instructors, compared to Hispanic or Latino, Caucasian/White, Asian, and African American/Black teachers in SL. Gender Related Questions The m ain question addressed in this section of questions wa s Does gender (of the student or teacher) influence teacher perceptions and exp ectations of student succ ess? Each question is described and addressed separately. Research question 4: Does perceived gender of the avatar predict teacher perceptions and expectations? Research question 4 was analyzed using ANOV As to examine the relationships between teacher expectations and perceptions and the ge nder of the student avatar. The relationship between the estimated attitude of the student toward school and the perceived gender of the student avatar was significant at p < .05 (F(1) = 7.15, p = .01; Table 4-24). Teachers in SL estimated male avatars as having higher levels of attitude toward school (M = 3.36) compared to female avatars (M = 3.15; Table 4-26). Research question 5: Does the gender of the particip ant predict teacher perceptions and expectations? Research question 5 was analyzed using ANOV As to examine the relationships between teacher expectations and perceptions and the gend er of the participant. No significance was found for any of these relationshi ps at p < .05 (Table 4-24). Research question 6: Is there an interaction between the gender of the avatar and the gender of the participant? Research question 6 was analyzed using ANOV As to examine the relationships between teacher expectations and perceptions and the inte raction between the gender of the avatar and the gender of the participant. The following re lationships were signi ficant (Table 4-24):

PAGE 76

76 A significant Participant gender by Avatar Gende r interaction emerged for attitude of the student toward school (F(1) = 5.846, p = .016; Table 4-24). o Male teachers in SL estimated male student av atars as having higher levels of attitude toward school (M = 3.54) compared to fema le student avatars (M = 3.00; Table 4-27). o Female teachers in SL estimated female student avatars as having higher levels of attitude toward school (M = 3.28) compared to male stude nt avatars (M = 3.23; Table 4-27). o The difference in attitude toward school between male and female avatars was larger for male teachers while the difference for female teachers was trivial. Summary: Gender Related Questions The gender-related questions asked the overarc hing question, Does gender (of the student or teacher) influence teacher perceptions and exp ectations of student succ ess? The results of this study found that under certain circumstances gender does influence teacher perceptions and expectations of student success. For example, male teachers in SL estimated male student avatars as having better attitudes toward school compared to female student avatars (Table 4-29), while the difference for female teachers was negligible. Interactions Between Gender and Ethnicity The m ain question addressed in this section of questions was Do any of the interactions between gender and ethnicity (of the student and/or teacher) in fluence teacher perceptions and expectations of student success? Each quest ion is described and addressed separately. Research question 7: Is there an interaction between the perceived ethnicity and gender of the avatar and teacher pe rceptions and expectations? Research question 7 was analyzed using ANOV As to examine the relationships between teacher expectations and percepti ons and the ethnicity and gender of the avatar. No significance was found for any of these relationships at p < .05 (Table 4-24).

PAGE 77

77 Research question 8: Is there an interaction between the perceived gender of the avatar and the ethnicity of the participant? Research question 8 was analyzed using ANOV As to examine the relationships between teacher expectations and perceptions and the inte raction between the gender of the avatar and the ethnicity of the participant. The fo llowing relationships were significant: A significant Perceived gender by Participant Et hnicity interaction emerged for Student IQ (F(3) = 2.67, p = .04; Table 4-24). o Caucasian/White (Mmales = 110.08; (Mfemales = 107.98), African American/Black (Mmales = 102.88; Mfemales = 102.50), and Other (Mmales = 115.00; Mfemales = 110.83) ethnicity teachers in SL estimated that ma le student avatars have a higher level of Student IQ compared to female st udent avatars (Table 4-28). o Hispanic or Latino (Mfemales = 117.50; (Mmales = 99.69) and Asian (Mfemales = 112.80; Mmales = 99.27) ethnicity teachers in SL estimated that female student avatars have a higher level of Student IQ compared to male student avatars (Table 4-28). A significant Perceived gender by Participant Et hnicity interaction emerged for attitude toward school (F(3) = 2. 81, p = .03; Table 4-24). o Caucasian/White (Mfemale = 3.28; Mmale = 3.25) and African American/Black (Mfemale = 3.25; Mmale = 3.00) and Other (Mfemale = 3.61; Mmale = 3.41) ethnicity teachers in SL estimated female student avatars as having a higher level of Attitude toward school compared to male stude nt avatars (Table 4-29). o Hispanic or Latino (Mmale = 3.68; Mfemale = 2.87) and Asian (Mmale = 3.28; Mfemale = 2.85) ethnicity teachers in SL estimated male student avatars as having a higher level of attitude toward school compared to female student avatars (Table 4-29). Research question 9: Is there an interaction between the ethnicity of the avatar and the gender of the participant? Research question 9 was analyzed using ANOV As to examine the relationships between teacher expectations and percep tions and the ethnicity of the avatar and the gender of the participant. No significance was found for any of these relationships at p < .05 (Table 4-24). Research question 10: Is there an interaction between the ethnicity of the participant and the gender of the participant?

PAGE 78

78 Research question 10 was analyzed using ANOV As to examine the relationships between teacher expectations and percepti ons and the ethnicity of the participant and the gender of the participant. No significance was found for any of these relationships at p < .05 (Table 4-24). Research question 11: Is there an interaction between the ethnicity of the avatar, the gender of the avatar, and the et hnicity of the participant? Research question 11 was analyzed using ANCO VAs to examine the relationships between teacher expectations and perceptions with the et hnicity of the avatar, the gender of the avatar, and the ethnicity of the participant. The following relationship was significant: A significant Perceived Avatar ethnicity by Perceived Av atar gender by participant ethnicity interaction emerged for Student IQ (F(8) = 2.89, p = .01; Table 4-24). o Caucasian/White teachers in SL estimated female, Caucasian/White student avatars (M = 110.50) as having higher levels of Student IQ compared to male, Caucasian/White student avatars (M = 103.02). Caucasian/White teachers in SL estimated male, Black/African-American (M = 117.16), Hisp anic/Latino (M = 109.90), and Asian (M = 110.23) ethnicity student avatars as having higher levels of Student IQ compared to female, Black/African-American (M = 109.73), Hispanic/Latino (M = 101.97), and Asian (M = 109.71) ethnicity student avatars (Table 4-30). o Hispanic or Latino teachers in SL esti mated female, Caucasian/White (M = 93), Black/African-American (M = 114.5), Hispan ic/Latino (M = 147.5) student avatars as having higher levels of St udent IQ compared to male, Caucasian/White (M = 101.018) and Black/African-American (M = 107.247) and Hispanic/Latino (M = 86.52) student avatars (Table 4-30). o Other ethnicity teachers in SL estimated female, Caucasian/White (M = 118.75) student avatars as having higher levels of Student IQ compared to male, Caucasian/White (M = 105) student avatars. Other ethnicity teachers in SL estimated male, Black/African-American (M = 132.50) and Hispanic/Latino (M = 111.25) student avatars as having higher levels of Student IQ compared to female, Black/African-American (M = 106.66) a nd Hispanic/Latino (M = 97.5) student avatars (Table 4-30). Research question 12: Is there an interaction between the ethnicity of the avatar, the gender of the avatar, and the gender of the participant?

PAGE 79

79 Research question 12 was analyzed using ANOV As to examine the relationships between teacher expectations and perceptions with the et hnicity of the avatar, the gender of the avatar, and the gender of the participant. No significance was found for a ny of these relationships at p < .05 (Table 4-24). Research question 13: Is there an interaction between the ethnicity of the avatar, the ethnicity of the participant, a nd the gender of the participant? Research question 13 was analyzed using ANOV As to examine the relationships between teacher expectations and perceptions with the ethnicity of the avatar, the ethnicity of the participant, and the gender of the participant. The followi ng relationship was significant: A significant Perceived Avatar ethnicity by Participant ge nder by participant ethnicity interaction emerged for Student IQ (F(8) = 2.13, p = .03; Table 4-24). o Male, Caucasian/White teachers in SL estimated Black/African American (M = 113.30) student avatars as having higher levels of Student IQ compared to Caucasian/White (M = 108.27) student avatar s, Asian (M = 106.83) student avatars, and Hispanic/Latino (M = 103.53) student avatars. Female, Caucasian/White teachers in SL estimated Black/African Am erican (M = 113.6) student avatars as having higher levels of Stude nt IQ compared to Asian (M = 113.11) student avatars, Hispanic/Latino (M = 108.34) student avatars, and Caucasian/White (M = 105.24) student avatars (Table 4-31). o Male, African American/Black teachers in SL estimated Black/African American (M = 115) student avatars as having higher levels of Student IQ comp ared to Asian (M = 111) student avatars. Female, African American/Black teachers in SL estimated Hispanic/Latino (M = 100) student avatar s as having higher levels of Student IQ compared to Asian (M = 94) student av atars and Caucasian (M = 93.66) student avatars (Table 4-31). o Male, Asian teachers in SL estimated Black/African American (M = 122.5) student avatars as having higher levels of Studen t IQ compared to Caucasian (M = 106) student avatars, Asian (M = 104.37) student avatars, and Hispanic/Latino (M = 98.25) student avatars. Female, Asian teachers in SL estimated Hispanic/Latino (M = 117.5) student avatars as having hi gher levels of Student IQ compared to Black/African American (M = 106.17) student avatars, Cau casian (M = 105.8) student avatars, and Asian (M = 99.33) student av atars (Table 4-31).

PAGE 80

80 o Male, Other ethnicity teachers in SL estim ated Caucasian (M = 107.5) student avatars as having higher levels of Student IQ co mpared to Hispanic/Latino (M = 106.25) and Black/African American (M = 100) student av atars. Female, Other ethnicity teachers in SL estimated Asian (M = 130) student av atars as having higher levels of Student IQ compared to Black/African American (M = 122.91) student avatars, Caucasian (M = 117.5) student avatars, and Hispanic/Lat ino (M = 102.5) student avatars (Table 431). Research question 14: Is there an interaction betw een the gender of the avatar, the ethnicity of the participant, a nd the gender of the participant? Research question 14 was analyzed using ANOV As to examine the relationships between teacher expectations and perceptions with the gender of the avatar, the ethnicity of the participant, and the gender of the participant. No signifi cance was found for any of these relationships at p < .05 (Table 4-24). Research question 15: Is there an interaction between the ethnicity of the avatar, the gender of the avatar, the ethnic ity of the participant, and th e gender of the participant? Research question 15 was analyzed using ANOV As to examine the relationships between teacher expectations and perceptions with the ethnicity of the avat ar, the gender of the avatar, the ethnicity of the participant, and the gender of th e participant. The follo wing relationships were significant: A significant Perceived Avatar gender by Perceived Avatar ethnicity by participant ethnicity by participant gende r interaction emerged for Student IQ (F(4) = 3.7, p = .01; Table 4-24). o Male, Other ethnicity teachers in SL es timated female, Hispanic/Latino (M = 110) student avatars as having higher levels of Student IQ compared to male, Hispanic/Latino student avat ars (M = 102.5; Table 4-32). o Female, Other ethnicity teachers in SL estimated male, Hispanic/Latino (M = 120) and Black/African-American (M = 132.5), stude nt avatars as having higher levels of Student IQ compared to female, Hispan ic/Latino (M = 85) and Black/AfricanAmerican (M = 113.33) student avatars. Fe male, Other ethnicity teachers in SL also estimated female Caucasian (M = 130) st udent avatars as having higher levels of Student IQ compared to male, Caucasian (M = 105) student avat ars (Table 4-32).

PAGE 81

81 o Male, Caucasian/White ethnicity teachers in SL estimated female, Caucasian (M = 113.61) student avatars as havi ng higher levels of Student IQ compared to male, Caucasian (M = 2.706) student avatars. Ma le, Caucasian/White ethnicity teachers in SL also estimated male, Black/African-A merican (M = 114.66), Hispanic/Latino (M = 109.85), and Asian (M = 108) student avatars as having higher levels of Student IQ compared to female, Black/African-Ame rican (M = 111.94), Hispanic/Latino (M = 97.21) and Asian (M = 105.66) studen t avatars (Table 4-32). o Female, Caucasian/White teachers in SL estimated female, Caucasian (M = 107.38) and Asian (M = 113.76) student avatars as having higher levels of Student IQ compared to male, Caucasian (M = 103.1) and Asian (M = 112.46) student avatars. Female, Caucasian/White teachers in SL also estimated male, Black/AfricanAmerican (M = 119.66) and Hispanic/Lati no (M = 109.95) student avatars as having higher levels of Student IQ compared to female, Black/African-American (M = 107.53) and Hispanic/Latino (M = 106.73) student avatars (Table 4-32). o Male, Hispanic/Latino teachers in SL es timated male, Caucasian (M = 104.6) and Black/African American (M = 120) student av atars as having higher levels of Student IQ compared to female, Caucasian (M = 86) and Black/African American (M = 115) student avatars (Table 4-32). o Female, Hispanic/Latino teachers in SL estimated female, Caucasian (M = 100), Black/African American (M = 115), and Hisp anic/Latino (M = 95) student avatars as having higher levels of student IQ compared to male, Caucasian (M = 80), Black/African American (M = 105), and Hispanic/Latino (M = 78.25) student avatars (Table 4-32). Summary: Interactions Between Gender and Ethnicity The questio ns dealing with the interacti ons between gender and ethnicity asked the overarching question, Do any of th e interactions between gender and ethnicity (of the student and/or teacher) influence teacher perceptions and expectations of student success? The results of this study found that under certain circumstances some of these interactions do influence teacher perceptions and expectati ons of student success. For example, Caucasian/White teachers in SL estimated female, Caucasian/White student avatars as having higher levels of Student IQ compared to male, Caucasian/White student avat ars. Caucasian/White teachers in SL also estimated male, Black/African-American, Hispanic/L atino, and Asian ethnicity student avatars as

PAGE 82

82 having higher levels of St udent IQ compared to female, Black/African-American, Hispanic/Latino, and Asian ethnicity student avatars (Table 4-34). Conclusions This chapter has attem pted to answer the ove rarching question of whether avatar choice influences teacher perceptions a nd expectations of student succe ss. In doing so, results have been given for the study of the influence of student avatar gender and ethnicity on teacher expectations and perceptions of student success. The results of this study found that under certain circumstances gender and ethnicity does in fluence teacher perceptions and expectations of student success. This study re ported on a) particip ant demographics, b) preliminary analyses, and c) main analyses. Chapter five focuse s on the discussion of th ese results and their implications.

PAGE 83

83 Table 4-1. Type of Students Taught Type of Students Frequency Percent K 12 Students 132 29.0 Colllege or University Students 93 20.4 Adult Students 172 37.8 Mixed age group 25 5.5 Subject Matter expert 25 5.5 Other 8 1.8 Total 455 100.0 Table 4-2. Education Completed Education Level Frequency Percent Cumulative Percent High School 48 10.5 10.5 Some college 127 27.9 38.5 Undergraduate Degree 121 26.6 65.1 Graduate degree (MA, MFA, Ph.D.) 137 30.1 95.2 Professional degree (CPA, etc.) 22 4.8 100.0 Total 455 100.0

PAGE 84

84 Table 4-3. Age Frequency Percent Valid Percent Cumulative Percent 18-21 81 17.8 19.2 19.2 21-25 64 14.1 15.2 34.4 26-30 72 15.8 17.1 51.5 31-35 46 10.1 10.9 62.5 36-40 55 12.1 13.1 75.5 41-50 59 13.0 14.0 89.5 51-60 36 7.9 8.6 98.1 60+ 8 1.8 1.9 100.0 Total 421 92.5 100.0 Missin G System 34 7.5 Total 455 100.0 Table 4-4. Gender Gender Frequency Percent Male 232 51.0 Female 223 49.0 Total 455 100.0 Table 4-5. Number of children Frequency Percent Cumulative Percent 0 303 66.6 66.6 1 63 13.8 80.4 2 62 13.6 94.1 3 25 5.5 99.6 4 or More 2 .4 100.0 Total 455 100.0

PAGE 85

85 Table 4-6. Marital status Frequency Percent Single 237 52.1 Married 148 32.5 Domestic Partners 32 7.0 Divorced 24 5.3 Widowed 2 .4 Other 12 2.6 Total 455 100.0 Table 4-7. Ethnicity Frequency Percent Caucasian/White 352 77.4 African American/Black 9 2.0 Hispanic or Latino 30 6.6 Asian 45 9.9 Other, please specify 19 4.2 Total 455 100.0 Table 4-8. Religious affiliation Frequency Percent Atheist 49 10.8 Agnostic 68 14.9 Protestant Christian 70 15.4 Roman Catholic 86 18.9 Evangelical Christian 22 4.8 Jewish 11 2.4 Muslim 5 1.1 Hindu 6 1.3 Buddhist 14 3.1 None 87 19.1

PAGE 86

86 Table 4-9. Continent of origin Frequency Percent Europe 159 34.9 Asia 36 7.9 North America 221 48.6 South America 21 4.6 Australia 14 3.1 Africa 2 .4 Other, please specify 2 .4 Total 455 100.0 Table 4-10. Annual family income Frequency Percent Cumulative Percent < $50k 124 27.3 27.3 $50 to 75k 117 25.7 53.0 $75 to 100k 102 22.4 75.4 $100 150k 63 13.8 89.2 $150 200k 27 5.9 95.2 $200k+ 22 4.8 100.0 Total 455 100.0 Table 4-11. Daily Internet use Frequency Percent Cumulative Percent < 1 hour 22 4.8 4.8 1 -2 hours 79 17.4 22.2 2 to 3 78 17.1 39.3 3 to 5 104 22.9 62.2 5 to 8 74 16.3 78.5 8 to 10 43 9.5 87.9 10 to 12 23 5.1 93.0 More than 12 32 7.0 100.0 Total 455 100.0

PAGE 87

87 Table 4-12. Time spent in Second Life each week Frequency Percent Cumulative Percent < 1 hour 18 4.0 4.0 1 to 2 hours 28 6.2 10.1 2 to 4 hours 55 12.1 22.2 4 to 6 hours 66 14.5 36.7 6 to 10 hours 68 14.9 51.6 11 to 15 Hours 52 11.4 63.1 16 to 20 Hours 59 13.0 76.0 21 to 30 Hours 51 11.2 87.3 31 to 40 Hours 22 4.8 92.1 More than 40 36 7.9 100.0 Total 455 100.0 Table 4-13. Residency length in Second Life Frequency Percent Cumulative Percent < 3 months 25 5.5 5.5 3 6 months 50 11.0 16.5 6 12 Months 121 26.6 43.1 12 24 Months 183 40.2 83.3 24 months+ 76 16.7 100.0 Total 455 100.0

PAGE 88

88 Table 4-14. Primary motivation for coming to SL Frequency Percent Fun 109 24.0 Work 32 7.0 Socializing 81 17.8 Playing 21 4.6 Education 42 9.2 Designing and Building 52 11.4 Experimenting/ Exploring 99 21.8 Other 19 4.2 Total 455 100.0 Table 4-15. Dependent Variab les Descriptive St atistics: without IQ Deletions N Mean Std. Deviation Estimation of Student's IQ 453 1738916.41 36537289.513 I would speculate that the student's social relationships with classmates are ___. 455 3.33 .736 I would speculate that the student's relationships with instructors are _____. 455 2.99 .792 I would guess that the student's attitude toward school is one of _____________. 455 3.25 .816 I would predict that the student would continue their education through _______. 455 4.83 1.650 Valid N (listwise) 453

PAGE 89

89 Table 4-16. Dependent Vari ables Descriptive Statistic s: with IQ Deletions N Mean Std. Deviation Estimation of Student's IQ 425 107.90 18.748 I would speculate that the student's social relationships with classmates are ____. 427 3.33 .722 I would speculate that the student's relationships with instructors are _____. 427 2.97 .781 I would guess that the student's attitude toward school is one of _____________. 427 3.26 .784 I would predict that the student would continue their education through _______. 427 4.79 1.608 Valid N (listwise) 425 Table 4-17. Independent vari ables Descriptive Statistics: without IQ Deletions N Mean Std. Deviation PerceivedAvatarGENDER 454 1.56 .498 PerceivedAvatarETHNICITY 454 2.29 1.095 What is your gender? 455 1.49 .500 What is your race/ethnicity? 455 1.62 1.211 Valid N (listwise) 454

PAGE 90

90 Table 4-18. Independent variables Descri ptive Statistics: with IQ Deletions N Mean Std. Deviation PerceivedAvatarGENDER 426 1.55 .498 PerceivedAvatarETHNICITY 426 2.29 1.100 What is your gender? 427 1.49 .500 What is your race/ethnicity? 427 1.62 1.222 Valid N (listwise) 426

PAGE 91

91 Table 4-19. Gender of the Avatar as Designated by the Researcher by Perceived Gender of the Avatar Crosstabs 17 221 3 241 7.1% 91.7% 1.2% 100.0% 48.6% 87.7% 1.5% 49.3% 3.5% 45.2% .6% 49.3% 18 31 199 248 7.3% 12.5% 80.2% 100.0% 51.4% 12.3% 98.5% 50.7% 3.7% 6.3% 40.7% 50.7% 35 252 202 489 7.2% 51.5% 41.3% 100.0% 100.0% 100.0% 100.0% 100.0% 7.2% 51.5% 41.3% 100.0% Count % within Gender of th e avatar as designated by the researcher % within What gender is the student in the picture and video? % of Total Count % within Gender of th e avatar as designated by the researcher % within What gender is the student in the picture and video? % of Total Count % within Gender of th e avatar as designated by the researcher % within What gender is the student in the picture and video? % of Total Female Male Gender of the avatar as designated by the researcher Total FEMALE MALE What gender is the student in the picture and video? Total

PAGE 92

92 Table 4-20. Gender of the Avatar as Designated by the Researcher by Perceived Ethnicity of the Avatar Crosstabs 17 49 68 67 40 241 7.1% 20.3% 28.2% 27.8% 16.6% 100.0% 48.6% 56.3% 50.7% 48.6% 42.1% 49.3% 3.5% 10.0% 13.9% 13.7% 8.2% 49.3% 18 38 66 71 55 248 7.3% 15.3% 26.6% 28.6% 22.2% 100.0% 51.4% 43.7% 49.3% 51.4% 57.9% 50.7% 3.7% 7.8% 13.5% 14.5% 11.2% 50.7% 35 87 134 138 95 489 7.2% 17.8% 27.4% 28.2% 19.4% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 7.2% 17.8% 27.4% 28.2% 19.4% 100.0% Count % within Gender of the avatar as designated by the researcher % within What race/ethnicity is the student in the picture and video? % of Total Count % within Gender of the avatar as designated by the researcher % within What race/ethnicity is the student in the picture and video? % of Total Count % within Gender of the avatar as designated by the researcher % within What race/ethnicity is the student in the picture and video? % of Total Female Male Gender of the avatar as designated by the researcher Total Asian Black/ AfricanAmerican Caucasian Hispanic/ Latino What race/ethnicity is the student in the picture and video? Total

PAGE 93

93 Table 4-21. Ethnicity of the Av atar as Designated by the Researcher by Perceived Gender of the Avatar Crosstabs 13 68 48 129 10.1% 52.7% 37.2% 100.0% 37.1% 27.0% 23.8% 26.4% 2.7% 13.9% 9.8% 26.4% 4 70 42 116 3.4% 60.3% 36.2% 100.0% 11.4% 27.8% 20.8% 23.7% .8% 14.3% 8.6% 23.7% 13 57 55 125 10.4% 45.6% 44.0% 100.0% 37.1% 22.6% 27.2% 25.6% 2.7% 11.7% 11.2% 25.6% 5 57 57 119 4.2% 47.9% 47.9% 100.0% 14.3% 22.6% 28.2% 24.3% 1.0% 11.7% 11.7% 24.3% 35 252 202 489 7.2% 51.5% 41.3% 100.0% 100.0% 100.0% 100.0% 100.0% 7.2% 51.5% 41.3% 100.0% Count % within Ethnicity of the avatar as designated by the researcher % within What gender is the student in the picture and video? % of Total Count % within Ethnicity of the avatar as designated by the researcher % within What gender is the student in the picture and video? % of Total Count % within Ethnicity of the avatar as designated by the researcher % within What gender is the student in the picture and video? % of Total Count % within Ethnicity of the avatar as designated by the researcher % within What gender is the student in the picture and video? % of Total Count % within Ethnicity of the avatar as designated by the researcher % within What gender is the student in the picture and video? % of Total African American Asian Caucasian Hispanic/Latino Ethnicity of the avatar as designated by the researcher Total FEMALE MALE What gender is th e student in the picture and video? Total

PAGE 94

94 Table 4-22. Ethnicity of the Avat ar as Designated by the Researcher Perceived Ethnicity of the Avatar Crosstabs 13 1 115 0 0 129 10.1% .8% 89.1% .0% .0% 100.0% 37.1% 1.1% 85.8% .0% .0% 26.4% 2.7% .2% 23.5% .0% .0% 26.4% 4 78 3 16 15 116 3.4% 67.2% 2.6% 13.8% 12.9% 100.0% 11.4% 89.7% 2.2% 11.6% 15.8% 23.7% .8% 16.0% .6% 3.3% 3.1% 23.7% 13 3 0 106 3 125 10.4% 2.4% .0% 84.8% 2.4% 100.0% 37.1% 3.4% .0% 76.8% 3.2% 25.6% 2.7% .6% .0% 21.7% .6% 25.6% 5 5 16 16 77 119 4.2% 4.2% 13.4% 13.4% 64.7% 100.0% 14.3% 5.7% 11.9% 11.6% 81.1% 24.3% 1.0% 1.0% 3.3% 3.3% 15.7% 24.3% 35 87 134 138 95 489 7.2% 17.8% 27.4% 28.2% 19.4% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 7.2% 17.8% 27.4% 28.2% 19.4% 100.0% Count % within Ethnicity of the avatar as designated by the researcher % within What race/ethnicity is the student in the picture and video? % of Total Count % within Ethnicity of the avatar as designated by the researcher % within What race/ethnicity is the student in the picture and video? % of Total Count % within Ethnicity of the avatar as designated by the researcher % within What race/ethnicity is the student in the picture and video? % of Total Count % within Ethnicity of the avatar as designated by the researcher % within What race/ethnicity is the student in the picture and video? % of Total Count % within Ethnicity of the avatar as designated by the researcher % within What race/ethnicity is the student in the picture and video? % of Total African American Asian Caucasian Hispanic/Latino Ethnicity of the avatar as designated by the researcher Total Asian Black/ AfricanAmerican Caucasian Hispanic/ Latino What race/ethnicity is the stude nt in the picture and video? Total

PAGE 95

95 Table 4-23. Random Assignment Versus Perc eived Assignment Category Differences Assigned Categories Perceived Categories 56 Asian females 55 Asian females 57 Asian males 32 Asian males 56 Black/African American females 84 Black/African American females 60 Black/African American males 50 Black/African American males 56 Caucasian female 70 Caucasian females 56 Caucasian male 70 Caucasian males 56 Hispanic Latino female 45 Hispanic Latino females 56 Hispanic Latino male 50 Hispanic Latino males

PAGE 96

96 Table 4-24. Continuous Data Set Hypotheses Results Hypotheses IQ Estimate Social Relationships with Classmates Relationships with Instructors Attitude toward School Education level Prediction F P F P F p F P F P Hypothesis #1: PerceivedAvatarEthnicity 1.67 .17 .72 .54 1.44 .23 1.09 .35 .41 .73 Hypothesis #2: PerceivedAvatarGender 2.51 .113 .01 .89 .08 .77 7.15 .01* .07 .78 Hypothesis #3: ParticipantGender .58 .44 1.15 .28 1.91 .16 .038 .53 .45 .50 Hypothesis #4: ParticipantEthnicity .31 .86 1.18 .31 2.65 .03* 1.67 .15 .26 .89 Hypothesis #5: PerceivedAvatarGender PerceivedAvatarEthnicity 1.65 .17 .67 .56 .10 .95 1.63 .18 .49 .68 Hypothesis #6: PerceivedAvatarGender ParticipantGender .54 .46 2.65 .10 .20 65 5. 38 .02* .50 .47 Hypothesis #7: PerceivedAvatarEthnicity ParticipantEthnicity .77 .65 1.58 .10 .74 .68 .46 .91 .60 .81 Hypothesis #8: PerceivedAvatarGender ParticipantEthnicity 2.67 .04* .63 .59 .14 .93 2.81 .03* 1.03 .37 Hypothesis #9: PerceivedAvatarEthnicity ParticipantGender 1.18 .31 .76 .51 .66 57 2.15 .09 1.33 .26 Hypothesis #10: ParticipantGender ParticipantEthnicity 2.16 .07 .74 .56 .98 .41 .38 .82 .80 .52 Hypothesis #11: PerceivedAvatarGender PerceivedAvatarEthnicity ParticipantEthnicity 2.89 .004 .50 .85 .95 .46 1.24 .27 .66 .72 Hypothesis #12: PerceivedAvatarGender PerceivedAvatarEthnicity ParticipantGender .27 .84 .31 .81 .75 .51 .29 .82 .76 .51 Hypothesis #13: PerceivedAvatarEthnicity ParticipantGender ParticipantEthnicity 2.13 .03* .56 .80 1.61 .12 .79 .60 .62 .75 Hypothesis #14: PerceivedAvatarGender ParticipantGender ParticipantEthnicity .65 .58 1.73 .15 .43 73 1.79 .14 1.00 .38 Hypothesis #15: PerceivedAvatarGender PerceivedAvatarEthnicity ParticipantGender ParticipantEthnicity 3. 70 .006* .28 .89 .95 .43 .61 .65 1.04 .38

PAGE 97

97 Table 4-25. Participant Ethnicity with Students Relationships with Instructors Estimated Marginal Means ________ 2.999 .046 2.907 3.090 2.267a .307 1.664 2.870 3.043a .171 2.706 3.380 2.714a .134 2.451 2.977 3.121a .197 2.734 3.509 What is your race/ethnicity? Caucasian/White African American/Black Hispanic or Latino Asian Other, please specify Mean Std. Error Lower Bound Upper Bound 95% Confidence Interval Based on modified population marginal mean. a. Table 4-26. Perceived Gender with Attitude toward School Estimated Marginal Means PerceivedAvatarGEND ER Mean Std. Error 95% Confidence Interval Lower Bound Upper Bound Male 3.361(a) .095 3.175 3.548 Female 3.149(a) .099 2.955 3.343 a Based on modified population marginal mean. Table 4-27. Perceived Gender Participant Ge nder Interaction for Attitude toward School Estimated Marginal Means ___ 3.542a .150 3.247 3.837 3.008a .146 2.721 3.295 3.233a .122 2.993 3.474 3.281a .134 3.018 3.544 PerceivedAvatar GENDER Male Female Male Female What is your gender? Male Female Mean Std. Error Lower Bound Upper Bound 95% Confidence Interval Based on modified population marginal mean. a.

PAGE 98

98 Table 4-28. Perceived Gender Participant Ethni city Interaction for Student IQ Estimated Marginal Means pQ 110.080 1.615 106.904 113.257 107.982 1.377 105.273 110.690 102.889a 9.131 84.933 120.845 102.500a 10.981 80.905 124.095 99.693a 5.302 89.267 110.119 117.500 5.715 106.262 128.738 99.274a 4.375 90.670 107.877 112.806 4.282 104.385 121.228 115.000a 7.088 101.061 128.939 110.833a 5.916 99.199 122.468 PerceivedAvatar GENDER Male Female Male Female Male Female Male Female Male Female What is your race/ethnicity? Caucasian/White African American/Black Hispanic or Latino Asian Other, please specify Mean Std. Error Lower Bound Upper Bound 95% Confidence Interval Based on modified population marginal mean. a. Table 4-29. Perceived Gender Participant Ethni city Interaction for Attitude toward School Estimated Marginal Means ___ 3.255 .070 3.117 3.394 3.284 .060 3.166 3.402 3.000a .398 2.217 3.783 3.250a .479 2.309 4.191 3.681a .231 3.226 4.135 2.875 .249 2.385 3.365 3.286a .191 2.911 3.661 2.850 .187 2.483 3.217 3.417a .299 2.830 4.004 3.619a .258 3.112 4.126 PerceivedAvatar GENDER Male Female Male Female Male Female Male Female Male Female What is your race/ethnicity? Caucasian/White African American/Black Hispanic or Latino Asian Other, please specify Mean Std. Error Lower Bound Upper Bound 95% Confidence Interval Based on modified population marginal mean. a.

PAGE 99

99 Table 4-30. Participant Ethnici ty Perceived Gender Percei ved Ethnicity Interaction for Student IQ Estimated Marginal Means Dependent Variable: Estimation of Student's IQ 103.019 2.556 97.993 108.045 110.500 2.487 105.610 115.390 117.167 3.245 110.786 123.547 109.737 2.246 105.321 114.154 109.905 3.094 103.821 115.988 101.972 3.097 95.882 108.062 110.231 3.888 102.585 117.877 109.718 3.088 103.646 115.790 93.667a 10.353 73.307 114.027 .b 115.000a 17.932 79.736 150.264 .b 100.000a 17.932 64.736 135.264 .b .b 102.500 10.981 80.905 124.095 92.300 9.822 72.985 111.615 93.000 12.680 68.064 117.936 112.500 10.981 90.905 134.095 114.500 12.680 89.564 139.436 86.625 10.024 66.912 106.338 147.500 10.981 125.905 169.095 115.000a 10.353 94.640 135.360 115.000 8.966 97.368 132.632 108.750 8.966 91.118 126.382 103.050 7.502 88.298 117.802 103.750a 8.966 86.118 121.382 115.550 7.502 100.798 130.302 92.500 8.966 74.868 110.132 123.250 10.981 101.655 144.845 94.333 7.321 79.937 108.730 109.375 7.765 94.105 124.645 105.000a 12.680 80.064 129.936 118.750 10.981 97.155 140.345 132.500a 12.680 107.564 157.436 106.667 10.353 86.307 127.027 111.250 10.981 89.655 132.845 97.500 10.981 75.905 119.095 .b 130.000a 17.932 94.736 165.264 PerceivedAvatar GENDER Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female PerceivedAvatar ETHNICITY Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian What is your race/ethnicity? Caucasian/White African American/Black Hispanic or Latino Asian Other, please specify Mean Std. Error Lower Bound Upper Bound 95% Confidence Interval Based on modified population marginal mean. a. This level combination of factors is not observed, thus the corresponding population marginal mean is not estimable. b.

PAGE 100

100 Table 4-31. Participant Ethnicity Participant Gender Percei ved Ethnicity Interaction for Student IQ Estimated Marginal Means pQ 108.276 2.367 103.621 112.932 113.304 2.488 108.410 118.198 103.534 3.158 97.323 109.744 106.833 3.505 99.941 113.725 105.242 2.667 99.998 110.487 113.600 3.063 107.577 119.623 108.343 3.031 102.382 114.304 113.115 3.517 106.199 120.031 .a 115.000b 17.932 79.736 150.264 .a 111.000b 17.932 75.736 146.264 93.667b 10.353 73.307 114.027 .a 100.000b 17.932 64.736 135.264 94.000b 12.680 69.064 118.936 95.300 9.822 75.985 114.615 117.500 12.680 92.564 142.436 147.500 12.680 122.564 172.436 115.000b 12.680 90.064 139.936 90.000 12.680 65.064 114.936 109.500 10.981 87.905 131.095 86.625 7.765 71.355 101.895 115.000 8.185 98.904 131.096 106.000 8.966 88.368 123.632 122.500b 12.680 97.564 147.436 98.250 8.966 80.618 115.882 104.375 6.848 90.908 117.842 105.800 7.502 91.048 120.552 106.175 6.015 94.347 118.003 117.500 10.981 95.905 139.095 99.333 8.185 83.237 115.429 107.500b 12.680 82.564 132.436 100.000b 17.932 64.736 135.264 106.250 10.981 84.655 127.845 .a 117.500 10.981 95.905 139.095 122.917 8.185 106.821 139.013 102.500 10.981 80.905 124.095 130.000b 17.932 94.736 165.264 PerceivedAvatar ETHNICITY Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian What is your gender? Male Female Male Female Male Female Male Female Male Female What is your race/ethnicity? Caucasian/White African American/Black Hispanic or Latino Asian Other, please specify Mean Std. Error Lower Bound Upper Bound 95% Confidence Interval This level combination of factors is not observed, thus the corresponding population marginal mean is not estimable. a. Based on modified population marginal mean. b

PAGE 101

101 Table 4-32. Participant Ethnicity Participant Gender Perceived Gender Percei ved Ethnicity Interactio n for Student IQ Est imated Marginal Means pQ 102.938 3.170 96.704 109.171 113.615 3.517 106.699 120.531 114.667 3.913 106.971 122.362 111.941 3.075 105.893 117.989 109.857 4.793 100.432 119.282 97.211 4.114 89.120 105.301 108.000 5.977 96.245 119.755 105.667 3.660 98.468 112.865 103.100 4.010 95.215 110.985 107.385 3.517 100.469 114.301 119.667 5.177 109.487 129.847 107.533 3.274 101.095 113.972 109.952 3.913 102.257 117.648 106.733 4.630 97.628 115.839 112.462 4.973 102.681 122.242 113.769 4.973 103.989 123.550 .a .a 115.000 17.932 79.736 150.264 .a .a .a .a 111.000 17.932 75.736 146.264 93.667 10.353 73.307 114.027 .a .a .a 100.000 17.932 64.736 135.264 .a .a 94.000 12.680 69.064 118.936 104.600 8.020 88.829 120.371 86.000 17.932 50.736 121.264 120.000 17.932 84.736 155.264 115.000 17.932 79.736 150.264 95.000 17.932 59.736 130.264 200.000 17.932 164.736 235.264 .a 115.000 12.680 90.064 139.936 80.000 17.932 44.736 115.264 100.000 17.932 64.736 135.264 105.000 12.680 80.064 129.936 114.000 17.932 78.736 149.264 78.250 8.966 60.618 95.882 95.000 12.680 70.064 119.936 115.000 10.353 94.640 135.360 115.000 12.680 90.064 139.936 102.500 12.680 77.564 127.436 109.500 12.680 84.564 134.436 .a 122.500 12.680 97.564 147.436 95.000 12.680 70.064 119.936 101.500 12.680 76.564 126.436 98.000 10.353 77.640 118.360 110.750 8.966 93.118 128.382 115.000 12.680 90.064 139.936 96.600 8.020 80.829 112.371 103.750 8.966 86.118 121.382 108.600 8.020 92.829 124.371 90.000 12.680 65.064 114.936 145.000 17.932 109.736 180.264 90.667 10.353 70.307 111.027 108.000 12.680 83.064 132.936 .a 107.500 12.680 82.564 132.436 .a 100.000 17.932 64.736 135.264 PerceivedAvatar GENDER Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female PerceivedAvatar ETHNICITY Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican Hispanic/Latino Asian Caucasian Black/AfricanAmerican What is your gender? Male Female Male Female Male Female Male Female Male What is your race/ethnicity? Caucasian/White African American/Black Hispanic or Latino Asian Other, please specify Mean Std. Error Lower Bound Upper Bound 95% Confidence Interval

PAGE 102

102 CHAPTER 5 IMPLICATIONS This chapter presents a summ ary of the study and conclusions drawn from the data presented in chapter four. It provides a discussion of th e implications for action and recommendations for further research. Social interaction is importa nt to learning (Vygotsky, 1978) and can occur in technology mediated forms (e.g., 5th dimension project Cole, 1995; Nicolopolou & Cole, 1993; CSILE Scardamalia & Bereiter, 1996; Scardamalia, et al., 1994; LaFerriere, 2002). One form of technology mediated interaction th at is increasingly being used in education is the Multi-User Virtual Environment (MUVE; New Media C onsortium, 2007; Castronova, 2001). These environments allow users to experience a graphically engaging world while interacting with others as an avatar (Lastowka & Hunter, 2004). Users can choose the physical appearance of their avatar by customizing minute details (Damer, 1998; Rehak, 2003; Lastowka & Hunter, 2004). Unfortunately, we have an inadequate understa nding about the effect of avatar choice on teacher expectations and percep tions of student success. The gender and ethnicity of students has been shown to specifically effect teacher perc eptions and expectations of students in face-toface classrooms (Clifford & Walster, 1973; Gu ttmann & Bar-Tal, 1982; Walther & Tidwell, 1996; Ferguson, 2003; Frawley, 2005; Van Du zer, 2006; Auwarter & Aruguete, 2008). Therefore, we need to study the influence of av atar choice (i.e. gender and ethnicity) on teacher perceptions and expectations of student success. The impact of the gender and ethnicity of the students avatar on the teachers expectations and perceptions begins in the classroom. T eachers expectations and perceptions of their students do change with regard to their students gender and ethnicity in a face-to-face classroom

PAGE 103

103 environment (Clifford & Walster, 1973; Gu ttmann & Bar-Tal, 1982; Ferguson, 2003; Frawley, 2005; Van Duzer, 2006; Auwarter & Aruguete, 2008). However, traditional Internet-based learning environments, such as learning manage ment systems, conceal the gender and ethnic identity of students by giving them anonymity (Wa lther & Tidwell, 1996). The lack of visual and audio cues in these environm ents acts to remove gender and et hnic identity to a large extent (Walther & Tidwell, 1996). However, MUVE s allow for the customization of gender and ethnicity in student avatars (L astowka & Hunter, 2004). With over 180 universities present in Second Life, it is easy to recognize that a la rge number of student s are entering these environments in search of an education. It is also easy to surmise that these students represent a wide variety of different ethniciti es and genders. The critical que stion then is: Are they faced with the prejudiced environment experienced by st udents of different ethn icity and gender in the face-to-face classroom (Clifford & Walster, 1973; Guttmann & Bar-Tal, 1982; Ferguson, 2003; Frawley, 2005; Van Duzer, 2006; Auwarter & Aruguete, 2008), by the environment presented by traditional Internet learning (Walther & Tidwell, 1996), or by another still to be determined environment? Research Questions and Discussion The overarching question in th is stu dy was whether avatar choice affected teacher expectations and perceptions of student academic success. Many references have been given that have convincingly shown that teacher biases ba sed on the gender and ethnicity of students do exist in face-to-face cl assrooms. However, this study was done in a MUVE. The sample selected mirrors the overall population of Sec ond Life, which is mostly Caucasian and from western nations. It should be noted that the Internet has been primarily adopted by Caucasian

PAGE 104

104 populations in the U.S. (77% Cau casian; Lenhart et al, 2002), so SL appears to be following the same pattern. The sample selected is also repr esentative of teachers in SL (Mallon, 2008). We will consider ethnicity related, gender related, and interactions between gender and ethnicity results in that order. Ethnicity Related Questions Researchers have shown that the ethn icity of students impacts teacher expectations of students in the face-to-face classrooms (Clifford & Walster, 1973; Guttmann & Bar-Tal, 1982; Ferguson, 2003; Auwarter & Aruguete, 2008). Studies have also shown that the ethnicity of the teacher is applicable to teachers' perceptions and expectations of academic achievement in classroom situations (Braun, 1976; Brophy & Good, 1974; Ferguson, 2003; Auwarter & Aruguete, 2008). Because most avatars in MUVEs have the capability to graphically display differences in ethnicity (Damer, 1998; Rehak, 2003; Lastowka & Hunter, 2004), it was deemed important to determine the relationship between the ethnicity of the student avatar and the teacher with teacher perceptions and expectations in a Multi User Virtual Environment. There was one significant result for the ethnicity-related questions: 1. Teachers in SL with an ethnic ity of Other estimated stude nts across all ethnicities and genders as having the highest levels of relationships with inst ructors, followed by Hispanic or Latino SL teachers, Caucasian/White SL teachers, Asian SL teachers and African American/Black SL teachers (Table 4-25). These results agree with current literature fr om face-to-face classrooms. Stereotypical bias based on teacher ethnicity is pervasive in th e educational system (Campbell, 1991; Fennema & Peterson, 1985; LaFrance, 1981; Masland, 1994). Edu cation provides, at l east in theory, equal opportunities to all individuals. In practice, however, differential expectations about students are typically present. Many physical characteristics are capable of evoking initial expectations and

PAGE 105

105 impressions, including teacher ethnicity (Braun, 1976; Brophy, 1983; Brophy & Good, 1974; Dusek, 1985; Finn, 1972; Ferguson, 2003; Auwarter & Aruguete, 2008). Teachers' ethnicities have been shown to relate to their perceptions of children's behavior, independent of the student's ethnicity (B eady & Hansell, 1981; Pigott & Cowen, 2000; Zimmerman, Khoury, Vega, Gil, & Warheit, 1995) In these studies teacher ethnicity was strongly related to college expectations. N on-Caucasian teachers had significantly greater expectations that their student s would successfully en ter and complete college than Caucasian teachers (Beady & Hansell, 1981). Also, NonCaucasian teachers rated all children more positively, as having more competencies and fewer problems than teachers of other ethnicities (Pigott & Cowen, 2000). As a result, this positive attit ude of Non-Caucasian teachers may have expressed itself in the Other ethnicity teacher s estimations of higher relationships with instructors. Results also showed that teacher ethnicit y did not influence their perceptions and expectations of Student IQ, relationships with classmates, attit ude toward school and prediction of future final education level. Thus, the ques tion is why were relationships with instructors affected by the ethnicity of the teacher and not these other areas? It may be because relationships with instructors are the only dependent variable that personally involves the instructor in their estimation. The attributes of each party are related to the quality of the relationship that develops between teachers a nd students (Kesner, 2000). For example, ethnic variation between teachers and students may in fluence how each percei ves the other. The teacher brings his or her own cultural values, be liefs and practices to the relationship (Bowlby, 1982). Thus, teacher ethnicity may be an issue if th e cultural norms of each party are in conflict.

PAGE 106

106 This is important because it may suggest that f actors that personally invol ve the instructor in their estimation are more subject to discrimination by teachers in MUVEs. Should all students seek out instructors from an Other ethnicity in order to put themselves in position for a better relationship with their instructor? These questions touch on the problem of unfair advantage mentioned earlier. But is it really unfair if a student uses the knowledge they have to give them the best opportunity to succe ed? It is if one belie ves that opportunity to succeed should be solely based on an objective standard of academic achievement. However, the literature cited in this study has clearly shown that teachers do not follow an objective standard in their perceptions and evaluations of students. Is it then unrealistic to attempt to hold students to a higher standard of academic su ccess than teachers themselves? The answer needs to address both sides. T eachers and students in MUVEs need to work together to develop new ways to measure student success that are not artifi cially biased. The IQ scale was created by Caucasians within one cultural context (American; Suzuki & Aronson, 2005), but the MUVE is clearly a completely different context. As test creators, teachers and students address this problem, there are many opportunities to think differently in examining what constitutes an intelligence measure and how to examine issues of bias (White, 2000). For example, given growing concerns regarding the us age of intelligence tests for selection purposes, Jensen (2000) suggested using cr iteria that go beyond standardized measures and the inclusion of indicators of past performance (i.e., portfolio, learning artifacts, etc.). Future studies should consider the use of different ki nds of measures of in telligence that may be more conducive to use in a MUVE in order to address this issue. Three possible reasons emerge for these conflicts with the literature. First, this may be a new finding that has to do with the unique proper ties and affordances that a MUVE has to offer.

PAGE 107

107 Research has found that MUVEs have unique stre ngths in simulation (Dieterle & Clark, in press), personal expression and creation (Perkins, 2007), real wo rld, collaborative learning, and role-playing (Rappa et al, 2008), and identity experimentation (Turkle, 1995). Thus, future research is needed to confirm these results a nd, if confirmed, determine the underlying factors in a MUVE that may have created this conflict with the literature. Second, the teachers in this study may be quite different than the teachers discussed in the above studies. For example, most of the above studies examined K-12 teachers, while our sample was more evenly split between adult and K-12 teachers (Table 5-1). Also, most of the cited literature surveyed teachers that were mo stly female, while our sample was evenly split between male and female representation. Finally, the above studies were either a) Run during a time period when the Internet eith er did not exist or was not as popular, or b) Did not measure teacher Internet usage. The combination of these may have been enough to explain these different results for teachers in SL. Third, the artificial experimental setting us ed for the experiment may have enabled teachers in SL to remove themselves from the setting (the classroom) of their discriminatory behavior. Elashoff and Snow ( 1971) suggest that teacher expect ations are more likely to be affected in their natural classroom environmen t. Strong attempts were made to make the experimental conditions a close simulation of natural MUVE condi tions. Teachers received an image and 3-D movie of a student avatar, as well as an actual student tran script. However, the study was not conducted in an actual MUVE cla ssroom, which could have separated teachers from the setting of their discriminatory behavior. This is important to consider because of the potential confounding effect of th e artificial experimental setting.

PAGE 108

108 Ethnicity biases inherent in real world teachers were not mirrored in these results. Future research should consider replicating this experiment for three reasons. First, because these results conflicted with a long hi story of literature on ethnicity bi ases, we need to replicate the experiment to determine if we get similar resu lts from a similar sample. Second, we need to replicate the experiment because the SL teachers were from a significantly different demographic than those in the literature. If we get the same results with a similar real life sample demographically, then future research will be need ed. This research should attempt to determine differences between real life and MUVE studies and the und erlying factors behind these differences (i.e. identity exploration in MUVEs, expe rimental setting, etc.). Finally, the experiment should be replicated in a more naturalistic setting (a classroom) in order to determine if the experimental setting led to these results. Th is research would be important in order to determine if these results are the result of the experimental methodology that was used, the teacher demographics, or if they are due to other reasons. Gender Related Questions Research has shown that the gender of teachers and students impacts teacher expectations of students in the face-to -face classrooms (Braun, 1976; Brophy & Good, 1974; Clifford & Walster, 1973; Guttmann & Bar-Tal, 1982; Fr awley, 2005; Van Duzer, 2006). Because most avatars in MUVEs have the capability to graphically display differences in gender (Damer, 1998; Rehak, 2003; Lastowka & Hunter, 2004), it was deemed important to determine the relationship between the gender of the student avatar and teacher perceptions and expectations in a Multi User Virtual Environment. There were three main significant results for the gender-related questions: 1. Teachers in Second Life estimate male student avatars as having a higher level of attitude toward school compared to female student avatars (Table 4-26).

PAGE 109

109 2. Male teachers in SL estimated male student avatars as having a higher attitude toward school compared to female student avatars (M = 3.27; Table 4-27). 3. Female teachers in SL estimated female student avatars as having a higher attitude toward school compared to male student av atars (Table 4-29; Table 4-27). It is important to note that the differences in attitude toward sc hool between male and female avatars were larger for male teachers while the difference for female teachers was negligible. These results are in conflict with current literatur e from face-to-face classrooms. Males appear to view school as less important in their lives as females (Clark et al., 2006). In a national survey, male students consistently reported that they did not listen in class, complete assignments, and did not do their best work (NCES, 2005). Females viewed their academic work as interesting and meaningful compared to males who did not appear to see the critical nature of their academic performance as it relates to future occupation (NCES, 2005). It is important to confirm these results because if teachers in MU VEs (especially males) estimate males as having higher attitudes toward school then this goes agai nst current data on acad emic performance that clearly shows males falling behind females (Burns and Bracey 2001; Clark et al. 2006; Kafer 2004). This result could indicate evidence of male teacher discrimination against female students in MUVEs, which shoul d be addressed through teacher education and professional development opportunities. Non-significant results for th e relationship between avatar and/or participant gender and Student IQ and future final education level may be explained by numer ous studies that have frequently failed to find signifi cant differences between male and female academic development and IQ (McNemar, 1942; Havighurst and Janke, 1944; Hughes, 1953; Jorm et al, 2004; Colom et al., 2000; Aluja-Fabregat et al., 2000). Thus, it ma y be that teachers in SL viewed both Student

PAGE 110

110 IQ and future final educational le vel as measures of intelligence that had no difference for males and females. Non-significant results for the relations hip between avatar gender and students relationships with classmates a nd instructors seem to conflict with current research from the face-to-face classroom. Research has shown that males in a classroom environment tend to be more aggressive and take more initiative in the classroom than females (Tuddenham, 1952; Spach, 1951; Sears, 1961; Feshbach, 1956; Sanf ord, Adkins, Miller, and Cobb, 1943; Digman, 1963), while females are seen as more sensitive to the feelings of others, more dependent, more motivated, more passive, and quieter than male s (Broverman et al., 1972; Feldman & Kiesler, 1974; Richmond-Abbott, 1979; Rim & Aloni, 1969; Ruble & Ruble, 1980). This literature would appear to indicate that teachers would pr eference females as having better relationships with students and instructors than males. Howeve r, results from this study did not agree. This disagreement between the litera ture and this study is important because knowing whether teachers preference students of a particular gend er as having better relationships with classmates and instructors could be an indicator of teacher bi as in MUVEs toward students of their gender. As for the student, the lesson to learn appears to be unfortunately all too clear: Be aware of your teachers gender and customize your avatar to the gender that is most advantageous to your success. Although this might come across as insensitive, results ha ve shown that teachers in MUVEs may discriminate on the basis of ge nder. However, MUVEs allow for detailed customization of avatar physical characteris tics, including gender (Las towka & Hunter, 2004). Students who are being discriminate d against in an MUVE have pow er that a student in a real life classroom does not. They can change their gender as they move from class to class and

PAGE 111

111 instructor to instructor. Therefore, the a ffordance of detailed customization in a MUVE empowers students to level th e academic playing field. However, it may be argued that it is unjust to require students to cha nge their gender as they move from instructor to instructor. Is this is the same thing as past real life efforts by public education to force female students into histor ically accepted feminine roles like homemaker and mother? During that time period it was argued that the mere availability of education empowered women by giving them th e choice to be educated (Beech er, 1842). We are repeating the mistakes of history by encouraging students to embrace changing their avatar gender for their own academic advantage. The solution is to focu s on the source of the apparent discrimination in this case the male teacher in SL. What methods have been used successfully to eliminate discrimination and increase acceptance of students from different genders in the face-to-face classroom? What programs have been created that have successfully help ed teachers to avoid judgment of students based on gender? Research seems to suggest th at analysis and argument is likely to challenge and change attitudes (Petty and Cacioppo, 1986). Observations of others actions may change normative beliefs and enha nce self-efficacy through social modeling and persuasion (Bandura, 1998), and, in doing so, is likely to prompt protective intention formation (Ajzen and Madden, 1986). Behavior-specific c ognitions and practice are most likely to be effective (Wight et al ., 1998). Four possible reasons emerge to explain these results. First, this new result may have to do with some interaction between teachers and the strengths that MUVEs allow. MUVEs allow teachers and students enhanced learning through simulations (Dieterle & Clark, in press), personal expression and creation (Perkins, 2007), collaborative le arning, and role-playing (Rappa et al, 2008) and identity experiment ation (Turkle, 1995). Therefore, future research is needed to

PAGE 112

112 confirm these results and, if confirmed, determ ine the underlying factors in a MUVE that may have created this conflict with the literature. Second, because this results conflicts with a la rge body of literature, it is important to ask whether there was something in the data, experime ntal setting, or analys is methods that might explain this difference. The teachers in this study (in SL) may be quite different than the teachers discussed in the above studies. For ex ample, most of the above studies examined K-12 teachers, while our sample was more evenly sp lit between adult and K-12 teachers (Table 5-1). Also, most of the cited literature surveyed teachers that were mostly female, while our sample was evenly split between male and female representation. Finally, the above studies were either a) Run during a time period when the Internet either di d not exist or was not as popular, or b) Did not measure teacher Internet usage. The combination of these may have been enough to explain these conf licting results. Third, the artificial experimental setting us ed for the experiment may have enabled teachers in SL to remove themselves from the setting (the classroom) of their discriminatory behavior. Elashoff and Snow ( 1971) suggest that teacher expect ations are more likely to be affected in their natural classroom environmen t. Strong attempts were made to make the experimental conditions a close simulation of natural MUVE condi tions. Teachers received an image and 3-D movie of a student avatar, as well as an actual student tran script. However, the study was not conducted in an actual MUVE cla ssroom, which could have separated teachers from the setting of their discriminatory behavior. This is important to consider because of the potential confounding effect of th e artificial experimental setting. Finally, the experimental methods may need to be examined. It may be that the deception used was not effective resulting in teachers having knowledge of the true purpose of the study. If

PAGE 113

113 this occurred, it would be likely that participants may choose to purposely bias their answers as they believe the researcher would want them to. Future research should consider replicating this experiment for three reasons. First, because these results conflicted with a long hist ory of literature on the subject, we need to replicate the experiment to determine if we get similar results from a similar sample. Second, we need to replicate the experiment because the SL teachers were from a significantly different demographic than those in the literature. If we get the same results with a similar real life sample demographically, then future research will be needed. This research should attempt to determine differences between real life and MUVE studies and the underlying factors behind these differences (i.e. identity explorati on in MUVEs, experimental setting, etc.). Finally, we should replicate the experiment in order to explore which aspects of males attitude toward school are appea ling to teachers in MUVE envi ronments. This is important because it may help us to develop effective inte rventions for teachers in MUVEs so they may be able to interact with students of every gender without prejudice. Further research is needed to determine if in terventions that have been successful with teachers in the face-to-face classr oom will also be successful in the MUVE classroom. Future research should also explore the effect of cha nging avatar gender on teachers expectations and perceptions of attitude toward school and relations hips with classmates. Research should also consider the effect of androgynous avatars on teacher expectations and perceptions. This is important to consider because th is study found that male teachers expectations and perceptions were biased toward student avatars of their ow n gender. Considering the effect of androgynous avatars on teacher expectations and perceptions ma y reveal a category that is free from gender bias or may open up completely new categories that may be subject to discrimination.

PAGE 114

114 Interactions Between Gender and Ethnicity Teachers in SL did not appear to b e impacted by the interaction between the ethnicity and gender of student avatars in their estimates of : Student IQ, relations hips with classmates, relationships with instructors and prediction of fu ture final education leve l. This is surprising due to the significant results th at were individually found for the gender and ethnicity of the avatar. It also conflicts with the literature from face-to-face classrooms that state that gender and ethnic equity has not yet been achieved in the face-to-face classroom (Campbell, 1991; Fennema & Peterson, 1985; LaFrance, 1981; Ma sland, 1994; Sadker & Sadker, 1994). In particular, the gender and ethnicity of students has been shown to impact teacher expectati ons of students in the face-to-face classrooms (Clifford & Walster, 1973; Guttmann & Bar-Tal, 1982; Ferguson, 2003; Frawley, 2005; Van Duzer, 2006; Auwarter & Aruguete, 2008). There were two main significant results for the interaction questions: 1. Caucasian/White and Other ethnicity teachers in SL estimated that male student avatars have a higher level of Student IQ compared to female student avatars (Table 4-28). 2. Hispanic or Latino and Asian ethnicity (and Non-Caucasian) teachers in SL estimated that female student avatars have a higher level of Student IQ compared to male student avatars (Table 4-28). 3. Caucasian/White teachers in SL estimated fe male, Caucasian/White student avatars as having higher levels of Student IQ compared to male, Caucasian/White student avatars. Caucasian/White teachers in SL estimated male, Black/African-American, Hispanic/Latino, and Asian ethnic ity student avatars as having higher levels of Student IQ compared to female, Black/African-American, Hispanic/Latino, and Asian ethnicity student avatars (Table 4-29). 4. Hispanic or Latino teachers in SL estima ted female, Caucasian/White, Black/AfricanAmerican, Hispanic/Latino student avatars as having higher levels of Student IQ compared to male, Caucasian/White, Black/African-Ameri can and Hispanic/Latino student avatars (Table 4-29). 5. Other ethnicity teachers in SL estimated fe male, Caucasian/White student avatars as having higher levels of Student IQ compared to male, Caucasian/White student avatars.

PAGE 115

115 Other ethnicity teachers in SL estimated male, Black/African-American and Hispanic/Latino student avatars as having higher levels of Student IQ compared to female, Black/African-American and Hispanic/Lati no student avatars (Table 4-29). Differences in IQ ratings were only negligible for African American/Black teachers (males = 102.88; females = 102.50; Table 4-30). The larges t differences in IQ ratings were from Hispanic/Latinos and Asians while African Am erican/Black teachers ra ted male and female avatars virtually the same. It appears that As ian and Hispanic/Latino t eachers have the largest differences in ratings of IQ. Also, Asian and Hispanic/Latino teachers who rated females over males in IQ had the largest effects. White teach ers gave higher IQ ratings to males of other ethnicities but not of their own ethnicity. Hisp anic/Latino teachers ap peared to be the only ethnicity of teacher that rated avatars the same across the board. Interestingly, Other ethnicity teachers agreed with Ca ucasian teachers. A large body of literature exists that has shown that the gender and ethnicity of the teacher does affect their perceptions and expectations of academic achievement in classroom situations (Braun, 1976; Brophy & Good, 1974; Ferguson, 2003; Auwarter & Arug uete, 2008; Frawley, 2005; Van Duzer, 2006). From this literature it w ould have been expected that results would have revealed teacher bias in SL against stude nts based on interactions between their own gender and ethnicity and the student avatars gender and ethnicity. Combining the results above, it appears that Cauc asian teachers in SL rate females of their own ethnicity as having a higher IQ than males, while they rate males of other ethnicities higher than females of those ethnicities. This is important because current data on academic performance clearly shows males falling behind females (Burns and Bracey 2001; Clark et al. 2006; Kafer 2004). Additionally, m ales appear to view school as less important in their lives as females (Clark et al., 2006, 2008). In a national survey, male student s consistently reported that

PAGE 116

116 they did not listen in class, complete assignmen ts, and did not do their best work (NCES, 2005). Females viewed their academic work as interesting and meaningful compared to males who did not appear to see the critical nature of their academic performance as it relates to future occupation (NCES, 2005). From this literature it would have been expected that Caucasian teachers in SL would have rated all females as havi ng a higher IQ than males. It is important to consider the reasons why these differe nces with the literature exist. The calculations involved in the ethnic choices of Caucasians seem to be different from those of other ethnic groups, si nce resources targeted for minor ity populations are generally not available to Caucasians (Nagel, 1994). In some situations, Caucasia n ethnicity can express itself as a reaction against perceived advantages of non-Caucasians (Bur stein 1991). In other situations, Caucasian ethnicity represents a pers onal choice exercised for social, emotional, or spiritual reasons (Waters 1990; Fischer 1986). Usi ng the criteria listed in the literature above, it makes sense to conclude that these Caucasian teac hers rated students in th is way due to one of the reasons listed above. If e ither is true, then these Caucas ian teachers in SL are allowing gender and ethnicity biases to cl oud their evaluation of students ac ademic success. Teachers in SL should receive training to he lp them to identify their personal motivations toward students academic success and to help them to see potential discriminatory effects of these motivations. Other results showed that Hispanic/Latino teach ers in SL rated female student avatars of every ethnicity as having a higher IQ than male student avatars of those ethnicities. Both attitudes and behaviors with respect to one's own ethnicity and others are conceptualized as changing as one develops and resolves issues and feelings about one's own and other groups (Nagel, 2000). These results may possibly be explained by Asian a nd Hispanic or Latino

PAGE 117

117 stereotypes that cast females from these cult ures as responsible, decisive, and ambitious (Hofstede, 1980, 1996). Non-significant results showed that the majority of these gender and ethnicity interactions did not appear to impact teachers in SL es timates of: Relationships with classmates, relationships with instructors, attitude toward school, and pred iction of future final education level. What caused teachers in SL to discrimina te on the basis of their own ethnicity and gender and the ethnicity and gender of the student avatar for estimates of intelligence (Student IQ) while seemingly being free from discrimination when it cam e to the other dependent variables? This is important because we may be able to focu s our attention toward the development of interventions that focus on teachers perspec tives of differing ethnicity and gender student intelligence. These results also conflict with literature from the face-to-face classroom which states that gender and ethnicity biases in these interactions should be present (Ferguson, 2003). These conflicting results are most likely due to a) a problem with the experi mental method or setting, b) a difference in the samples demographics vs. real world teacher demographics, or c) true conflicting results. We need to replicate this study in order to determine if this result is true. If such a replication has similar findings, then stud ies into the underlying fact ors that caused these differences would be in order. Of particular interest would be the exploration of why teacher biases in SL seem to have an effect on estimates of intelligence. This is important because future study may reveal the factors that underlie gender and ethnicity in a MUVE and answer questions as to why some biases seem to mirror results from face-to-face classroom s while others do not. As MUVE use becomes more and more frequent in education, business, and society in general, the impact of gender and ethnicity in these dive rse environments will need to be taken into

PAGE 118

118 account in order to create fair, impartial educational, work, and life settings in which we can interact and live. Future research should fo cus on the development and testing of impartial educational settings in MUVEs that benef it students of every gender and ethnicity. Summary: Direct Implications Much has been revealed by this study of th e im pact of student avatars gender and ethnicity on teachers expectati ons and perceptions. Our overarching question coming into this study was whether avatar choice in SL affected teacher perceptions and ex pectations of student success. The answer was yes, in a few contexts. Overall, only 8 out of a possible 75 results showed that gender or ethnicity or an interaction had an infl uence on teacher perceptions and expectations of student success. Some teacher s showed the real world gender and ethnicity biases talked about in the litera ture when they rated students on IQ (4 significant results), attitude toward school (3 significant resu lts), and relationships with inst ructors (1 significant result). However, no teachers showed the biases descri bed in the literature when rating students on relationships with classmates and prediction of future education level, and many did not show biases on the other measures. Als o, only 1 result was significant for ethnicity as a main effect, 2 results were significant for gender as a main effect and 5 results were significant for interactions between gender and ethnicity. Several reasons fo r this lack of agreem ent with the literature exist: 1. First, these may be new findings that have to do with the unique properties and affordances that a MUVE has to offer. Research ha s found that MUVEs have unique strengths in simulation (Dieterle & Clark, in press), personal expression and crea tion (Perkins, 2007), real world, collaborative learning, and role-playing (Rappa et al, 2008), and identity experimentation (Turkle, 1995). Thus, future re search is needed to confirm these results and, if confirmed, determine the underlying factor s in a MUVE that may have created this conflict with the literature.

PAGE 119

119 2. Because some of these results conflict with a large body of literature, it is important to ask whether there was something in the data, experi mental setting, or an alysis methods that might explain this difference. Teachers in SL may be quite different than the teachers discussed in the above studies. For example, most of the above studies examined K-12 teachers, while our sample was more evenly split between adult a nd K-12 teachers (Table 5-1). Also, most of the cited literature surv eyed teachers that were mostly female, while our sample was evenly split between male and female representation. Finally, the above studies were either a) Run during a time period when the Inte rnet either did not exist or was not as popular, or b) Did not measure teacher Internet usage. The combination of these may have been enough to e xplain these conflicting results. 3. The artificial experimental setting used for th e experiment may have enabled teachers to remove themselves from the setting (the clas sroom) of their discriminatory behavior. Elashoff and Snow (1971) suggest that teacher expectations are more likely to be affected in their natural classroom environment. Strong attempts were made to make the experimental conditions a close simulati on of natural MUVE conditions. Teachers received an image and 3-D movie of a stude nt avatar, as well as an actual student transcript. However, the st udy was not conducted in an ac tual MUVE classroom, which could have separated teachers from the setting of their discriminatory behavior. This is important to consider because of the pot ential confounding effect of the artificial experimental setting. 4. The experimental methods may need to be ex amined. It may be that the deception used was not effective resulting in te achers having knowledge of th e true purpose of the study. If this occurred, it would be likely that pa rticipants may choose to purposely bias their answers as they believe the researcher would want them to. 5. Some of these estimated marginal means may be a problem, or perhaps suggests a bias on the part of the participants (Miller, 1991). In trad itional research desi gn, you use a sample big enough to detect the smallest worthwhile effect (Miller, 1991). However due to the demographics of Second Life, this analysis had low numbers of participants in some categories (i.e. only 9 African Americans; Table 4-7). As a result, some significant effects may not have been detected. Additionally, a bias may exist due to these low numbers, which would in turn cause the results to be skewed in the direction of those few participants responses, rather than those of a larger group. Future research should consider replicating this experiment in or der to address these differences. First, because these results conflicted with a long hi story of literature on the subject, we need to replicate the experiment to determine if we get similar results from a similar sample.

PAGE 120

120 Second, we should replicate the experiment in or der to explore the specific aspects of student gender and ethnicity that may underlie teacher bias in MUVE environments. This is important because it may help us to develop effective inte rventions for teachers in MUVEs so they may be able to interact with students of every gender without prejudice. Finally, we need to replicate the study with a more diverse population in order to en sure that these results were not due to an unrepresentative sample for some of the categories. Results of this study revealed that student gender and et hnicity on the whole does not influence teacher expectations and perceptions in a MUVE. As a result, the problems of unintentional disadvantage and unf air advantage may exist to a much lower degree in a MUVE than originally expected. If this is correct, then the academic playing field in MUVEs may be more level than that of the face-to-face cla ssroom (Auwarter & Aruguete, 2008; Rist, 1970). It could actually be free of many of the gender and et hnicity biases currently present in face-to-face classrooms. Thus, future research should explor e the reasons why this bias doesnt appear to exist in MUVEs. However, this study showed that in a few cases, teachers do show gender and ethnicity biases in their perceptions and expectations of students in MUVEs. Unfortunately, this means that SL teacher biases in thei r perceptions of students coul d lead to some problems with unintentional disadvantage and unfair advantage. For example, male and female teachers in SL estimated that their own gender st udent avatars have a higher level of attitude toward school as compared to student avatars of the other ge nder. Students who didnt know this information could end up choosing a male avatar when ta ught by a female teacher, which could result in unintentional disadvantage to this student. However, a st udent who did know this information could end up choosing a male avatar when taught by a male teacher, resulting in unfair

PAGE 121

121 advantage. Both possibilities are troubling becau se the academic playing field in MUVEs is not level it is biased toward student avatars of particular genders and ethnicities. Furthermore, teacher bias in their perceptions of students often results in lower academic performance by students (Auwarter & Aruguete, 2008; Rist, 1970). Thus, this is more than just a concern about a few students who may or may not have knowledge of the results of th is study. The results of this study concern all teachers and students in MUVEs b ecause all students can be affected in their academic performance by teacher expectations and perceptions. Because of these problems, support programs need to be developed that help teachers in SL acquire teaching and interacti on techniques to address imbalances in attention and instruction that students of different genders and ethnicities receive. For ex ample, African American/Black teachers in SL estimated the lowest levels of stude nts relationships with instructors regardless of avatar gender or ethnicit y. Whether due to personal experi ences with racism or some other factor, African American/Black teachers in SL te nded to see students as being either unwilling or unable to build relationships with their instructors. These resu lts can be used as a springboard for development of a professional development program designed to help Black/African American teachers in SL to change this pers pective. When considering development of a program that addresses these kind of biases, it is important to consider research that seems to suggest that analysis and argument is likely to challenge and change attitudes (Petty and Cacioppo, 1986), social modeling and role playin g may change normative beliefs and enhance self-efficacy (Bandura, 1998; Ajzen and Madden, 1986), and practice in a safe, yet authentic environment is most likely to be effective in changing biases (Wight et al ., 1998). Interestingly, the pedagogical strengths of MUVE s lie in simulation, role play a nd practice, so the development of a program to help these te achers that is located in a MUVE would provide an intriguing

PAGE 122

122 solution. Programs like these would directly ad dress teacher biases in MUVEs and indirectly address the problems of unintenti onal disadvantage and unfair a dvantage mentioned earlier. Future research should consider determining the factors that may explain why some teachers in MUVEs hold these perceptions about student s relationships with instructors. Based on the results of this study, it is also important to determine which aspects of students intelligence, attitude, and the way they relate to clas smates and instructors affect particular groups of teachers in MUVEs. For ex ample, male, Asian teachers in SL estimate that Black/African American student avatars have the highest level of Student IQ, followed by: Asian student avatars, Caucasian student avatars, Asian student avatars, and Hispanic or Latino student avatars. If we knew that male, Asian teachers in SL considered particular stereotypical behavior or appearance that is typically attributed to Af rican Americans as markers of intelligence, we could address that in a training pr ogram that could help teachers to evaluate students more fairly. This is important because teachers without biases can interact with students of every gender and ethnicity without prejudice. Richardson (1996) sugge sted that personal beliefs of teachers such as gender and ethnicity biases are well esta blished and are primarily shaped by personal experience. As a result, we should not simply e xpect that an increase in multicultural knowledge would necessarily enhance the development of culturally diverse educators. Instead, it is important that we consider the development of in terventions that focus on the acquisition of a set of personal experiences that will help to shape teacher expectati ons and perceptions of students of every gender and ethnicity in MUVEs. Future research should consider a variety of approaches to measuring MUVE teacher perceptions and expectations of Student IQ, relationships with classmates, relationships with instructors, attitude toward sc hool, and prediction of future fi nal education level in MUVEs.

PAGE 123

123 Achieving triangulation in research is an impor tant goal in our progre ss toward validating the results of a study, as well as in deepeni ng our understanding. A lthough much thought and planning went into the selection of these measures and variables, it may be that other measures, variables and settings may more accurately ga uge these aspects of teacher perceptions and expectations in MUVEs. Future studies should vary the measures used and mix qualitative and quantitative methodologies in order to ach ieve a well-rounded set of results. Teachers and students in MUVEs need to work together to develop new ways to measure student success that are not artificially biased. The IQ scal e was created by Caucasians within one cultural context (American; Suzuki & Aronson, 2005). A MUVE is clearly a different context than the one in which the IQ scale was created. As test creators, teachers and students address this problem, there are many opportunitie s to think differently in examining what constitutes an intelligence measure and how to examine issues of bias (White, 2000). For example, given growing concerns regarding the us age of intelligence tests for selection purposes, Jensen (2000) suggested using cr iteria that go beyond standardized measures and the inclusion of indicators of past performance (i.e., portfolio, learning artifacts, etc.). Future studies should consider the use of different kinds of measures of intelligence that may work better in MUVEs in order to address this issue. More broadly, the results of this study have pr ovided evidence that choice of avatar does affect some aspects of teachers expectations and perceptions. We turn to a more indirect discussion of these ramifications next. Indirect Implications: Detailed Customization The results of this study have confirm ed that the gender and ethnicity of student avatars do have an impact on some aspects of SL teachers expectations and perceptions, including student

PAGE 124

124 IQ, relationships with classmates, and relationships with instructors, and attitude toward school. Taking a step back, gender and ethnicity are only two aspects of a wide range of physical details that can be customized within most MUVEs. For example, in Second Life, an avatar can be changed on a 0 to 100 scale for shape (body, head, eyes, ears, nose, mouth, chin, torso, legs, etc.), skin (skin color, face detail, makeup, body detail etc.), hair (color, style, eyebrows, facial), and eyes (color, texture, etc.; Damer, 1998; Rehak, 2003; Lastowka & Hunter, 2004). The question then becomes: Does this research pr esent sufficient evidence to support an agenda of exploring the impact of detailed customization of avatar physical characteristics on teachers expectations and perceptions? This study does present sufficient evidence to su pport this research ag enda. First, many of the details of avatar customization are strongly re lated to gender and ethnicity stereotypes, so an attempt to extract them from the stereotypical pe rspective in which they are viewed would be extremely difficult. At the same time, these st rong connections provide a logical link for the continuation and broadening of this research into the overall physical characteristics of avatars and their impact on teachers ex pectations and perceptions. It may be possible to say that stereotypical gender and ethnic attributes are what actually have an impact on teacher expectations and perc eptions in MUVEs. Thus, one could potentially state that these physical characteristics of avatar s were what actually had the effect on teacher expectations and perceptions. For example, one of our results states th at regardless of their ethnicity or gender, teachers in SL estimate ma le student avatars as ha ving a higher level of attitude toward school than female student avatars. Following this line of logic, we could then make the proposition that stereotypically masculine attributes such as large physical size, a high level of musculature, short hair length, casual dress, a low level of make-up, high weight level,

PAGE 125

125 rugged attractiveness, tall height, and hairy face (Hegelson, 1994; Brigham, 1971; Devine & Elliot, 1995) were what effected teacher expectati ons and perceptions. It would then be possible to test each of these stereotypical traits to determine which one had an impact on teachers expectations and perceptions in MUVEs. It woul d also be expected that the impact of these characteristics would vary according to the type of measure that teachers are estimating about students (relational, attitudina l, intelligence, etc.). Thus, there is a wealth of future research that could be accomplished to determine the effect of specific stereotypical gender and ethnic avatar physical characteristics on specific areas of teachers expectations and perceptions in MUVEs. This research is important because it will help to a) map out the factors involved in teacher s perspective on gender a nd ethnicity traits in MUVEs, b) enable us to test potential gender and ethnicity medi ation effects of MUVEs on these specific stereotypical physical characteristics, and c) make possible the development of specific interventions that are focused on the specific stereo typical physical characteristics that seem to trigger gender and ethnicity bi as in teachers in MUVEs. Ambient Implications Cumulative folder method Som e broader based implications of this study also need to be discussed. First, the study was patterned after a face-to-face classroom study (Clifford & Walster, 1973). It used a common approach used in many education-based research st udies that manipulate ethni city and gender. It chose head shots of student s and put them in a file that also contained th e school record of the student. The reasoning is that t eachers commonly use a student's image and academic transcript to form impressions. The file is used to observe teachers' expectations and perceptions about students of different genders and ethnicities. The academic description is held constant while the

PAGE 126

126 attached photo is varied to por tray a different gender/ethnicity combination. Patzer (1985) felt that the cumulative folder technique was an acceptable method for studying ethnicity and gender differences because, when teachers review st udents, they commonly receive an academic description and a photo in the stude nt's record. The cumulative folder procedure, however, is not without limitations. Relying on photographs fo r evaluating gender and ethnicity may be problematic because photographs provide a static cue for basing attributions and evaluations. Providing only a static cue oversimplifies the co nceptualization of gender and ethnicity as onedimensional constructs. Argyle and McHenry (1971) showed that photographs and brief exposure time do not simulate a real world situation. This study used a variation on the cumulative fo lder design. A transc ript was substituted for the academic record, and an image of an avatar for the photograph. However, in order to provide a less static representati on of the avatar, a short video of the avatar was also included. The video showed the same background for each avat ar, gave a view of the avatar from afar, and then slowly zoomed in to a close face shot of the avatar. The Fraps (version 2.9.4) screen recording software was used to record the videos. This variation on the cumulative folder desi gn appeared to work well. Results were collected quickly with little negative feedback from participants. Re sults showed significant effects of the gender and ethnicity of student avatars on teachers perceptions and expectations. However, as with any experiment that utilizes an artificial setting, (s omething other than the classroom), the possibility exists that the experime ntal setting used may ha ve enabled teachers to remove themselves from the setting (the classroom) of their discriminatory behavior. Giving credence to this argument, Elashoff and Snow ( 1971) suggest that teacher expectations are more likely to be affected in their na tural classroom environment. With that said, strong attempts were

PAGE 127

127 made to make the experimental conditions a clos e simulation of natural MUVE conditions. Both the image and the 3-D movie accurately portrayed MUVE conditions, and teachers received an actual student transcript. Future research shoul d pursue studies that test these results yet are conducted in a natural MUVE classroom setting. This is important because it will assure us that the results of this study are due to actual discriminatory behavior on the part of teachers and not the result of the study format. Demographics Implications Other broader based im plications of this study touch on the demographics of the SL teachers who were surveyed. For example, onl y 29% of the respondents were K-12 teachers, 20.4% were college or university instructors and 37.8% were instru ctors of adult students (Table 4-1). In this case, the large percentage of teacher s of adult instructors may have skewed the data. It may also be that the large number of inst ructors who were under 25 years of age (34.4%) as well as only possessing a high school education (10.5%, Table 4-2) or an undergraduate degree (26.6%, Table 4-2) may reflect a less well trained body of instructors when it comes to gender and ethnicity biases. Other demographic trends re vealed that 66% of the teachers surveyed had zero children and 52.1% were single. This indicates a highly mob ile group that would correlate with the relative youth of the sample. Perhaps one of the most provoking aspects of th e demographics was the data on ethnicity. 77.4 percent of the 455 participants were Caucasian/White, 2 percent were African American/Black, 6.6 percent were Hispanic or Latino, 9.9 percent were Asian, and 4 percent considered themselves in an Other category. Data on religious affiliation confirms the ethnicity demographics, revealing that 39.1% of the teachers were of some type of Christian affiliation (Table 4-8), a common affiliation among Caucasian/ Whites. Other demographic trends of note

PAGE 128

128 are that 83.5% of the teachers su rveyed were from either North America or Europe (Table 4-9), areas in which Caucasian/White ethnicity populati ons are in the majority These demographics are provoking because of their slant toward a Western, Caucasian, Christian population. We may have received very different results if th e population had been more evenly divided among ethnicities, geographic areas, and religious affiliation. Interestingly, 56.9% of the SL teachers surveyed had resided more than one year in Second Life (Table 4-13), yet only 9.2% had a prim ary motivation for coming to Second Life of education (Table 4-14). This leaves 90.8% of the teachers surv eyed as having a noneducational motive for coming to Second Life. Th ese statistics would make sense if the younger individuals in the survey were originally drawn to this MUVE for entertainment. Demographic data confirms this as 68.2% of the teachers su rveyed were primarily motivated to come to Second Life for fun, socializing, playing, and expe rimenting/exploring. As a result, it appears that younger teachers with a gaming background ma y have composed the majority of this population. We may have received very different results if the population had been more evenly divided among motivations and length of residency in SL. Future studies that focus on the ethnicity of the MUVE teacher as an important variable should ensure that enough teachers from non-Caucasian ethnicities, different age groups, and geographic areas are recruited to participate. Th is will be a difficult objective to meet due to the current demographics of Second Life and othe r MUVEs. The current demographics possibly could color the results of this study toward expe ctations and perceptions of gender and ethnicity that are part of the majority young professional culture in Europe and North America. It appears that the perceptions and expecta tions of student avatar gender and ethnicity expressed in this study may be primarily those of young, Caucasia ns who reside in the Western world.

PAGE 129

129 Implications of Virtual Representations of Self Other broader implications of this study touch on how the virtual representation of the self in MUVEs impacts the percep tions and expectations of others This section considers the effects of gender and ethnicity misrepresentatio n on students, identity exploration on teachers and others; as well as several other implications. The sample for this study was not randoml y assigned to the eight different ethnic categories due to the misrepresentation of avatar gender/ethnicity by severa l participants (Table 4-23). As a result, because the purpose of this study was to determine the impact of student avatars gender and ethnicity on teachers expecta tions and perceptions, it was logical to use the teachers perceptions to self-assign them to th e category that they thought they were in. The Asian male and Hispanic/Latino female were the most difficult for participants to identify. This is important because students in MUVEs who are Asian males and Hispanic/Latino females may be judged as being in a different ethnicity and/ or gender than they were intended. This could result in these students being misunderstood by t eachers who are seeking to be ethnically and gender conscious. For example, if an Asian male student avatar was per ceived by a teacher as a Hispanic/Latino female, then he would most likel y be treated as a Hispanic/Latino female. This could result in the student experiencing unintenti onal disadvantage due to stereotypes that exist on academic performance for Asian males (Hof stede, 1980, 1996). Future research should consider the impact that such misrepresenta tions have on student avatars in MUVEs. Identity experimentation appears to be prev alent in MUVEs (Turkle, 1995). In these environments teachers interact with students as avatars. The teacher can explore new roles, beliefs, and positions in an authentic situation while interacting with other individuals, a situation that has been shown to have significant psycho logical advantages (Turkle, 1995). When teachers

PAGE 130

130 use avatars, they choose to use certain relevant characteristics of their identities as strategic resources to enhance their participation and the overall e ffectiveness of the community (Widdicombe, 1998). Thus, teachers may have eff ectively overcome their discriminatory biases based on their own ethnicity when it came to the their perceptions and ex pectations of Student IQ, relationships with classmates attitude toward school and pred iction of future final education level, through the construction of a new identity that enables them to interact with students without prejudice. This is crucia l because if identity experimentat ion acts to mediate the effects of discriminatory biases, then it could be appl ied in virtual and face-to -face classrooms, thus giving teachers of different ethnicities a more e qual opportunity to help all students to succeed. Future research should explore wh ether online instructors more conduc ive to this type of identity experimentation than those in the face-to-face classroom. This is also important because it implie s that the face-to-face classroom may be less conducive to identity experimentation and explor ation than a virtual classroom. This makes sense because in a MUVE students and teachers ha ve a wider range of possibilities for altering their persona than in a face-to -face classroom. For example, a face-to-face classroom teacher may encourage a role play of a historical event, even to the point of dressing the part, but the simulation will still fall short of reality. In the MUVE, the environment and avatars can be customized to how that historical event looked, thus allowing students to walk in the shoes of historical figures. Identity experimentation and exploration li ke this has an additional implication for teachers. Teachers interacting with students in MUVEs now have the opportunity to not only discuss important issues and teach critical concep ts; they also can poten tially develop a deeper sense of empathy for their students. Teacher education programs often emphasize being fair to

PAGE 131

131 students by putting teachers, in their shoes, but it is taken metaphorically. In MUVEs, putting a teacher in a students shoes is literally possible. A teacher and student can swap avatars and roles and role play a discussion, activity, or anything they desire. This has signifi cant potential for how we think about teacher education. Ho w do we believe teachers should strive to understand (cognitively and emotiona lly) their students? What conditions are best for fostering that kind of inclination? Future research is needed to consider th ese important questions. It is also important to consider the ramifications of this study in light of a larger population and wider range of characteristics. Beyond ge nder and ethnicity, physical characteristics of bodily attractiveness and physical disabilities pres ent intriguing possibilities to explore. For example, one of the results of this study was th at male, Caucasian/White teachers in SL estimate that male, Asian student avatars have the sec ond lowest level of Stude nt IQ; despite common stereotypes that portray Asian ma les as intelligent and hard working. A potential explanation for this could be summed up in a remark from a colle ague when looking at the avatars, who said that the Asian male looked unattractive. As a result the Asian male could have possibly been rated lower on Student IQ due to his lack of physical attractiveness. Other aspects of how the virtual represen tation of the self in MUVEs impacts the perceptions and expectations of others may in clude psychological, emo tional, and intellectual characteristics of the virtual self. Constructs such as personality traits, emotional intelligence, internal and external locus of control, positiv e attitudes, negative attitudes and many more attributes could be potentially m easured and their impact determin ed on others expectations and perceptions. As an example, one of the results fro m this study was that teachers in SL rate male avatars as having a higher level of attitude toward school than female avatars. This may indicate a higher level of emotional intelligence for male avatars because they may be possibly better at

PAGE 132

132 being aware of and managing their own emotions in a healthy and productive manner. Thus, it would be important to focus interventions on help ing female avatars to develop their emotional intelligence in order to improve their attitude toward school. Finally, because some gender and ethnicity biases are present in MUVEs in some circumstances, what old forms of prejudice might reassert themselves due to the unique affordances of MUVEs? For example, the Gore an simulations in Second Life boast a rather large slave community that involves sadomasochi sm. Another question is what new types of discrimination will emerge due to the distinct af fordances of MUVEs? For example, in SL, a person can be whatever they want animal, vegeta ble, mineral, etc. As a result, a new race has surfaced called furries. This race is not limited to male or female gender, but also includes an androgynous gender. Thus new biases may emerge based on these new categories of ethnicity and gender. Other Future Study Ideas Several other ideas for f uture study occurred that are not connected with one of the sections above. They are presented here. This was a deception study in which teachers were led to believe that they were attempting to evaluate different forms of transc ripts used in MUVEs. However, some of the teachers may have guessed the researchers motives for this study from the content of the survey. Thus, the study should be replicated and includ e the following safeguards to ensure deception: 1. Ask a question at the end of the surv ey about what this study was about. 2. Ask a question about at what point in the survey they understood the true purpose. Ensuring deception is important because participan ts tend to think about what the researcher wants them to do. As a result, if a significan t number of participants knew the studys true purpose when they were evaluating the avatars it would bias the study.

PAGE 133

133 Another related question to ask is that if teachers in SL know that their students might not represent themselves accurately, why then would they react with any bias? In other words, if I knew that what I was seeing did not represent my st udents physical self, even if I held a bias, why would I express that bias? Research by Black et al. (2008) has revealed that different people create different avatars. Some are exact representations of self, others are ideal selves, and still others are extremely diff erent (Black et al., 2008). What is interesting is that biases in this study did appear despite the fact that the teachers surveyed did not know if the avatar represented the students real lif e appearance. Future research should consider tapping into subconscious gender and ethnicity bi ases through the use of other methodologies better suited to the task. Other research should l ook at whether the participant believes that the person they are interacting with is the actual person with whom they are interacting. For exam ple, if I am talking to an Asian female in SL, is that really who I believe Im interacting with ? And does that make a difference in terms of how I am interacting with that person? Another important consideration is the questi on that was asked at the beginning of this study in order to qualify a person to be a participant. This question could have been interpreted in many different ways. Are you teaching in Se cond Life? could have been interpreted by a person who uses SL as a tool, lecture platform or a variety of different teaching methods. Although this study did not touch on the impact of the teaching method used on teacher perspectives and expectations, it is still an interesting thought. Future research should consider controlling for the primary teaching method used as a way to determine if gender and ethnicity biases are more potent with one method than another.

PAGE 134

134 Conclusion Our purpose was to determ ine the influence of avatar choice on teacher expectations and perceptions of student success. Results showed that beliefs a bout ethnicity and gender in the physical world translated into the virtual world in surprising ways. Several suggestions for how this occurred were given, as well as a rich discu ssion of indirect and ambien t implications of this study. Biases/prejudices held by teachers must be identified, challenged, and reconstructed if educational institutions are going to improve their ability to meet the needs and interests of every student, particularly students from ethnic a nd gender groups that have received much discrimination in the American educational system. Results of this study lead to deep questions about ethnic and gender identity that border on spiritual and moral considerations. If identities are socially constructe d, then what does that mean? The constructivist argues that every person has an intellectual bias and everything that they see and interact with is influenced by that bias. The constructivist claims that the products of human action and speech are social categories, their memb ership rules, content, and valuations; and that these categor ies can and do change over time. Yet the results of this study appear to challenge this claim in some areas with regard to ethnicity and ge nder. The key here is in the phrase, over time. Over time, the rules, content, and valuations of ethnicity and gender do change. Yet it is helpful to remember th at gender and ethnicity have been around for thousands of years and have changed over that time period. Can we really expect conceptions about gender and ethnicity to change over a few years or decades? Results from this study may indicate that although quick ch ange can occur, it does not oc cur without the presence of recalcitrant areas that resist transformation. The mere introd uction of virtual world technology was hailed my many as a locale for identity expe rimentation and re-creation, but the prophets of

PAGE 135

135 technology forgot that although technology may change quickly, people do not, and when they do change they do so in rather surprising ways The evolution of ethnicity and gender is a process inherently dependent on humanity, and al though gender and ethnicity may be subject to potentially significant mediation e ffects of MUVEs, much research is still needed to prove and quantify these effects.

PAGE 136

136 Table 5-1. Similarities and Differences in dem ographics of Real Life vs. Second Life Teachers Real Life Teachers SL Teachers K-12 teachers Split between adult and K-12 teachers From North America From North America and Europe Caucasian Caucasian Sample Size = 396 Sample Size = 453 66 male, 330 female 232 male, 223 female Average age was 32.7 Under 30 years of age ? Graduate or undergraduate degree ? Single ? No children None > 3 hours daily on the Internet None > 6 hours per week in SL None Resident of SL for more than a year

PAGE 137

137 APPENDIX A IRB UFIRB 02 Social & Behavioral Research Protocol Submission Title of Protocol: Education in a Multi-User Virtual Envi ronment: The Importance of the Student Record Principal Investigator: Dennis Beck UFID #: 4638-1920 Degree / Title: Doctoral Candidate Department: Educational Technology Mailing Address: 13604 NW 137th Place Alachua, FL 32615 Email Address & Telephone Number: denbeck@ufl.edu 352-219 -0223 Co-Investigator(s): UFID#: Supervisor: Dr. Rick Ferdig UFID#: 9525-2390 Degree / Title: PhD/Associate Professor Department: Educational Technology Mailing Address: School of Teaching and Learning College of Education University of Florida PO BOX 117048 GAINESVILLE FL US 32611 7048 Email Address & Telephone Number:

PAGE 138

138 rferdig@ufl.edu 352 392 9191 ext. 275 Date of Proposed Research: May 1, 2008 to December 31, 2008 Source of Funding (A copy of the grant proposal must be s ubmitted with this protocol if funding is involved): Unfunded Scientific Purpose of the Study: Social interaction is important to learning (Vygotsky, 1978). Social interaction can occur in technology mediated forms (e.g., 5th dimension project Cole, 1995; Nicolopolou & Cole, 1993; CSILE Scardamalia & Bereiter, 1996; Scardamalia, et al ., 1994; LaFerriere, 2002). One kind of technology mediated social interaction is Internet-based social interaction (Repman, Zinskie, & Carlson, 2005). This type of social interaction is as pervasive as traditional approaches to communication (Bakardjieva, 2003). Therefore, it is important t hat we study the impact of Internet based social interaction on learning. One form of Internet bas ed social interaction that has the potential to be used for learning is the Multi-User Virtual Environm ent (MUVE) (The Horizon Report, 2007; Castronova, 2001). Experts predict that 80 Percent of active In ternet users will regularly participate in a MUVE by the end of 2011 (Gartner, 2007). Therefore, it is important that our study of the impact of Internet based social interaction on learning be extended to the study of the impact of social interaction in a MUVE. As part of a MUVE, people socially interact via avatars (Lastowka & Hunter, 2004). These avatars have the capability to be customized to minute details (Damer, 1998; Rehak, 2003; Lastowka & Hunter, 2004). Unfortunately, we know little about the effect of the detailed customization of avatars on learning. Therefore, it is impo rtant that we study the effect of the detailed customiz ation of avatars on learning (Lastowka & Hunter, 2004). We also know little about the effect of detailed customization of student avatars on teacher expect ations of students. In particular, gender and ethnicity have been shown to specifically effect teacher expectations of students in real life classrooms (Clifford & Walster, 1973; Guttmann & Bar-Tal, 1982). If we were to disc over this information, we would be able to help teachers in MUVEs to discover potential biases and pr ejudices toward some students, as well as level the playing field for student avatars of all ranges of detailed customization. Therefore, we need to study the effect of detailed customization of student avatars on expected current and future performance of the student by the teacher. Describe the Research Methodology in Non-Technical Language: ( Explain what will be done with or to the research participant. ) During the first phase of the st udy, teachers will be given a letter explaining the study, a student transcript with an attached photograph, a recommendation letter from a high school teacher, and an opinion survey (Clifford & Walster, 1973). Each teacher will be randomly assigned to one of 8 possible

PAGE 139

139 conditions. Conditions are based on the ethnicity and gender of the avatar photographs. Avatar photographs will be four male and four female avatar s of the following ethnic categories: (U.S. Census Bureau Statistics, 2007): Asian, Hispanic or Latin o, Black or African American, White or Caucasian American (see Appendix A). The letter explaining the study to the teacher will be used primarily to seek his or her cooperation. It begins by questioning the value of school records, the purpose of permanent record files, and transcripts. The letter then proceeds to explain t hat in an attempt to answer these questions, we are examining a variety of transcript forms used by different universities in Second Life. The teacher's reactions will guide us in identifying the best forms. The full text of this letter can be found in Appendix B. High school and college student transcripts will be used in place of a student summary report since the Clifford & Walster study was used with 5th grade teachers. High school and college student transcripts best approximate the information given regarding achievement on the 5th grade student summary report. See Appendices C and D for the full text of these transcripts. These transcripts were madeup for the study purpose and do not belong to a real person. The opinion survey consists of the following five items: (1) "I would estimate that the student has an IQ of. In the original st udy (Clifford & Walster, 1973), possible answers ranged from 1 (96-100) to 7 (126-130). Our study will rescale the IQ variable as a ratio variable in order to achieve greater accuracy. (2) "I would speculate that the student's social relationships wi th classmates are ---." Range of possible answers: from 5 (very good) to 1 (very bad). (3) "I would speculate that the student's relations hips with instructors ar e ---." Range of possible answers: from 5 (very good) to 1 (very bad). (4) "I would guess that the student's attitude to ward school is one of -." Range: from 6 (strong interest) to 1 (strong indifference). (5) "I would predict that the student would continue their educati on through ---." Range: from 1 (1 year of college) to 11 (Ph.D.). On the next page of the survey, teachers are asked to complete some demographic information regarding the control variables used in the study. Space will also be provid ed for the teachers to comment on their reactions to the transcript format and the type of information it provided (See Appendix E). Phase two of the study will occur immediately afte r the completion of phase one. This phase will involve the completion of some demographic inform ation as well as completion of the 25 item Professional Beliefs About Diversity Scale and the 15 item Personal Beliefs about Diversity Scale (Pohan & Aguilar, 2001). These scales assess te achers personal and professional beliefs about diversity, including attitudinal measures of ethnicity and gender. Participants will respond to each item by rating their response on a 5-point Likert-type scale ranging from 1 ( strongly disagree ) to 5 ( strongly

PAGE 140

140 agree). The full text of the Personal Beliefs About Diversity Scale can be found in Appendix F. The full test of the Professional Beliefs About Dive rsity Scale can be found in Appendix G. At the end of the study, a short explan ation of the studys true purpose w ill be read by the participants. They will also be assured that t heir responses will be kept in the strictest confidentiality and their names will not be associated wi th their responses. Describe Potential Benefits and Anticipated Risks: ( If risk of physical, psychological or economic harm may be involved, describe the steps taken to protect participant.) No anticipated risks or harm will befall the participants. Potential benefits include helping teachers in MultiUser Virtual Environments to discover potential biases and prejudices toward some students, as well as level the playing field for student avatars of all ranges of gender and ethnicity. Describe How Participant(s) Will Be Recruited, the Number and AGE of the Participants, and Proposed Compensation: A partnership with the Social Research Foundation has been established to facilitate the collection of data. The Social Research foundation is a non-profit organization dedicated to changing lives through online, interactive education programs. This organization operates the First Opinions Panel, which is the largest consumer research panel in Second Life. Through this organization panelists have been identified who are teachers in Second Life or who plan to teach in Second Life. If not enough teachers are identified through the First Opinions Panel, other sources of data will be the Second Life for Researchers and Second Life for Educators listser vs, and the following Second Life groups that operate in world: Real Life Education in Second Life, Second Life Grad Student Colony, Second Life Research, K-12 Educators, and Community Colleges in Second Life. There will be approximately 500 participants, ranging in age from 18 to 65. Participants will be compensated approximately 500 Linden dollars ($2 US) to complete the survey. Describe the Informed Consent Process. Include a Copy of the Informed Consent Document: Participants will be directed to the survey site through an email sent by the Social Research Foundation. The first page of what they see w ill be the Informed Consent Form (see Appendix F). There will be two buttons on the bottom of the form : I Consent and I DO NOT Consent. If a participant chooses the I Consent button, they will be immediately directed to the online survey. However, if a participant chooses the I DO NOT Co nsent button, they will be directed to a screen that thanks them for their time and wishes them well in their endeavors. Those consenting to the study will then be directed to the survey, which is explained in detail in the section entitled, Describe the Research Methodology in Non-Technical Language.

PAGE 141

141 Principal Investigator(s) Signature: Supervisor Signature: Department Chair/Center Director Signature: Date:

PAGE 142

142 APPENDIX B PRE-SURVEY NOTI FICATION LETTER A Survey opportunity exists for you from the Fi rst Opinions Panel of the Social Research Foundation. The Survey will begin on July 7, 2008. It will pay 1000 Linden for completion of the survey. Please answer the question below to s ee if you qualify to take this survey: Do you teach or plan to teach in Second Life? o Yes o No {We sent this as a question in a screener, rea lly a one question survey where they identified which response they matched. The Yes responses were then sent the appropriate survey.}

PAGE 143

143 APPENDIX C INFORMED CONSENT FORM Informed C onsent Protocol Title: Education in a Multi-User Virtual Environment: The Import ance of the Student Record Please read this consent document carefully befo re you decide to participate in this study. Purpose of the research study: The purpose of this study is to examine a variety of tran script forms used by different schools and universities in Second Life as to their effectiveness in providing information that really helps us understand the student as an individual. What you will be asked to do in the study: You will be given a letter explaining the study, a student high school and college transcript, an attached photograph, and an opinion survey (consisting of 5 questions) for you to complete. You will be asked to examine all of the materials and then complete the opinion survey. After completing the opinion survey, you will be asked to complete a follow up survey consisting of 40 questions. After completion of the follow up survey, you will read a short conclusion to the study. Time required: 30 minutes Risks and Benefits: No anticipated risks or benefits Compensation: You will be paid 500 Linden in compensation for participating in this research. Confidentiality: Your identity will be kept confidential to the extent provided by law. Your information will be assigned a code number. The list connecting your name to this number will be kept in a locked file in my faculty supervisor's office. When the study is completed and the data have been analyzed, the list will be destroyed. Your name will not be used in any report. Voluntary participation: Your participation in this study is completely voluntary. There is no penalty for not participating. Right to withdraw from the study: You have the right to withdraw from the study at anytime without consequence. Whom to contact if you have questions about the study: Dennis Beck, Graduate Student, Educational Technology Program Area, School of Teaching and Learning, College of Education, University of Florida, PO BOX 117048, GAINESVILLE FL US 32611 7048, phone 352-219-0223.

PAGE 144

144 Dr. Rick Ferdig, School of Teaching and Learning, College of Education, Un iversity of Florid a, PO BOX 117048, GAINESVILLE FL US 32611 7048, phone 352 392 9191 ext. 275 Whom to contact about your rights as a research participant in the study: IRB02 Office, Box 112250, University of Florida, Gainesville, FL 32611-2250; phone 392-0433. Agreement: I have read the procedure described abov e. I voluntarily agree to participate in the procedure and I have received a copy of this description. I CONSENT I DO NOT CONSENT

PAGE 145

145 APPENDIX D AVATAR PHOTOGRAPHS

PAGE 146

146 Figure C-1. Avatar Photographs Hispanic or Latino Male Hispanic or Latino Female AfricanAmerican/Black Male Asian Male Caucasian/White Male AfricanAmerican/Black Female Asian Female Caucasian/White Female

PAGE 147

147 APPENDIX E LETTER TO TEACHERS Dear Instruc tor, How purposeful are permanent record files? How revealing are transcripts? Do they provide information that really helps us understand the student as an individual? All of us educators realize the im portance of dealing with students on a one-to-one basisthe importance of establishing a unique, personalized relationship w ith each student. Does the permanent record file, summary report card, or transcript facilitate "getting acquainted?" Can the teacher, confronted with a new class of st udents, use the files to get a "head start?" As a result, we are asking you to examine the student materials provided and give their best estimate of four important pieces of information: (1) student's IQ (2) student's social status with peers, (3) students attitudes toward education, and (4) student's future educational accomplishments. Thank you for participating in th is study. We hope that it will result in bette r information available to all instructors and better rela tionships between inst ructors and students. Sincerely, Dennis E. Beck Researcher University of Florida

PAGE 148

148 APPENDIX F STUDENT HIGH SCHOOL TRANSCRIPT School of Record School Name: High School School Address: 123 Main St. Anywhere, USA 12345 School Telephone: (757) 555-1212 District Report Date Hampton 4/25/08 HIGH SCHOOL TRANSCRIPT (Grades 9-12) Student Information Name (Last, First, MI): Grad Yr: Johnson, Lee J. 2007 Address: 123 Main St. Anywhere, USA 12345 Parent/Guardian: John P. Johnson Student ID/SSN: Sex: Birth Date: Home Telephone: 123-45-6789 F 02/17/89 (123) 555-1234 COURSE STUDY 06/96 English 9 B 1 1 09 06/96 Algebra 1 B1 1 09 06/96 United States History B+ 1 1 09 06/96 Biology/Lab C 1 1 09 06/96 Personal Fitness/PE A 1 1 09 06/96 Guitar A 1 1 09 Total Credits 6 COURSE STUDY 06/97 English 10 B1 1 10 06/97 Algebra 2 C 1 1 10 06/97 World History A 1 1 10 06/97 Chemistry/Lab C 1 1 10 06/97 Personal Fitness/PE A 0.50 0.50 10 06/97 Spanish 1 B 1 1 10 06/97 Music Theory A 0.25 0.25 10 06/97 Art/Drawing B0.25 0.25 10 06/97 Health A 0.50 0.50 10 Total Credits 6.5 SUMMARY BY YEAR Mo/Yr GPA Earn Attempt COURSE STUDY 06/98 English 11 B 1 1 11 06/98 Geometry C 1 1 11 06/98 US Government B 0.50 0.50 11 06/98 Economics C 0.50 0.50 11 06/98 Physics C 1 1 11 06/98 Spanish 2 B 1 1 11 06/98 Piano A 0.50 0.50 11 06/98 Art History B 0.50 0.50 11 Total Credits 6 COURSE STUDY 06/99 Pre-Calculus C 1 1 12 06/99 Spanish 3 B 1 1 12 06/99 Anatomy C 1 1 12 06/99 English 101 C 1 1 12 06/99 Western Civ. B 1 1 12 06/99 Constitutional Law A 0.50 0.50 12 06/99 Speech 101 B1 1 12 Total Credits 6.5 END OF TRANSCRIPT Enroll Date: 09/01/03 Graduation Date: 06/01/07 06/96 3.16 6 6 06/97 2.98 6.5 6.5 06/98 2.66 6 6 06/99 2.56 6.5 6.5 Grade Table Grade Points A+ = 4.3 A = 4.0 A= 3.7 B+ = 3.3 B = 3.0 B= 2.7 C+ = 2.3 C = 2.0 C= 1.7 D+ = 1.3 D = 1.0 F = 0 Cumulative Total Credits GPA Credits GPA Points Cum. GPA Summary 25 25 68.075 Mo/Yr Course Title Final Credits Credits Grade Mo/Yr Course Title Final Credits Credits Grade Grade Earned Attempted Level Grade Earned Attempted Level 2.723

PAGE 149

149 APPENDIX G TRANSCRIPT COMMENTS AND DEMOGR APHIC S URVEY REGARDING CONTROL VARIABLES Please provide your comments and reactions to the transcript format: _________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ _____________________________________________________________________________ Please take a moment now to answer a few questions regarding yourself: 1) What is your gender? a. Male b. Female 2) What is your age? a. 18-21 b. 22-25 c. 26-30 d. 31-35 e. 36-40 f. 41-50 g. 51-60 h. 60+ 3) How many children do you have in your household a. 0 b. 1 c. 2 d. 3 e. 4 or more 4) What is your marital status? a. Single b. Married c. Domestic Partners d. Divorced e. Widowed

PAGE 150

150 f. Other 5) What level of Education have you completed? a. High school b. Some college c. Undergraduate degree, d. Graduate degree (MA, MFA, Ph.D.) e. Professional degree (CPA, etc.) 6) What is your race/ethnicity? a. Caucasian/White b. African American/Black c. Hispanic or Latino d. Asian e. Other 7) What is your religious affiliation? a. Protestant Christian b. Roman Catholic c. Evangelical Christian d. Jewish e. Muslim f. Hindu g. Buddhist h. Other 8) What is your annual family Income? a. < $50k b. $50 to 75k c. $75 to 100k d. $100 150k e. $150 200k f. $200k+ 9) What is your Internet daily use? ( NOT including tim e spent in Second Life) a. < 1 hour b. 1 2 hours c. 2 to 3 d. 3 to 5 e. 5 to 8 f. 8 to 10 g. 10 to 12 h. More than 12 10) How much time do you spend in Second Life each week? a. < 1 hour b. 1 to 2 hours c. 2 to 4 hours

PAGE 151

151 d. 4 to 6 hours e. 6 to 10 hours f. 11 to 15 hours g. 16 to 20 hours h. 21 to 30 hours i. 31 to 40 hours j. More than 40 11) How long have you been a resident of Second Life? a. < 3 months b. 3-6 months c. 6-12 months d. 12 24 months e. 24 months+ 12) What is your primary motivation or reason for coming to SL? a. Fun b. Work c. Socializing d. Playing e. Education f. Designing and building g. Experimenting/Exploring h. Other 13) You are a teacher of what kind of students? a. K-12 Students b. College or University Students c. Adult Stu dents d. Other

PAGE 152

152 APPENDIX H DEBRIEF EMAIL Thank you for participating in the recent First O pinions Panel Survey entitled, Education in a Multi-User Virtual Environment: The Importa nce of the Student Record. Please read the email below for more information on the purpose of the study. Sometimes in research it is necessary not to tell the participants the hypothesis. However, now that you are finished participating, we w ould like to tell you the pur pose of this study. We are not really interested in the importanc e of the student record for education in a Multi-User Virtual Environm ent. What we are most interested in is the effect of detailed customization of student avatars on teacher exp ectations and perceptions of students. In particular, gender and ethnicity have been shown to specifically effect te acher expectations of students in real life classrooms. If we di scover this information, we will be able to help teachers in Multi User Virtual Environments to discover potential biases and prejudices toward some students, as well as level the playing fiel d for student avatars of all ranges of detailed customization. For this study, you looked at transcripts for a st udent, along with an image and video of the student which had a specific gende r and ethnicity. You then responded to an opinion survey and answered some demographic questions. The transcripts you examined were made-up for the study purpose and do not belong to a real person. The data gathered in this study is co mpletely anonymous. It will be used to help determine if teachers in Multi User Virtual Envi ronments hold biases or prejudices toward students of varying gender and ethni city. It will also be used to help determine the nature of these biases or prejudices. Thank you very much for your participation in this study.

PAGE 153

153 If you have any questions about this study, de sire further information on this study, or do not understand the true purpose of this study, please contact: Dennis Beck School of Teaching and Learning College of Education University of Florida PO BOX 117048 GAINESVILLE FL US 32611 7048

PAGE 154

154 REFERENCE LIST Ada ms, G. R. (1978). Racial membership and physical attractivene ss effects on preschool teachers' expectations. Child Study Journal, 8, 29-41. Alexander, K. L., Entwisle, D. R., & Thompson, M. S. (1987). School performance, status relations, and the stru cture of sentiment: Bringi ng the teacher back in. American Sociological Review, 52, 665 682. Aluja-Fabregat, A., Colom, R., Abad, F. & Juan-E spinosa, M. (2000). Sex differences in general intelligence defined as g among young adolescents. Personality and Individual Differences, 28, 813 Anderson, J. (1988). The education of Bla cks in the South, 1860-1935. Chapel Hill: University of North Carolina Press. Appiah, K. A. (1994). Identity, Authenticity, Survival: Multicu ltural Societies and Social Reproduction in Charles Taylor, Multiculturalism and The Po litics of Recognition, ed. Amy Gutmann (Princeton: Princet on University Press, 1994), pp.149-63. Ajzen, I. and Madden, T. J. (1986). Prediction of goa l-directed behaviour: attitudes, intentions and perceived behavioural control. Journal of Experiment al Social Psychology 22, 453 474. Auwarter, A. E. & Aruguete, M. S. (2008). Effe cts of Student Gender and Socioeconomic Status on Teacher Perceptions. The Journal of Educational Research, 101(4), 243-246. Backe-Hansen, E. & Ogden, T. (1996) Competent girls and problematic boys ? Sex differences in two cohorts of Norwegian 10and 13-year-olds. Childhood, 3, 331. Bakardjieva, M. (2003). Virtual Togetherness: an Everyday-life Perspective. Media Culture Society, 25, 291-313 Bandura, A. (1998). Health promotion from the perspective of social cognitive theory. Psychology and Health 13, 623. Bartle, R. (2003). Designing Virtual Worlds. New Riders: Indianapolis. Bartle, R. (1999). Early MUD hist ory Retrieved April 3, 2008, from http://www.mud.co.uk/richard/mudhist.htm Beady, C. H. & Hansell, S. (1981). Teacher Race and Expectations for Stu dent Achievement. American Educational Re search Journal, 18(2), 191-206. Beaman, R., Wheldall, K. & Kemp, C. (2006). Diffe rential teacher attention to boys and girls in the classroom. Educational Review, 58(3), 339

PAGE 155

155 Beecher, C. E. (1842). A Treatise on Domestic Economy, Fo r the Use of Young Ladies at Home, and At School. Revised Edition, Boston: T. H. Webb, 1842. Bell, M. (2001). Online role-play: Anonymity, engagement and risk. Education Media International, 38 (4), 251-260. Belsey, C. (1988) Marking by numbers. AUT Women: a newsletter published by the Association of University Teachers, 15, Autumn, pp. 1-2. Benson, P. L., Karabenick, S. A., & Lerner, R. M. (1976). Pretty pleases: The effects of physical attractiveness, race, and sex on receiving help. Journal of Experimental Social Psychology, 12, 409-415 Berge, Z. (1995). Facilitating computer conf erencing: Recommendations from the field. Educational Technology, 35, 22-30 Berger, C. (2005). Interpersonal communication: Theoretical perspectives, future prospects. Journal of Communication, 55, 415-447. Bird, C. P., & Fisher, T. D. (1986). Thirty Ye ars Later: Attitudes Toward the Employment of Older Workers. Journal of Applied Psychology, 71, 515-517. Black, E. W., Ferdig, R. E., DiPietro, J. C., Liu, F., Whalen, B. (2008). Visual analyses of the creation of avatars. International Journal of Gaming and Computer-Mediated Simulations, 1(1), 90-107 Blascovich, J., Loomis, J., Beall, A., et al. (2002). Immersive virtual environment technology as a methodological tool for social psychology. Psychological Inquiry 13(2), 103. Blizzard Entertainment, Inc. (2007). World of Warcraft. Retrieved October 8, 2007 from World of Warcraft Web site: http://www.worldofwarcraft.com Bowers, J., Pycock, J. and OBrien, J. (1996): T alk and Embodiment in Collaborative Virtual Environments. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems: Common Ground. New York: ACM Press, 58. Braun, C. (1976). Teacher expecta tion: Sociopsychological dynamics. Review of Educational Research, 46, 185-213 Brigham, J. C. (1971). Ethnic stereotypes. Psychological Bulletin, 76(1) 15-38. Brophy, J. (1983). Research on the self-fulf illing prophecy and teacher expectations. Journal of Educational Psychology, 75, 633-661. Brophy, J., & Evertson, C. M. (1981). Student characteristics and teaching. New York: Longman Press.

PAGE 156

156 Brophy, J. E., & Good, T. L. (1974). Teacher-student relationships: Causes and consequences. New York: Holt, Reinhart, & Winston Broverman, I. K., Vogel, S. R., Broverman, D. M ., Clarkson, F. E., & Rosenkrantz, P. S. (1972). Sex-role stereotypes: A current appraisal. Journal of Social Issues, 28(2), 59-78 Brown, J. S. and Duguid, P.(2000). The Social Life of Information Harvard, MA: Harvard Business School Press. Brown, J. S. & Duguid, P. (1991). Organizational learning and communities-of-practice: Toward a unified view of work ing, learning, and innovation. Organization Science, 2, 40-57. Brown, J. S. & Thomas, D. (2006). You Play World of Warcraft? You're Hired! Why multiplayer games may be the best kind of job training. Wired, 14(4). Retrieved September 9, 2008 from http://www.wired.com/wired/archive/14.04/learn.html Brubaker, R. (1996). Nationalism Reframed: Nationhood and the National Question in the New Europe. Cambridge, UK. Burns, J., & Bracey, P. (2001). Boys underachieve m ent: Issues, challenges and possible ways forward. Westminister Studies in Education, 24(2), 155. Carey, J. (1980). Paralanguage in computer mediated communication. 18th Annual Meeting of the Association for Computational Linguistics and Parasession on Topics in Interactive Discourse: Proceedings of the Conference, ed. K. Sondheimer, pp. 67-69, Philadelphia, PA. Carter, D. M. (2004). Living in Virtua l Communities: Making Friends Online. Journal of Urban Technology, 11(3) 109-125 Castronova, E. (2001): Virtual Worlds: A FirstHand Account of Market and Society on the Cyberian Frontier, (December 2001). CESifo Working Paper Series No. 618. Retrieved April 3, 2008 from: http://ssrn.com/abstract=294828 Castronova, E. (2004). Trade Volume Data at http://business.fullerton.e du/ecastronova/Synthetic%20W or lds%20Economic%20Data/eco nomic_data.htm Castronova, E. (2006). Synthetic Worlds: The Business and Culture of Online Games Chicago: University of Chicago Press. Chang, L., & Krosnick, J. A. (2003). Measuring th e frequency of regular behaviors: Comparing the 'typical week' to the 'past week.' Sociological Methodology, 33, 55-80.

PAGE 157

157 Chavous, T.M., Rivas-Drake, D., Smalls, C., Griffin, T., & Cogburn, C. (2008). Gender matters, too: the influences of school racial disc rimination and racial identity on academic engagement outcomes among Africa n American adolescents. Developmental Psychology, 44(3), 637-654 Cherny, L. (1999). Conversation and Community: Chat in a Virtual World. Palo Alto, CA: CSLI Publications. Clapham, M. M. & Fulford, M. D. (1997). Age Bias in Assessment Center Ratings. Journal of Managerial Issues, 9(3) 373-387 Clark, K. (1963). Educational stimulation of ra cially disadvantaged children. In A.H. Sassow (Ed.), Education in depressed areas (pp. 142-162). New York: Teachers College. Clark, M. A., Thompson, P. & Vialle, W. ( 2008). Examining the gender gap in educational outcomes in public education: Involving preservice school counselors and teachers in cross-cultural and interdisciplinary research. International Journal for the Advancement of Counseling, 30(1), 52-66 Clark, M. A., Oakley, E., & Adams, H. ( 2006). The gender achievement gap challenge. ASCA School Counselor, 43(3), 20. Clark, E. R. and Flores, B. B. (2001). Who Am I? The Social Construction of Ethnic Identity and Self-Perceptions in Latino Preservice Teachers. The Urban Review, 33(2), 69-86 Clifford, M. M., & Walster, E. (1973). The effect of physical attractiveness on teacher expectations. Sociology of Education, 46, 248 258. Clifton, R. A,, & Baksh, I. J. (1978). Physical attractiveness, year of university, and the expectations of student-teachers. Canadian Journal of Education, 3, 37-46 Cleveland, J. N., & Murphy, K. R. (1992). Analyz ing performance appraisal as goal-directed behavior. In. G. Ferris, & K. Rowland (Eds.), Research in personnel and human resources management (Vol. 10, pp. 121). Greenwich, CT: JAI Press. Cohen, E. G. (1972). Sociology and the Classroom: Setting the Conditions for Teacher-Student Interaction. Review of Educational Research, 42(4), 441-452 Cole, M. (1995). Socio-cultural-historical psyc hology: some general remarks and a proposal for a new kind of cultural-genetic methodology. In J.V. Wertsch, P. Del Ri, & A. Alverez (Eds.), Sociocultural Studies of Mind (pp. 187-214) Cambridge: Cambridge University Press. Colom, R., Juan-Espinosa, M., Abad, F. & Garcia L. F. (2000). Negligible sex differences in general intelligence. Intelligence, 28, 57.

PAGE 158

158 Cooper, H. M., Baron, R. M., & Lowe, C. A. (1 975). The importance of race and social class information in the formation of expectancies about academic performance. Journal of Educational Psychology, 67, 312-319. Cremin, L. A. (ed.) (1957). The Republic and the School: Horace Mann and the Education of Free Men Classics in Education No. 1 (New York: Teachers College Press). Cross, T. L. (1997). Psychological and Social Aspects of Educating Gifted Students. Peabody Journal of Education, 72(3/4), 180-200. Cross, J. F., & Cross, J. (1971). Age, sex, race, and perception of facial beauty. Developmental Psychology, 5, 433-439. Csikszentmihalyi, M. (1990). Flow: The Psychology of Optimal Experience New York: Harper and Row. ISBN 0-06-092043-2 Curtis, P. (1996). Com unicazione via MUD: i feno meni sociali delle realt virtuali basate su testo [MUD communication: social aspect s of text-based virtual reality]. Sistemi Intelligenti, 8(2), 229-251. Dailey, C. A. (1952). The effects of premature conclusion upon the acquisition of understanding of a person. Journal of Psychology 33 133-152. Damer, B.F. (1998). Avatars, Exploring and Building Vi rtual Worlds on the Internet. AddisonWesley Longman/Peachpit Press. Dauber, S. L., Alexander, K. L., Entwisle, D. R. (1996). Tracking and transitions through the middle grades: Channeling educational trajectories. Sociology of Education, 69, 290-307. Dawson, K. & Ferdig, R.E. (2006). Commentary : Expanding notions of acceptable research evidence in educational technology : A Response to Schrum et al. Contemporary Issues in Technology and Teacher Education 6(1), 133-142. Available: http://www.citejournal.org/vol6/iss1/general/article2.cfm Dede, C., Nelson, B., Ketelhut, D., Clarke, J ., & Bowm an, C. (2004). Design-based research strategies for studying situated learning in a multi-user virtual environment. Proceedings of the sixth international conf erence on the learning sciences, 158-165. Mahweh, NJ: Lawrence Erlbaum Dee, T. S. (2004). Teachers, race, and student achievement in a randomized experiment. The Review of Economics an d Statistics, 86(1), 195-210 Dee, T. S. (2005). A Teacher like Me: Does Race, Ethnicity, or Gender Matter? The American Economic Review, 95(2) 158-165

PAGE 159

159 Delwiche, A. (2006). Massively multiplayer onlin e games (MMOs) in the new media classroom. Educational Technology & Society, 9(3) 160-172. DeMeis, D. K., & Turner, R. R. (1978). Effects of student's race, physic al attractiveness, and dialect on teacher evaluations. Contemporary Educational Psychology, 3, 77-86. Devine, P.G. & Elliot, A. J. (1995). Are Raci al Stereotypes Really Fading? The Princeton Trilogy Revisited. Journal of Personality and Social Psychology, 21 1139-1150 Dewberry, C. (2001). Performance disparities between whites and ethnic minorities: Real differences or assessment bias? Journal of Occupational and Organizational Psychology, 74, 659. Dieterle, E., & Clarke, J. (in press). Multi-user virtua l environments for teaching and learning. In M. Pagani (Ed.), Encyclopedia of multimedia technology and networking (2nd ed). Hershey, PA: Idea Group, Inc. Retrieved online from http://muve.gse.harvard.edu/rivercityproj ec t/documents/MUVE-for-TandL-DieterleClarke.pdf on November 26, 2008 Digman, J. M. (1963). Principal dimensions of child personality as inferred from teacher's judgments. Child Development, 34 43-60. Dion, KK; Berscheid, E; Walster, E. (1972). What is beautiful is good. Journal of Personality & Social Psychology, 24, 285. Dipboye, R. L., Arvey, R. D., & Terpstra, D. E. (1977). Sex and physical at tractiveness of raters and applicants as determinan ts of resume evaluations. Journal of Applied Psychology, 4, 288. Dooley, D. (2001). Social research methods (Fourth ed.). Upper Saddle River, NJ: Prentice-Hall Inc. Douse, N., & McManus, I. (1993). The personality of fantasy gamers. British Journal of Psychology, 84(4) 505-509. Dubrovsky, V. J., Kiesler, S., & Sethna, B. N. (1991). The equalization phenomenon: Status effects in computer-mediated and face-to-face decision making groups. Human Computer Interaction, 6, 119-146. Duncan, O. D. (1975). Introduction to Structural Equation Models. New York: Academic Press. Dunn, J. & McGuire, S. (2006). Sibling and Peer Relationships in Childhood. Journal of Child Psychology and Psychi atry, 33(1), 67-105

PAGE 160

160 Durbridge, N. (2004) Is VRML a Forgotten Language? MA Thesis. Retrieved April 3, 2008 from : http://emedia.brookes.ac.uk/neil/em7/diss2_final.pdf Dusek, J. B. (1985). Teacher expectancies. Hillsd ale, NJ: Erlbaum. Eagly, A. H. & Carli, L. L. (1981) Sex of researchers and sex-typed communications as determinants of sex differences in influen ceability: a meta-analysis of social influence studies, Psychological Bulletin, 90(1) 1-20. Eakins, B. and Eakins, G. 1978. Sex Differences in Human Communication. Boston: Houghton Mifflin. Edwards, C. (2006). Another World Engineering & Technology, 1(9) 28-32. Electronic Arts, Inc. (2007). SimCity. Retrie ved October 8, 2007 from SimCity Societies Web site: http://simcity.ea.com Ethnologue (2008). Languages of the worl d. Retrieved Septem ber 22, 2008 from http://www.ethnologue.com/home.asp Exley, Beryl (2005) Learner Characteristics of A sian EFL Students: Exceptions to the Norm In Young, Janelle, Eds. Proceedings Pleasure Passion Provocation. Joint National Conference AATE & ALEA 2005 pages 1-16, Gold Coast, Australia. Falchikov, N. and Magin, D. (1997). Detecting Gender Bias in P eer Marking of Students' Group Process Work, Assessment & Evaluation in Higher Education, 22(4), 385-396 Feldman, S. A., & Kiesler, S. B. (1974). Those who are number two try harder: The effect of sex on attributions of causality. Journal of Personality and Social Psychology, 30, 846855 Felson, R. B. (1980). Physical attractiveness, gr ades and teacher attri butions of ability. Representative Research in Social Psychology, 11, 64-71 Fennema, E. & Peterson, P. (1985). Autonomous learning behavior: A possible explanation of gender-related differences in mathematics. In: L.C. Wilkinson a nd C.B. Marrett (Eds.), Gender influences in classroom interaction Academic Press, Orlando, FL, 17. Ferguson, R. F. (2003). Teachers' Perceptions an d Expectations and the Black-White Test Score Gap. Urban Education, 38, 460-507. Feshbach, S. (1956). The catharsis hypothesis an d some consequences of interaction with aggressive and neutral play objects. Journal of Personality, 24 449-462. Finkelstein, L. M., Burke, M. J., & Raju, N. S. (1995). Age discrimination in simulated employment contexts: an integrative analysis, Journal of applied psychology, 80(6), 652662.

PAGE 161

161 Finholt, T., Sproull, L., & Kiesler, S. (1989). Communication and performance in ad hoc task groups, In Jolene Galegher, Robert Krau t, and Carmen Egido (Eds.), Intellectual Teamwork: Social and Technical Bases of Collaboration, Erlbaum, Hillsdale, NJ. Finn, J. D. (1972). Expectations and the educational environment. Review of Educational Research, 42, 387-410 Ford, J. K., Kraiger, K., & Schechtman, S. L. (1986). Study of race eVects in objective indexes and subjective evaluations of performancea meta-analysi s of performance criteria. Psychological Bulletin, 99(3), 330. Foreman, J. (2003). Next-generation: Edu cational technology versus the lecture. Educause Review, July/August, 12-22. Foster, A. L. (2007). Make-Believe College Em erges in Second Life, The Chronicle of Higher Education, October 26, 2007. Frawley, T. (2005). Gender bias in the classroom: Current cont roversies and implications for teachers. Childhood Education, 81(4), 221-227. Fulford, C. P. & Zhang, S. (1993). Perceptions of interaction: Th e critical predic tor in distance education. American Journal of Distance Education, 7(3), 8-21 Galimberti, C. (1994). Dalla comunicazione al la conversazione [From communication to conversation]. Ricerche di Psicologia, 18(1) 113-152. Gamoran, A. & Berends, M. (1987). The effect s of stratification in secondary schools: Synthesis of survey and ethnographic research. Review of Educational Research, 57, 415435. Garcia, A. & Jacobs, J. (1999). The Eyes of the Beholder: Understanding the Turn taking System in Quasi-synchronous CMC. Research on Language and Social Interaction, 32(4) 337 367. Gartner (2007). Retrieved April 24th, 2007 from http://www.gartner.com/it/section.jsp?type=press _releases&format=xhtml&year=2007 Gatti, G. G., & Harwell, H. R. (1998). Advantag es of computer programs over power charts for the estimation of power. Journal of Statistics Education, 6(3), Retrieved September 11, 2008 from http://www.amstat.org/publications/jse/v6n3/gatti.htm l Gay, G. (1975). Teachers achievement expecta tions of and classroom interactions with ethnically different students. Contemporary Education, 46, 166-171 Gee, J. P. (2003). What Video-games have to Teac h us about Learning and Literacy New York, USA: Palgrave Macmillan.

PAGE 162

162 Gioia, D. A., & Longenecker, C. O. (1994). Delvi ng into the dark side: Th e politics of executive appraisal. Organizational Dynamics, 22, 47. Glidewell, J. C., et al (1966). Socialization and social st ructure in the classroom. In M. L. Hoffman and L. N. Hoffman (Eds.), Review of Research in Child Development, No. 2. New York, N. Y.: Russell Sage Foundation, pp. 221-256. Gold, M. (1958). Power in the classroom. Sociometry, 27, 50-60. Goodwin, Charles (1986): Gesture as a Resource for the Organization of Mutual Orientation. Semiotica, 62 29. Goodwin, C., & Heritage, J. (1990). Conversation analysis. Annual Review of Anthropology, 19 283-307. Grantham, T., & Ford, D. (1998). Social needs of Danisha: An underach ieving gifted African American female. Roeper Review, 21(2) 96. Gunawardena, C. N., & Zittle, F. (1997). Social presence as a predictor of satisfaction within a computer mediated conferencing environment. American Journal of Distance Education, 11(3), 8-26 Guttmann, J. & Bar-Tal, D. (1982). Stereotypic Perceptions of Teachers. American Educational Research Journal, 19(4), 519-528 Habermas, J. (1991). A Reply, in Axel Honneth and Hans Joas, eds., Communicative Action: Essays on Jrgen Habermass The Theory of Communicative Action, translated by Jeremy Gaines and Doris L. Jones. Cambridge, Ma ssachusetts: MIT Press, pp. 214-264 (first published in German in 1986; book also published in English in 1991 by Polity Press, Cambridge, UK). Hall, R. M., & Sandler, B. R. (1984). A chilly cl imate in the classroom. In A.G. Sargent (Ed.), Beyond sex roles (pp. 503-510). St. Paul, MN: West. Hammersley, M. (1990) An evaluation of tw o studies of gender imbalance in primary classrooms, British Educational Research Journal, 16, 125. Harmon, J. (1998). Electronic meeting and intens e group conflict: Effects of a policy-modeling performance support system and an audio co mmunication support system on satisfaction and agreement. Group Decision and Negotiations 7, 131-155. Harris, M.,& Lowendahl, J (2007). Second Life university classes fo r real-life cred it. Retrieved from Gartner database on May 9, 2007.

PAGE 163

163 Havighurst, R. J., and Janke, L. L. (1944). Relations between ability a nd social status in a midwestern community, I: 10 year old children. Journal of Educational Psychology, 35 357-368. Helgeson, V. S. (1994). Prot otypes and dimensions of ma sculinity and femininity, Sex Roles, 31 653-682 Held, R. M., & Durlach, N. I. (1992). Telepresence. Presence: Teleoperators and Virtual Environments, 1 (1), 109. Hemp, P. (2006). Avatar-based marketing, Harvard Business Review Herman, A., Coombe, R. J. & Kaye, L. (2006). Your second life? Goodwill and the performativity of intellectual pr operty in online digital gaming. Cultural Studies 20(2), 184-210 Herring, S. (1999): Interactional Coherence in CMC. Journal of Computer-Mediated Communication, 4(4) Retrieved from http://www.ascusc.org/jcmc/vol4/issue4/herring.html Hiltz, S. R., Johnson, K., & Turoff, M. (1986) Experim ents in group decision making, 1: Communication process and outcome in face to face vs. computerized conferences. Human Communication Research 13, 225-252. Hindmarsh, J., Fraser, M., Heath, C., Benfor d, S. & Greenhalgh, C. (1998). Fragmented Interaction: Establishing Mutu al Orientation in Virtual Envi ronments. In Proceedings of the 1998 ACM Conference on Computer Supported Cooperative Work. New York: ACM Press, pp. 217. Hoffman, D. L., & Novak, T. P. (1996). Marketing in hypermedia computer-mediated environments: conceptual foundations. Journal of Marketing 60, 50-68. Hofstede, G. (1980). Cultures consequences: International differences in work-related values. Beverly Hills, CA: Sage. Hofstede, G. (1996). Gender Stereotypes a nd Partner Preferences of Asian Women in Masculine and Feminine Cultures. Journal of Cross-Cultural Psychology, 27, 533-546 Hore, T. (1971). Assessment of teaching practices: An "attractive" hypothesis. Brirish Journal of Educational Psychology, 4, 327-328 Houghton, S., Wheldall, K. & Merrett, F. ( 1988) Classroom behaviour problems which secondary teachers say they find most troublesome, British Educational Research Journal, 14, 297.

PAGE 164

164 Howes, C. (1988). Peer interaction in young children. Monograph of the Society for Research in Child Development Serial No. 217. Hughes, J. (1998). Therapy Is Fantasy: Roleplay ing, Healing and the Construction of Symbolic Order. Paper presented in Anthropology IV Honours, Medical Anthropology Seminar retrieved April 3, 2008 from: http://www.rpgstudies.net/hughes/therapy_is_fantasy.html Hughes, M. C. (1953). Sex differences in reading achievem ent in the elementary grades. Supplementary Educational Monograph, 77, 102-106. Irvine, J. J. (1986). Teacher-stude nt interactions: Effects of stude nt race, sex and grade level. Journal of Educational Psychology, 78, 14-21. Jakobsson, M. & Taylor, T.L. (2003). The Sopra nos meets Everquest social networking in massively multiuser networking games. FineArt forum, 17(8) 81-90 Janangelo, J. (1991). Technopower and technoppressi on: Some abuses of power and control in computer assisted writing environments. Comps and Composit, 9(1), 47-63. Jefferson, G. (1989). Preliminary Notes on a Poss ible Metric which Provides for a Standard Maximum Silence of Approximately One S econd in Conversation. In D. Roger and P. Bull (eds): Conversation: An Interdisciplinary Perspec tive. Clevedon: Multilingual Matters, pp. 166. Jennings, N & Collins, C. (2007). Virtual or Vi rtually U: Educational Institutions in Second Life. International Journal of Social Sciences, 2(3), 180-186. Jensen, A. R. (2000). Testing: Th e dilemma of group differences. Psychology, PublicPolicy, and Law, 6, 121. Jorm, A. F., Anstey, K. J., Christensen, H. & Rodgers, B. (2004). Gender differences in cognitive abilities: The medi ating role of health st ate and health habits. Intelligence, 32(1), 7-23 Kaeter, M. (1995). Age-old Myths. Training, 32(1), 61-66. Kafer, K. (2004). U.S. girl students outperform boys in most subjects, study finds. School Reform News. Chicago: Heartland Institute. Kanuka, H., & Anderson, T. (1998). On-line so cial interchange, discord and knowledge construction. Journal of Distance Education, 13(1), 57-74. Kaplan, N., Farrell, E. (1994). Weavers of we bs: A portrait of young women on the net. The Arachnet Electronic Journal on Virtual Culture

PAGE 165

165 Katz, M. (1972). Attitudinal modernity cla ssroom power and status characteristics: An investigation. Paper presented at the meeti ng of the American Educational Research Association, Chicago, Illinois. Kearsley, G. (1995). The Nature and Value of Interaction in Distance Learning (ACSDE Research Monograph No. 12). State College, Pennsylvania: Pennsylvania State University, American Center for the St udy of Distance Education. Kearsley, G., & Shneiderman, B. (1999). Enga gement theory: A framework for technologybased teaching and learning. Retrieved January 23, 2006, from http://home.sprynet.com/ gkearsley/engage.htm Keesey, C. (2007). The path less traveled: Th inking asynchronous for learning in Second Life. Paper presented at The Innovate-L ive Seminar Series, June 5, 2007. Kehle, T. J., Bramble, W. J., & Mason, E. J. (1 974). Teachers' expectations: Ratings of students' performance as biased by student characteristics. Journal of Experime ntal Education, 43, 54-60. Kelly, A. (1988) Gender differen ces in teacherpupil interac tions: a meta-analytic review, Research in Education, 39 1. Kendon, A. (1980): Gesticulation and Speech: Two Asp ects of the Process of Utterance. In M.R. Key (ed.): The Relation between Verbal and Nonverbal Communication. Mouton, pp. 207 227 Kerr D.S. & Murthy U.S. (2004). Divergent a nd convergent idea generation in teams: a comparison of computer mediated and face-to-face communication. Group Decision and Negotiation, 13, 381. Kesner, J. E. (2000). Teacher ch aracteristics and the quality of childteacher relationships. Journal of School Psychology, 28(2), 133 Kiesler, S. (1998). An Internet paradox (pan el): A social medium that may undermine sociability. In Proceedings of the 1998 ACM Conference on Computer Supported Cooperative Work, Seattle, Washington, November 14 (p. 403). New York: ACM Press. Kiesler, S., Siegel, J., & McGuire, T. W. (1984). Social psychological aspects of computermediated communication. American Psychologist, 39, 1123. Kiesler, S., Zubrow, D., Moses, A. M., & Ge ller, V. (1985). Affect in computer-mediated communication. Human Computer Interaction, 1, 77-104. Koster, R. (2002). Online World Timeline Retrieved April 3, 2008 from: http://www.legendmud.org/raph/gaming/mudtimeline.html

PAGE 166

166 Kranz, P. L. (1970), The relationship between t eacher perception of pupils and teacher behavior toward those pupils. Paper presente d at AERA meeting, Minneapolis. Laferrire, T. (2002). Tele learning: Distance and telos. Journal of Distance Education, 17(3), 29-45 Laferriere, R. J., Keller, M. L., Gossweiler, & R ., Pausch, R. (1994). An introductory tutorial for developing multi-user virtual environments. PRESENCE: Teleoperators and Virtual Environments 3(4), 255-264. LaFrance, M. (1981). Gender gestures: Sex, sex-role and non-verbal communication. In C. Mayo, & N. Henley (Eds.), Gender and non-verbal behaviour (pp. 129). New York: Springer-Verlag. Lakoff, R. (1975). Language and Woman's Place. New York: Harper & Row. Landau, J. (1995). The relationship of race and gender to managers ratings of promotion potential. Journal of Organizational Behavior, 16 (4), 391. Lastowka, F & Hunter, D. (2004). The laws of the virtual world. California Law Review, 92(1), Lee, O. & Shin, M. (2004). Addictive Consumption of Avatars in Cyberspace. Cyberpsychology & Behavior, 7(4) 417-420. Lefkowitz, J., & Battista, M. (1995). Potential sources of criterion bias in supervisor ratings used for test validation. Journal of Business and Psychology, 9(4) 389. Levy, J., Wubbels, T., Den Brok, P. & Brekelma ns, M. (2003). Students' perceptions of interpersonal aspects of the learning environment. Learning Environments Research, 6(1), 5-36 Longenecker, C. O., Sims, H. P., & Gioia, D. A. (1987). Behind the mask: The politics of employee appraisal. Academy of Management Executive, 1(3) 183. Luff, I. (2000). "I've been in th e Reichstag": Rethinking roleplay. Teaching History, 100, 8-18. Mallon, A. (2008). 2008 Second Life Surve y. Retrieved November 29, 2008 from http://www.socialresearchfoundation.org Mantovani, G. (1995). Virtual reality as a comm unication environment: consensual hallucination, fiction and possible selves. Human Relations, 48(6), 669-683. Marsh, H.W., Balla, J.R., & McDonald, R. P. ( 1988). Goodness of fit indices in confirmatory factor analysis: The effect of sample size. Psychological Bulletin 103, 391-410.

PAGE 167

167 Marwit, K. L., Marwit, S. J., & Walker, E. F. (1978). Effects of student race and physical attractiveness on teachers j udgments of transgressions. Journal of Educational Psychology, 70, 911-915. Marwit, S. J. (1982). Students' race, physical attractiveness and teachers' judgments of transgressions: Followup and clarification. Psychological Reports, 50, 242 Masland, S. (1994). Gender equity in classrooms: The teacher factor. Equity & excellence in education 27 (3), 19-27. Maslow, A. (1954). Motivation and personality New York: Harper. Matheson, K. & Zanna, M. P. (1990). Computer-m ediated communications: The focus is on me, Social Science Computer Review, 8(1), 1-12. Matusitz, J. (2005). Deception in the virtual world: A semiotic analysis of identity. Journal of New Media and Culture, 3(1) 54-63. Mayer, R. (2003). Learning environments: The case for evidence-based practice and issue-driven research. Educational Psychology Review, 15, 359. Mayer, R. & Moreno, (2003). Nine Ways to Re duce Cognitive Load in Multimedia Learning. Educational Psychologist, 38(1), 43. McDonald J. & Gibson, C. C. (1998). Interp ersonal dynamics and group development in computer conferencing. The American Journal of Distance Education, 12(1), 6-24. McEvoy, G. M. & Cascio, W. F. (1989). Cumu lative evidence of the relationship between employee age and job performance. Journal of Applied Psychology 74 11-17. McGrath, J. E. (1990). Time matters in groups. In J. Galegher, R. E. Kraut, & C. Egidio (Eds.), Intellectual teamwork: Social and technological foundations of cooperative work (pp. 2361). Hillsdale, NJ: Erlbaum. McGuire, T. W., Kiesler, S. and Siegel, J. (1987). Group and computer-mediated discussion effects in risk decision making, Journal of Personality and Social Psychology, 52(5), 91730 Argyle, M., & McHenry, R. (1971). Do spectacle s really affect judgm ents of intelligence? British Journal of Social and Clinical Psychology, 10, 27-29. McNeill, D. (1992). Hand and Mind: What Gestures Reveal About Thought. Chicago: Chicago University Press. McNemar, Q. (1942). The revision of the Stanford-Bi net scale: an analysis of the standardization data. Boston: Houghton Mifflin.

PAGE 168

168 McQuillen, J. S. (2003). The influence of technology on the initiation of interpersonal relationships. Education, 123(3), 616-623 Meadows, M. S. (2008). I, Avatar; The Culture and Consequences Of Having A Second Life. Peachpit/New Riders Merrett, F. & Wheldall, K. (1992) Teachers use of praise and reprimands to boys and girls, Educational Review, 44 73. Miller, C., McLaughlin, J., Haddon, J., & Chansky, N. (1968). Socio-economic class and teacher bias. Psychological Reports, 23, 806. Miller, J. (1991). Reaction time analysis with out lier exclusion: Bias varies with sample size. The Quarterly Journal of Ex perimental Psychology, 43A (4), 907-912. Moore, R.J., Ducheneaut, N. & Nickell, E. (2007). Doing Virtually Nothing: Awareness and Accountability in Massively Multiplayer Online Worlds. Computer Supported Cooperative Work, 16(3) 265-305. Murphy, K. R., & Cleveland, J. N. (1991). Performance appraisal: An organizational perspective. Needham Heights, MA: Allyn & Bacon. Murphy, K. R., & Cleveland, J. N. (1995). Understanding performance appraisal: Soci al, organizational, and goal-based perspectives. Thousand Oaks, CA: Sage. Murphy, K. R., Cleveland, J. N., Skattebo, A. L., & Kinney, T. B. (2004). Raters who pursue different goals give different ratings. Journal of Applied Psychology, 89, 158. Muijs, D., & Reynolds, D. (2005). Effective Teaching: Evidence and Practice. SAGE, 2005 ISBN 1412901650. Nagel, J. (1994). Constructing Ethnicity: Creating and Recreati ng Ethnic Identity and Culture, Social Problems, 41(1) 152-176 National Center for Educational Statistics (2005). Student effort and educational progress. The condition of education. Retrieved November 11, 2006 from http://www.nces.ed.gov/pubsearch /pubsinfo.asp?pubid=2005094 Newcom b, T. M. (1947). Autistic hostility and social reality. Human Relations, 1, 69-86. New Media Consortium (2007). 2007 Horizon Report Retrieved April 3, 2008 from: http://www.nmc.org/pdf/2007_Horizon_Report.pdf

PAGE 169

169 Nicolopolou, A. & Cole, M. (1993). The Fifth Dimension, its playworl d, and its in stitutional contexts: The generation and transmission of shared knowledge in a culture of collaborative learning. In E. A. Foreman, N. Minnick, & C.A. Stone (Eds.), Contexts for learning: Sociocultural dynamics in ch ildren's development (pp. 283-314). New York: Oxford University Press. Noddings, N. (1992). The challenge to care in schools: An alternative approach to education. New York: Teachers College Press. Palardy, J. M. (1969). What teachers believe-what children achieve. Elementary School Journal, 168-169, 370-374. Phillips, B. N. (1962). Sex, social class, and anxiety as sources of variation in school achievement. Journal of Educational Psychology, 53 316-322. Reeves, B. & Nass, C. (2002). The media equation: How people treat computers, television, and new media like real people and places. CSLI Publications. Oakes, J. (1985). Keeping track: How schools structure inequality. Yale University Press. Oliverio, J. C. & Beck, D. (2009). Enhanced inte raction in Mixed Social Environments. In R.F. Ferdig (Ed.), Handbook of Research on Effec tive Electronic Gaming in Education (pp. 146-162). Hershey, PA: Information Science Publishing. Orengo,V., Zornoza, A., Prieto, F.,& Peir, J. M. (2000). The influen ce of familiarity among group members, group atmosphere and asser tiveness on uninhibited behavior through 3 different communication media. Computers in Human Behavior 16, 141-159. Oyserman, D., Harrison, K., & Bybee, D. (2001). Can racial identity be promotive of academic efficacy? International Journal of Behavioral Development, 25 (4), 379. Park, J. R. (2007). Interpersonal and Af fective Communication in Synchronous Online Discourse. The Library Quarterly, 77(2), 133 Patzer, G. (1985). The physical attrac tiveness phenomena. New York: Plenum. Peng, W. (2004). Is playing games all bad? Positi ve effects of computer and video games in learning. Paper presented at 54th Annual meeting of the In ternational Communication Association New Orleans, USA, May 27-31. Peres, Y. (1971). Ethnic relations in Israel. American Journal of Sociology, 76, 1021-1047. Perkins, R. A. (2007). ICT international. TechTrends, 51(5), 20-23 Peterson, M. (2006). Learner interaction management in an avatar and ch at-based virtual world. Computer Assisted Language Learning, 19(1) 79-103.

PAGE 170

170 Petty R. E., Cacioppo, J. T. (1986). The ela boration likelihood mode l of persuasion. In Berkowitz, L. (ed.), Advances in Experimental Social Psychology Academic Press, New York, vol. 19, pp. 123. Pigott, R. L., & Cowen, E. L. (2000). Teacher ra ce, child race, racial congruence, and teacher ratings of children's school adjustment. Journal of School Psychology, 38(2), 177-196 Poole, M. S., Holmes, M., & De Sanctis, G. (1991). Conflict management in a computer supported meeting environment. Management Science, 37, 926-953. Qvortrup, L. (Ed.) (2001). Virtual interaction: Interaction in virtual inhabited 3D worlds. London: Springer. Qvortrup, L. (2002). Virtual space: Spatiality in virtual inhabited 3D worlds. London: Springer. Rafaeli, S. (1986). The electronic bulletin board: A computer driven mass medium. Computers and the Social Sciences, 2(3), 123-136. Rafaeli, S. (1988). Interactivity : From new media to communication, In Sage Annual Review of Communication Research: Adva ncing Communication Science, Vol. 16, eds R. P. Hawkins, J. M. Wiemann and S. Pingree, 110-134. Beverly Hills, CA: Sage. Rappa, N. A., Kok, D., Yip, H., & Baey, S. C. (2008) The role of teacher, student and ICT in enhancing student engagement in multiuser virtual environments. British Journal of Educational Technology, 1-9 Rehak, B. (2003). Playing at being: psychoanalysis and the avatar. In: Wo lf Mark J. P., Perron Bernard, editors. The video game theory reader. London: Routledge Reilly, J. & Mulhern, G. (1995). Ge nder differences in self-estim ated IQ: The need for care in interpreting group data. Personality and Individual Differences, 18(2) 189-192 Repman, J., Zinskie, C., & Carlson, R. D. (2005) Effective Use of CMC Tools in Interactive Online Learning. Computers in the Schools, 22(1/2), 57-69 Rey, R., Smith, A. L., Yoon, J., Somers, C. and Barnett, D. (2007). Relationships between teachers and urban African American children: The role of informant. School Psychology International, 28(3) 346. Reyhner, J. & Eder, J. (1989). A History of Indian Education. Billings, MT: Eastern Montana College. Rice, R. E. & Love, G. (1987). Electronic emo tion: Socioemotional content in a computermediated network. Communication Research, 14 85-108.

PAGE 171

171 Richmond-Abbott., T. (1979). Stereotypes of me n and women in the American culture. In M. Richmond-Abbott (Ed.), The American woman. New York: Holt, Rinehart, & Winston Riggio, R. E., Widaman, K. F., Tucker, J. S., & Salinas, C. (1991). Beauty is more than skin deep. Basic and Applied Social Psychology, 12, 423-439. Rim, Y., & Aloni, R. (1969). Stereotypes accord ing to ethnic origin, so cial class and sex. Acta Psychologica, 31(3) 12-325. Rist, R. C. (1970). Student social class and te acher expectations: the se lf-fulfilling prophecy in ghetto education. Harvard Educational Review 40 411-451. Ritts, V, Patterson, M. L., Tubbs, M. E. (1992). Expectations, Impressions, and Judgments of Physically Attractive Students: A Review. Review of Educational Research 62(4) 413426). Riva, G. & Galimberti, C. (1998). Computer-m ediated communication: Identity and social interaction in an electronic environment. Genetic, Social & General Psychology Monographs, 124(4), 434-463 Robinette, W. (1992). Synthetic experience: A proposed taxonomy. Presence: Teleoperators and Virtual Environments, 1(2) 229. Rogers, E. (2003). Diffusion of Innovations. (5th ed.). New York: Free Press. Rogoff, B. (1995). Observing Sociocultural Activity on Three Planes: Participatory Appropriation, Guided Participation, and Apprenticeship, in Sociocultural Studies of Mind, J. V. Wertsch, P. del Rio, and A. Alvarez (e ds.), Cambridge University Press, Cambridge, United Kingdom., pp. 139-164. Roscigno, V. J. (1998). Race and the Repr oduction of Educational Disadvantage. Social Forces, 76(3) 1033-1061 Ruble, D. N., & Ruble,T. L. (1980). Se x stereotypes. In A. G. Miller (Ed.), In the eye of the beholder: Contemporary issues in stereotyping. New York: Holt, Rinehart, & Winston Rubovits, P. C., & Maehr, M. L. (1 973). Pygmalion in black and white. Journal of Personality and Social Psychology, 25, 210-218. Sacks, H., Schegloff, E. A. and Jefferson, G. (1974). A Simplest Systematics for the Organization of Turn-taking for Conversation. Language, 50, 696. Sadker, M.P. & Sadker, D.M. (1994). Failing at fairness: how Americas schools cheat girls, Scribners Sons, New York

PAGE 172

172 Saft, E.W., & Pianta, R.C. (2001) Teachers perceptions of thei r relationships with students: Effects of child age, gender, and et hnicity of teachers and children. School Psychology Quarterly, 16, 125. Salvia,.J., Algozzine, R., & Sheare, J. B. (1977). Attractiveness and school achievement. Journal of School Psychology, 15, 60-67. Sanford, R.N., Adkins, M. M., Muller, R.B. & Cobb, E. A. (1943). Physique, Personality, and Scholarship. Monographs of the Society for Research in Child Development, 13 1-105 Scardamalia M. & Bereiter, C. (1996). Engaging students in a knowledge society. Educational Leadership, 54(3) 6-10. Scardamalia, M., Bereiter, C., & Lamon, M. (1994) CSILE: Trying to bri ng students into world 3. In K. McGilley (Ed.), Classroom lessons: Integrating cognitive theory and classroom practice. (pp. 201-228). Cambridge, MA: MIT Press. Schegloff, E. A. & Sacks, H. (1973). Opening up closings. Semiotica, 7, 289-327 Schiano, D. J. (1999). Lessons from LambdaMOO: A social, text-based virtual environment. Presence: Teleoperators & Virtual Environments, 8(2) 127-170. Schloerb, D. W. (1995). A quantita tive measure of telepresence. Presence: Teleoperators and Virtual Environments, 4(1) 64. Schmuck, R. (1962). Sociometric status and utilization of academic abilities. Merrill-Palmer Quarterly, 8, 165-172. Schmuck, R. (1963). Some relationship of peer li king patterns in the cla ssroom to pupil attitudes and achievement. School Review, 71, 337-359. Schmuck, R., and E. Van Egmond (1965). Sex differe nces in the relationship of interpersonal perceptions to academic performance. Psychology in the School, 2 32-40. Schroeder, R. (2002). The Social Life of Avatars. Springer-Verlag London Limited. Schrum, L., Thompson, A., Sprague, D., Maddux, C ., McAnear, A., Bell, L., & Bull, G. (2005). Advancing the field: Considering acceptable evidence in educational technology research. Contemporary Issues in Tec hnology and Teacher Education 5(3/4) Retrieved September 11, 2008, from http://www.citejournal.org/vol 5/iss3/editori al/article1.cfm Sears, R. R. (1961). Relation of early socialization experiences to aggression in m iddle childhood. Journal of Abnormal and Social Psychology, 63 466-492. Shank, P. (2004). New social interaction tools for online learning. ITFORUM PAPER #81. Retrieved from http://it.coe.uga.edu/itforum /paper81/paper81.html on October 28, 2008

PAGE 173

173 Sherbloom, J. (1988). Direction, function, and signature in electronic mail. Journal of Business Communication, 25 39-54. Sheridan, T. B. (1996). Further musi ngs on the psychophysics of presence. Presence: Teleoperators and Virtual Environments, 5(2) 241. Siegel, J., Dubrovsky, V., Kiesler, S., & McGuire, T. W. (1986). Group processes in computermediated communication. Organizational Behaviour and Human Decision Processes, 37, 157-187. Sigall, H. and Ostrove, N. (1975) Beautiful but dangerous: Effect s of offender's attractiveness and nature of the crime on juridic judgment. Journal of Personality and Social Psychology, 31, 410-414. Siggaard Jensen, S. (2007). Ref lective Designing for Actors and Avatars in Virtual Worlds. In Simon Heilesen & Sisse Siggaard Je nsen (Eds). Designing for Networked Communications, Idea Group Inc. Simpson, A., & Erickson, M. (1983). Teachers' ve rbal and nonverbal communication patterns as a function of teacher race, student gender, and student race. American Educational Research Journal, 20(2), 183-198. SimTeach (2007). Retrieved October 8, 2007 from Second Life Education Wiki Web site: http://www.simteach.com/wiki/index.php?title=Second_Life_Education_W iki Singer, E. (1996). Espoused teaching paradigms of college faculty. Research in Higher Education, 37(6), 659-79 Slater, M., Usoh, M., & Steed, A. (1994). Dept h of presence in virtual environments. Presence: Teleoperators and Virtual Environments, 3(2) 130. Spache, G. (1951). Sex differences in the Rosenzweig P-F study, children's form. Journal of Clinical Psychology, 7 235-238. Spear, M. G. (1984). Sex bias in science teacher s' ratings of work and pupil characteristics, European Journal of Science Education, 6(4) 369-377. Spears, R., & Lea, M. (1992). Social influence a nd the influence of the "social" in computermediated communication. In L. Lea (ed.), Cont exts of computer mediated communication. Hemel Hempstead: Harvester Wheatsheaf, pp. 30-65. Sproull, L., & Kiesler, S. (1991). Connections: New ways of working in the networked organization Cambridge, MA: MIT Press.

PAGE 174

174 Steinkuehler, C. A. (2004). Learning in massively multiplayer online games. In Kafai, Y. B., Sandoval, W. A., Enyedy, N., Nixon, A. S. & Herrera, F. (Eds.), Proceedings of the 6th International Conference of the Learning Sciences Mahwah, NJ, USA: Erlbaum, 521-528. Steinkuehler, C. A. (2006). Why game (culture) studies now? Games & Culture 1(1) 1-6. Stephenson, N. (1992). Snow Crash. New York: Bantam Books. Strauss, S. G. (1997). Technol ogy, group process and group outcome s: Testing the connections in computer-mediated and face to face groups. Human Computer Interaction 12, 227-266. Suzuki, L. & Aronson, J. (2005). The cultural mall eability of intelligence and its impact on the racial/ethnic hierarchy. Psychology, Public Policy, and Law, 11(2), 320 Tabachnick, B.G. & Fidell, L.S. (2000). Using multivariate statistics (4th Ed.). New York: Harper-Collins. Talamo, A. & Ligorio, B. (2001). St rategic identities in cyberspace. Cyberpsychology and Behavior, 4(1), 109 Tannen, D. (1990). You Just Don't Understand. New York: Ballantine. Taylor, C. (1995) `The politics of recognition', in A. Gutmann (ed.) Multiculturalism: Examining the Politics of Recogniti on, pp. 25. Princeton: Prince ton University Press. Taylor, T.L. (1999). Life in virtual worlds: Pl ural existence, multimodalities, and other online research challenges. American Behavioral Scientist, 43(3), 436. Tharp, R., & Gallimore, R. (1988). Rousing minds to life: Teaching, learning, and schooling in a social context. Cambridge, MA: Cambridge University Press. Thurlow, C., Lengel, L. & Tomic, A. (2004). Computer mediated communication: Social interaction and the internet. London: Sage. Tuddenham, R. D. (1952). Studies in reputation: sex and grade differences in school children's evaluations of their peers. Psychology Monographs, 333 l-39. Turkle, S. (1995). Life on the Screen: Identity in the Age of the Internet New York, USA: Simon & Schuster. Tyack, D. (1974). The one best system, a history of American urban education. Cambridge, MA: Harvard University Press. United Nations (2008). World Population Pros pects. Retrieved September 22, 2008 from http://esa.un.org/unpp/

PAGE 175

175 University of Florida (2007). I RB. Retrieved November 28, 2008 from http://irb.ufl.edu/irb02/index.html U.S. Census Bureau (2006). Hispanic or Lati no by race: Population estim ates. Retrieved on May 30, 2008. U.S. Census Bureau (2008). The 2008 Statisti cal Abstract. Retrieved September 12, 2008 from http://www.census.gov/compendia/statab/ U.S. Departm ent of Education, ( 1994). Dropout rates in the United States. National Center for Education Statistics. Van Duzer, E. (2006). Gender bias in pr edictions of boys academic failure. Action in Teacher Education, 28(1), 94-98 Vygotsky, L.S. (1978). Mind and society: The development of higher psychological processes Cambridge, MA: Harvard University Press. Waldman, D. A., & Avolio, B. J. (1991). Race effects in performance evaluationscontrolling for ability, educati on, and experience. Journal of Applied Psychology, 76(6) 897. Walther, J. B. (1992). Interpersonal effects in computer mediated inte raction: A relational perspective. Communication Research 19, 52-90. Walther, J.B., Tidwell, L.C., (1996). When is mediated communication not interpersonal? In: Galvin, K.M., Cooper, P.J. (Eds.), Making Connections: Readings in Relational Communication. Roxbury, Los Angeles, pp. 300. Walther, J., and Burgoon, J. K. (1992). Relati onal communication in computer-mediated interaction. Human Communication Research, 19(1), 50-88. Warschaur, M. (1997) Computer-mediated colla borative learning: theo ry and practice in Modern Language Journal 81(3), 470-481 Waters, M. (1990). Ethnic Options: Choosing Identities in America. Berkeley: University of California Press. We, G. (1993). Cross-gender communication in cyberspace. Arachne t Journal on Virtual Culture Weber, M. (1946). From Max Weber: Essays in Sociology trans. H. H. Gerth and C. Wright Mills New York: Oxford UP Weisband, S. P. (1992). Group disc ussion and first advocacy effect s in computer-mediated and face-to-face decision making groups. Organizational Behavior and Human Decision Processes, 53, 352-380.

PAGE 176

176 Wheldall, K. & Merrett, F. (1988) Which cla ssroom behaviours do primary school teachers say they find most troublesome? Educational Review, 40 13. White, S. H. (2000). Conceptual foundations of IQ testing. Psychology, Public Policy, andLaw, 6, 33. Widdicombe, S. (1998). But Y ou Dont Class Yourself: The Interactional Management of Category Membership and Non-Membership. pp. 52 in Identities in Talk edited by C. Antaki and S. Widdicombe. London: Sage. Wight, D., Abraham, C. and Scott, S. (1998) To wards a psycho-social th eoretical framework for sexual health promotion. Health Education Research 13, 317. Wiley, M H., & Eskilson, A. (1978). Why did you learn in school today? Teachers' perceptions of causality. Sociology of Education, 51, 261-269. Wollstonecraft, M. (1792). A Vindication of the Ri ghts of Men and A Vindica tion of the Rights of Woman. (Ed.) Sylvia Tomaselli. Cambridge: Cambridge University Press, 1995. Wong, K. F. E., & Kwong, J. Y. Y. (2005). Between-individual compar isons in performance evaluation: A perspective from prospect theory. Journal of Applied Psychology, 90, 284 294. Wood, N. T., Solomon, M. R. & Englis, B. G. ( 2005). Personalization of online avatars: Is the messenger as important as the message? International Journal of Internet Marketing and Advertising, 2(1/2), 143-161. Worrall, N., Worrall, C., & Meldrum, C. (1988). Ch ildren's reciprocations of teacher evaluations. British Journal of Educational Psychology, 58, 78-88. Yee, N. (2006). The Demographics, Motivati ons and Derived Experiences of Users of Massively-Multiuser Online Graphical Environments. PRESENCE: Teleoperators and Virtual Environments 15 309-329. Yee, N. Bailenson, J. N., Urbanek, M. Chang, F., Merget, D. (2007). The Unbearable Likeness of Being Digital: The Persis tence of Nonverbal Social Norms in Online Virtual Environments. Cyberpsychology & Behav ior, 10(1), 115-121 Yellowlees, P. (2006). Pedagogy and educat ional technologies of the future. Academic Psychiatry, 30(6) 439-441 Yinon,Y., Abenda, A., & Chirera, A. (1975). Indire ct measurement of prej udice towards Israelis of oriental origin among Is raelis of western origin. Megamont, 21, 314-323.

PAGE 177

177 Younger, M. & Warrington, M. (1996) Differentia l achievement of girls and boys at GCSE: some observations from the perspective of one school. British Journal of Sociology of Education, 17 299. Zajonc, R. B. (1980). Feeling and thinki ng: Preferences need no inferences. American Psychologist, 35, 151-175 Zimmerman, R. S., Khoury, E. L., Vega, W. A., G il, A. G., & Warheit, G. J. (1995). Teacher and parent perceptions of behavior problems am ong a sample of African American, Hispanic, and Non-Hispanic students. American Journal of Community Psychology, 23, 181-197 Zornoza, A., Ripoll, P., & Peir, J. M. (2002). Conflict Management in Groups that Work in Two Different Communication Contexts: Face-To-Face and Computer-Mediated Communication. Small Group Research, 33, 481-508.

PAGE 178

178 BIOGRAPHICAL SKETCH Dennis Beck com pleted his PhD in educational technology at the University of Florida. He also has over 17 years of experience as an educational technologist in the non-profit industry, focusing on implementing and using basic technol ogies to maximize trai ning and learning. He enjoys teaching technology and learning classes and is interested in research ing social constructs in virtual world environments and the development of narrative ident ity in virtual world environments.