Citation
Examining Influence of Instructor Presence in Instructional Videos

Material Information

Title:
Examining Influence of Instructor Presence in Instructional Videos An Individual Differences Perspective
Creator:
Wang, Jiahui
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (184 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Curriculum and Instruction
Teaching and Learning
Committee Chair:
ANTONENKO,PAVLO
Committee Co-Chair:
BEAL,CAROLE R
Committee Members:
DAWSON,KARA MARIEHOPKINS
KEIL,ANDREAS
SCHNEPS,MATTHEW H

Subjects

Subjects / Keywords:
instructor -- presence -- video
Teaching and Learning -- Dissertations, Academic -- UF
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Curriculum and Instruction thesis, Ph.D.

Notes

Abstract:
With the continued expansion of online learning, many popular online education platforms use instructional videos that integrate a live recording of a real instructor. The current study aimed to explore how instructor presence in videos influences learning, visual attention distribution, cognitive dynamics, and learner perceptions and how these effects are moderated by individual differences in working memory capacity (WMC) and inhibitory control. This study used a design with the content difficulty as a within subjects variable and instructor presence as a between subjects variable. Participants watched two statistics instructional videos of varied content difficulty, each with or without instructor presence, while the experimenter recorded their electrical brain activity via electroencephalography and eye movements using an eye tracker. Afterwards, participants self-reported their cognitive load, perceived learning, satisfaction, situational interest, social presence for both videos and their perceptions of the instructor for the videos featuring a recording of the instructor. Learning from the two videos was measured using retention and transfer questions. Individual differences in WMC were assessed using the Automated Operation Span task and differences in inhibitory control were measured using the Flanker test before the video session. Results indicated the instructor frame attracted a significant amount of attention for both easy and difficult topic videos. Instructor present difficult topic video led to lower level of theta activity, which indicated decreased working memory load. Findings also showed instructor presence improved the ability to transfer information from the difficult topic. Instructor presence produced a positive effect on reducing cognitive load for the difficult topic, and increasing their perceived learning, satisfaction, and situational interest for both topics. Instructor presence was largely perceived as helpful, entertaining, and engaging for both topics. Besides, participants who had higher WMC scores performed significantly better on the retention test for the easy topic; and those who had higher inhibitory control scores excelled on the transfer test for the difficult topic. Last, process measures such as cognitive dynamics and visual attention distribution predicted product measures such as learning and learner perception. The study provided positive evidence for including an instructor in instructional video, especially for difficult content. ( en )
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.
Thesis:
Thesis (Ph.D.)--University of Florida, 2018.
Local:
Adviser: ANTONENKO,PAVLO.
Local:
Co-adviser: BEAL,CAROLE R.
Statement of Responsibility:
by Jiahui Wang.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
LD1780 2018 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

EXAMINING INFLUENCE OF INSTRUCTOR PRESENCE IN INSTRUCTIONAL VIDEOS: AN INDIVIDUAL DIFFERENCES PERSPECTIVE By JIAHUI WANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENT S FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2018

PAGE 2

2018 Jia hui Wang

PAGE 3

To my Mom

PAGE 4

4 ACKNOWLEDGMENTS I would like to express my heartfelt gratitude and sincere thanks to my committee members, Dr. Pasha Antone nko, Dr. Kara Dawson, Dr. Carole Beal, Dr. Andreas Keil, and Dr. Matthew Schneps. Without their guidance constant supervision, and valuable suggestions throughout the entire period of the work, it would have been impossible to complete the research work. I am also thankful to my parents and my husband for all their love and support along my Ph.D. journey

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURE S ................................ ................................ ................................ ........ 11 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 Context ................................ ................................ ................................ ................... 15 Statement of Problem ................................ ................................ ............................. 17 Purpose of Research ................................ ................................ .............................. 17 Research Questions ................................ ................................ ............................... 18 Significance ................................ ................................ ................................ ............ 18 2 LITERATURE REVIEW ................................ ................................ .......................... 19 Theoretical Foundations ................................ ................................ ......................... 19 Models of Memory ................................ ................................ ............................ 20 Cognitive Load Theory ................................ ................................ ..................... 21 Dual Coding Theory ................................ ................................ ......................... 23 Cognitive Theory of Multimedia Learning ................................ ......................... 24 Human Cognitive Architecture and Video Based Instruction ............................ 25 Video based instruction designed to manage essential processing ........... 27 Video based instruction designed to reduce extraneous processing ......... 29 Video based instruction designed to foster generative processing ............ 33 Instructional Video with Instructor Presence ................................ ........................... 35 Positive and Negative Influences of Instructor Presence ................................ 35 Factors That Influence the Efficacy of Instructor Presence .............................. 41 Types of knowledge to be learned ................................ ............................. 41 Content difficulty ................................ ................................ ........................ 42 Frame size of the instructor ................................ ................................ ........ 43 Interactional style of the instructor ................................ ............................. 43 Intermittent vs. permanent instructor presence ................................ .......... 44 Learner Cognitive Differences ................................ ................................ .......... 44 Working memory capacity ................................ ................................ .......... 46 Inhibitory control ................................ ................................ ......................... 49 Measurement Considerations ................................ ................................ ................. 51 Product Measures ................................ ................................ ............................ 51

PAGE 6

6 Learning ................................ ................................ ................................ ..... 51 Learner perceptions ................................ ................................ ................... 52 Process Measures ................................ ................................ ............................ 54 Visual attention distribution ................................ ................................ ........ 54 Cognitive dynamics ................................ ................................ .................... 57 3 METHODOLOGY ................................ ................................ ................................ ... 60 Research Design ................................ ................................ ................................ .... 60 Participants ................................ ................................ ................................ ............. 61 Materials ................................ ................................ ................................ ................. 62 Apparatus ................................ ................................ ................................ ............... 63 Measures ................................ ................................ ................................ ................ 64 Pre Intervention Measures ................................ ................................ ............... 64 Pre screening survey ................................ ................................ ................. 64 Inhibitory control ................................ ................................ ......................... 64 Working memory capacity ................................ ................................ .......... 65 Measures Used During Intervention ................................ ................................ 66 Visual attention distribution ................................ ................................ ........ 66 Cognitive dynamics ................................ ................................ .................... 66 Post Intervention Measures ................................ ................................ .............. 66 Learning test ................................ ................................ .............................. 66 Learner perceptions ................................ ................................ ................... 68 Data Collection Procedures ................................ ................................ .................... 71 Data Cleaning and Scoring Procedures ................................ ................................ .. 72 Inhibitory Control Test Data ................................ ................................ .............. 72 Working Memory Capacity Test Data ................................ ............................... 72 Cognitive Dynamics Data ................................ ................................ ................. 73 Visual Attention Distribution Data ................................ ................................ ..... 73 Learning Test Data ................................ ................................ ........................... 73 Learner Perception Data ................................ ................................ .................. 74 Data Analysis ................................ ................................ ................................ .......... 74 Step 1: Checking Assumptions for Statistical Tests ................................ ......... 74 Step 2: Exploring Group Homogeneity ................................ ............................. 75 Step 3: Examine the Effects of Instructor Presence ................................ ......... 75 Step 4: Examine the Moderating Effects of Individual Differences ................... 76 Step 5: Examine the Effect of Individual Differences on Attentional Dynamics ................................ ................................ ................................ ...... 77 Step 6: Examine the Relationship between Attentional Dynamics and Other Dependent Variables ................................ ................................ ..................... 78 4 RESULTS ................................ ................................ ................................ ............... 84 Participants Demographics ................................ ................................ ..................... 84 Influences of Instructor Prese nce on Learning, Learner Perceptions, Visual Attention Distribution, and Cognitive Dynamics ................................ ................... 84

PAGE 7

7 Learning ................................ ................................ ................................ ........... 84 Retention ................................ ................................ ................................ .... 85 Transfer ................................ ................................ ................................ ...... 85 Learner Perceptions ................................ ................................ ......................... 85 Cognitive load ................................ ................................ ............................ 85 Satisfaction ................................ ................................ ................................ 86 Perceived learning ................................ ................................ ..................... 87 Situational interest ................................ ................................ ..................... 87 Social presence ................................ ................................ ......................... 88 ................................ ................. 89 eptions of instructor presence ................................ ......... 90 Visual Attention Distribution ................................ ................................ .............. 91 Cognitive Dynamics ................................ ................................ .......................... 93 Predictors of Learning and Learner Perceptions ................................ ..................... 95 Visual Attention Distribution Predicted Learner Perception .............................. 95 Easy topic video with instructor present ................................ ..................... 96 Difficult topic video with instructor present ................................ ................. 96 Cognitive Dynamics Predict ed Learning ................................ ........................... 97 Easy topic video ................................ ................................ ......................... 97 Difficult topic video ................................ ................................ ..................... 98 Cogn itive Dynamics Predicted Learner Perception ................................ .......... 98 Easy topic video ................................ ................................ ......................... 99 Difficult topic video ................................ ................................ ..................... 99 Moderating Effects of Individual Differences ................................ ......................... 100 Working Memory Capacity ................................ ................................ ............. 100 Individual differences in working memory capacity ................................ .. 100 Moderating effects of individual differences in working memory capacity 100 Inhibitory Control ................................ ................................ ............................ 10 2 Individual differences in inhibitory control ................................ ................ 102 Moderating effects of individual differences in inhibitory control .............. 102 Summary of Findings ................................ ................................ ............................ 103 5 DISCUSSION ................................ ................................ ................................ ....... 122 Influences of Instructor Pres ence on Learning, Learner Perception, Visual Attention Distribution, and Cognitive Dynamics ................................ ................. 122 Learning ................................ ................................ ................................ ......... 123 Learner Perceptions ................................ ................................ ....................... 125 Cognitive load ................................ ................................ .......................... 126 Satisfaction ................................ ................................ .............................. 128 Perceived learning ................................ ................................ ................... 129 Situational interest ................................ ................................ ................... 130 Social presence ................................ ................................ ....................... 130 Perceptions of inst ructor presence ................................ ........................... 132 Visual Attention Distribution Predicted Learner Perception ............................ 132 Cognitive Dynamics Predicted Learning ................................ ......................... 134

PAGE 8

8 Moderating Effects of Individual Differences ................................ ......................... 136 Working Memory Capacity Score Predicted Retention Performance ............. 137 Inhibitory Control Score Predicted Transfer Performance .............................. 139 Implications ................................ ................................ ................................ ........... 142 Lim itations ................................ ................................ ................................ ............. 144 Conclusion ................................ ................................ ................................ ............ 144 APPENDIX A EASY TOPIC VIDEO DETAILS ................................ ................................ ............ 146 B DIFFICULT TOPIC VIDEO DETAILS ................................ ................................ .... 148 C PRE SCREENING SUR V EY ................................ ................................ ................ 149 D LEARNING TESTS ................................ ................................ ............................... 152 E LEA R NER PERCEPTIONS SUR V EY ................................ ................................ ... 165 F INFORMED CONSENT ................................ ................................ ........................ 168 LIST OF REFERENCES ................................ ................................ ............................. 171 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 184

PAGE 9

9 LIST OF TABLES Table page 3 1 Variables in the current study ................................ ................................ ............. 79 3 2 Phases of the study ................................ ................................ ............................ 79 3 3 Data format for difficult topic ................................ ................................ ............... 80 4 1 Participant c haracteristics ................................ ................................ ................. 106 4 2 Mean accuracy (%) and standard deviations on retention test ......................... 106 4 3 Mean accuracy (%) and standard deviat ions on transfer test ........................... 106 4 4 Means and standard deviations on four types of cognitive load for the easy topic video ................................ ................................ ................................ ........ 107 4 5 Mea ns and standard deviations on four types of cognitive load for the difficult topic video ................................ ................................ ................................ ........ 107 4 6 Mean and standard deviations on satisfaction ratings with the easy and difficult topic videos ................................ ................................ .......................... 108 4 7 Mean and standard deviations on perceived learning for the easy and difficult topic videos ................................ ................................ ................................ ....... 108 4 8 Mean and standard devi ations on self reported situational interest level for the easy and difficult topic videos ................................ ................................ ..... 108 4 9 Mean and standard deviations on level of agreement with the social presence statements for th e easy topic video ................................ .................. 109 4 10 Mean and standard deviations on level of agreement with the social presence statements for the difficult topic video ................................ ............... 109 4 11 Visual attention distribution statistics for the easy topic video .......................... 110 4 12 Visual attention distribution statistics for the difficult topic video ....................... 110 4 13 Number of interpretable EEG data and number of noise free segments for each video condition ................................ ................................ ......................... 111 4 14 and alpha bands for the easy topic video ................................ ................................ ................................ ................. 111

PAGE 10

10 4 15 video ................................ ................................ ................................ ................. 112

PAGE 11

11 LIST OF FIGURES Figu re page 3 1 Group assignment for the study. ................................ ................................ ......... 81 3 2 Screenshots of the videos on the easy topic: Terminology associate d with experiments and observational studies. ................................ ............................. 81 3 3 Screenshots of the videos on the difficult topic: Rationale for the Analysis of Variance (ANOVA). ................................ ................................ ............................ 81 3 4 Experimental setup. ................................ ................................ ............................ 82 3 5 Examples of the Flanker Inhibi tory Control and Attention test ............................ 82 3 6 An example of th e Automated Operation Span task ................................ ........... 83 3 7 EEG 10 20 system. ................................ ................................ ............................ 83 4 1 e easy and difficult topic videos in the instructor present and instructor absent conditions. .................... 113 4 2 videos in the instructor present and instructor absent conditions. .................... 113 4 3 reported cognitive load for the easy topic video. ................... 114 4 4 reported cognitive load for the difficult topic video. ................ 114 4 5 ............ 115 4 6 reported perceived learning. ................................ .................. 115 4 7 reported situational interest level. ................................ .......... 116 4 8 .................... 116 4 9 ................ 117 4 10 .. 117 4 11 rib ution for the easy topic videos .................. 110 4 12 for the difficult topic videos ................ 118 4 13 Example Independent Component Analysis output from one participant. ........ 118

PAGE 12

12 4 14 and absent (perforated line) in the ea sy topic video.. ................................ ....... 119 4 15 and absent (perforated line) in the difficult topic video ................................ ..... 119 4 16 video when instructor was present (solid line) and absent (perforated line). .... 120 4 17 .......................... 120 4 18 ................................ ........ 121

PAGE 13

13 Abstract o f Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EXAMINING INFLUENCE OF INSTRUCTOR PRESENCE IN INSTRUCTIONAL VIDEOS: AN INDIVIDUA L DIFFERENCES PERSPECTIVE By Jiahui Wang May 2018 Chair: Pa vlo Antonenko Major: Curriculum and Instruction With the continued expansion of online learning, many popular online education platforms use instructional videos that integrate a live record ing of a real instructor The current study aimed to explore how instructor presence in videos influences learning, visual attention distribution, cognitive dynamics, and learner perceptions and how these effects are moderated by individual differences in working memory capacity (WMC) and inhibitory control. This study used a design with the content difficulty as a within subjects variable and instructor presence as a between subjects variable. Participants watch ed two statistics instructional vid eos of varied content difficulty, each with or without instructor presence, while the experimenter record ed their electrical brain activity via electroencephalography and eye movements using an eye tracker. Afterwards, participants self report ed their cogn itive load, perceived learning, satisfaction, situational interest social presence for both videos and their perceptions of the instructor for the video s featuring a recording of the instructor Learning from the two videos was measured using retention an d transfer questions. Individual differences in

PAGE 14

14 WMC were assessed using the Automated Operation Span task and differences in inhibitory control were measured using the Flanker test before the video session Results indicated the instructor frame attracted a significant amount of attention for both easy and difficult topic videos. Instructor present difficult topic video led to lower level of theta activity, which indicated decreased working memory load. Findings also showed instructor presence i mproved lea difficult topic. Instructor presence produced a positive effect on reducing cognitive load for the difficult topic, and increasing their perceived l earning, satisfaction and situational interest fo r both topics. Instructor presence was largely p erceived as helpful, entertaining, and engaging for both topics. Besides, participants who had higher WMC score s performed significantly better on the retention test for the easy topic; and those who ha d high er inhibitory control score s excelled on the transfer test for the difficult topic. Last, process measures such as cognitive dynamics and visual attention distribution predicted product measures such as learning and learner perception. The study provided p ositive evidence for including an instructor in instructional video, especially for difficult content.

PAGE 15

15 CHAPTER 1 INTRODUCTION Context With the continued expansion of online learning in both K 12 and higher education (Allen, Sea man, Poulin, & Straut, 2016) a growing number of instructional units are delivered online using video format on a multitude of platforms such as Coursera Khan Academy and edX (Guo, Kim, & Rubin, 2014) Striving to enhance student learning experience in the online environment, educ ators are placing much emphasis on designing and developing quality instructional videos (Crook & Schofield, 2017) Khan Academy started out as a resource that o ffered instructional videos in m athematics, with the intention of improving K 12 math ematics achievement in the United States. One prominent feature of Khan Academy videos is that they are designed without a visible instructor and inste ad format (Herold, Stahovich, Lin, & Calfee, 2011) to explain math ematics concepts that is, the learner and drawing and writing out formulas, concepts, and graphs Unlike Khan Academy, many other existing instructional videos (e.g., those created by e dX and Coursera ) include a video of the instructor as a picture in picture effect next to the learning cont ent with a substantial increase in production cost (Kizilcec, Papadopoulos & Sritanyaratana, 2014) On one hand, the presence of instructor can elicit beneficial socio emotional interaction (Clark & Mayer, 2016) On the other hand, the presence of a real instructor

PAGE 16

16 on the screen provides a group of complex visual stimuli that might distract learners and already imposed a relatively high intrinsic load (Wang & Antonenko, 2017) It is possible that the potential benefits from eliciting socio emotional responses by adding an instructor to the video ma y be offset by extraneous visual and cognitive processing associated with attending to the instructor in the video, because it conveys little content related information. Although several studies have found positive influences of showing instructor in inst ructional video s other studies revealed no or negative effects of instructor presence on learning or self reported cognitive load. The empirical evidence for the support of incorporating an instructor in instructional video s is limited and conflicting. Th erefore, further study is needed to comprehensively explore the influence of instructor presence in instructional video s I nstructional size fits individual students vary in many ways and individual differences in attention and cognition can affect the way students learn with the same media Positive (or negative) effects of instructor presence in instructional video s may be especially pronounced when a learner has diminished (or enhanced) working me mory and visual attention capacity processing speed inhibitory control, and a host of other cognitive and non cognitive variables Considering instructional videos are used by millions of learners who exhibit a wide range of attentional and cognitive dif ferences, it is important to examine how the influence s of instructor presence are moderated by individual differences in variables such as working memory capacity and inhibitory control and

PAGE 17

17 whether and how the design of videos can be improved to accommoda te the needs of a wide range of learners. Statement of Problem Empirical evidence is limited and mixed concerning the influence of instructor presence in instructional video s Therefore, further research is needed to explore the influence of instructor p resence in instructional videos. Also, i t is time to discard the one size fits all approach in designing instructional videos for all learners, and conduct rigorous research to understand how individual differences in learner attention and cognition can mo derate the influence of instr uctor presence in instructional videos. Purpose of Research T he proposed study aim s to explore how instructor presence in videos influences learning, visual attention distribution, cognitive dynamics, and learner perceptions and how these effects are moderated by individual differences in working memory capacity and inhibitory control. Existing evidence suggest s learners with differences in attention and cognition might respond to instructor presence in differential ways (Sanchez & Wiley, 2006; Rothbart, & Posner, 1985 ). Th e study focuses on the videos produced by Study Edge an online learning community that offer s instructional support on multiple topics such as Statistics, Economics, Organic Chemistry and so on. The main frame of each video is devoted to a Khan Academy s tyle pencast, whereas the bottom right hand corner always shows a shoulder up view of the instructor The instructor frame shows the instructor explaining the conte nt and displaying non verbal cues ( e.g., eye gaze, gesturing, and facial expressions).

PAGE 18

18 Resea rch Questions T he current study is designed to explore the following two research questions. Question 1. To what extent does instructor presence in easy and difficult instructional videos influence learning, visual attention distribution, cognitive dynam ics, and learner perceptions ( cognitive load, perceived learning, satisfaction, situational interest social presence, and perceptions of the instructor ) ? Question 2. How do individual differences in working memory capacity and inhibitory control moderate instructor presence effects in easy and difficult instructional videos? Significance The current study is designed to generate evidence on the influence of instructor presence on learning, visual attention distribution, cognitive dynamics, and learner pe rceptions in instructional videos. The study also contributes to the understanding of how individual differences in important cognitive variables such as working memory capacity and inhibitory control moderate these effects Findings from this study can ad vance our understanding of individual cognitive effects of instructor presence in instructional videos. Furthermore, the study will generate design implications for videos that accommodate individual differences and fur ther research on individual differences in multimedia learning.

PAGE 19

19 CHAPTER 2 LITERATURE REVIEW This chapter discusses the theoretical foundation s of multimedia learning and how the multimedia learning principles inform the design of instructional video. Next, I reviewed the empirical studies that examined the effect s of instructor presence in instructional video. Following a brief overview of various factors that interact with the influence of instructor presence on learning, perceptions, and engagement, I discuss ed how individual differences such as working memory capacity and inhibitory control can possibly moderate the influence of instructor presence in easy and difficult topic instructional videos. Finally I examined methodological considerations and compared the product measures with process measures and what additional information process measures of cognitive dynamics and visual attention distribution can reveal in this Theoretical Foundations Multimedia learning has been define d as learning from verbal and pictorial information (Mayer, 2014) Verbal information can be presented in the form of on screen text and sound (e.g., narration). Pictorial information c an be retrieved from pictures, diagrams, video, animation, and so on. Multimedia learning includes learning from animation, video, screen casting, games and simulations among other instructional media that involve pictorial and verbal information. A fundam ental hypothesis underlying research on multimedia learning is that humans learn better when multimedia is designed based on how human mind works to optimize cognitive processing of complex multimedia information (Ma yer, 2009)

PAGE 20

20 An important theoretical perspective informing research on multimedia learning is Cognitive Theory of Multimedia Learning (CTML, Mayer, 2014) It is based on several theoret ical frameworks that explain and predict how human memory processes information. They are dual coding theory ( Paivio, 1986) cognitive load theory (Sweller, van Merri nboer, & Paas, 1998) and theories of working memory (Baddeley & Hitch, 1974) Models of Memory Atkinson and Shiffrin (1968) posited t hat human memory is divided into sensory, short term and long term systems. Sensory memory is a highly transient information store that holds sensory information received via vision hearing the olfactory system, and so on If the information is attended to, it enters the short term memory. Short term memory is a central processing unit to process incoming information This processing involves filtering information by discarding irrelevant inform ation and selecting only the most relevant information units, and organizing these information units for efficient and effective integrat ion with prior knowledge that is stored in the long term memory. Long term memory stores schema ta which are mental str ucture s to organize knowledge. This multi store model of memory has also been referred to as the information processing model as it conceives human memory system as a computer, which receives input from the environment, processes it, and produces output. W orking memory is a term coined by Baddeley and Hitch in 1974 to describe the processing and storage performed in short term memory as the central processing unit of human cognitive architecture Specifically, Baddeley and Hitch (1974) proposed that similar to the overall human cognitive architecture, working memory is multi component

PAGE 21

21 system for the short term storage and manipulation of information Working memory was assumed to have sub units to process different types of information: a) visuospatial sketchpad for storing and processing visuospatial input, b) phonological loop for manipulat ing auditory information (Baddeley & Hitch, 1974) and c) central executive as an attentional control system to oversee and control the entire working memory system, and perform more integrative and thus cognitively demanding tasks such as problem solving. Later, Baddeley (1992) complemented the working memory system with a fourth component episodic buffer, which holds representations that integrate phonological, vis uo spatial information, and possibly other information that is not handled by the phonological loop or visuospatial sketchpad, for example, semantic information. The component is episodic because it is assumed to bind information into a unitary episodic representati on reflecting the specific context of information processing Episodic buffer, visuo spatial sketchpad, phonological loop, and central executive are essential components of the working memory system that altogether temporarily store and process incoming in formation, as well as integrate information that is deemed most relevant with prior knowledge in the long term memory. Other theories of working memory have been proposed however, because CTML us focuses on this particular conceptualization. Cognitive Load Theory Baddeley s (1974) working memory model discussed the structure of working memory and scholars such as Miller (1956) and Cowan (2001) suggested working memory is limited in its processing capacity. Miller (1956) found that our

PAGE 22

22 working memory system can process about seven distinc t information units. A more recent description of the limited working memory capacity was proposed by Cowan (2001) who experimentally determined that when we have to compare and contrast information, working memory only allows about four chunks of information to be processed at a time. Based on this delineation of limited working memory capacity, cogn itive load theory (Sweller, 1988) defines the amount of information being stored and manipulated in working memory as and explains and predicts how working memory processes information and interacts with long term memory (Sweller, van Merri nboer & Paas, 1998). When the information the learner needs to process exceeds the limited working memory capacity, information is not processed optimally and so learning is hindered due to excessive demands on the cognitive system To account for this, c ognitive load theory distinguishes between three types of cognitive load : intrin sic load, extraneous load, and germane load (Paas, Tuovinen, Tabbers, & Van Gerven, 2003) Different topics (e.g., solving a C alculus problem vs. adding two single digits) differ in how many information elements t hey contain and how these elements interact to produce understanding and learning. The amount of information elements and the level of element interactivity (Sweller, 2010) thus impose different levels of intrinsic cognitive load on the learner some content is more difficult to learn because there are more interacting e lements and so the intrinsic load it imposes on the cognitive system is higher For example, solving an optimization problem in Calculus involves determining the function identifying the constraints to the optimization problem, calculating the derivatives and so on whereas adding t w o single digits only requires processing two information elements using a single non complex mathematical

PAGE 23

23 operation. Thus, intrinsic cognitive load is determined by the complexity of the material (i.e., element interactivity) and is moderated by the prior knowledge the learner possesses. In contrast, extraneous cognitive load is mainly associated with a poor presentation of information and design of learning materials. For example, presentation slides designed using poor contra st between the text and background, or slides that include a lot extraneous animations that distract the learner from the relevant content, cause high levels of extraneous processing, which results in extraneous load on the cognitive resources and missed o pportunities for learning. Finally, g ermane load occurs wh en the learning content and instructions are designed to help people concentrate on learning and integrate new information with existing knowledge more effectively and efficiently without overwhelmi ng them In this sense, one could say that the art of teaching is about designing learning experiences to encourage germane load for all learners in the classroom. To conclude whereas germane load facilitates learning, intrinsic and extraneou s types of lo ad hinder learning (Plass, Moreno, & Brnken, 2010) From the perspective of multimedia learning design, multimedia instructional materials sho uld be designed to minimize extraneous cognitive load and allow more cognitive resources for germane processing of content imposing either high or low level of intrinsic load. Dual Coding Theory Because multimedia learning involves processing verbal and pictorial information, a nother important cognitive theory that underlies CTML is the dual coding theory ( Paivio 1986) The dual coding theory suggests that ver bal and pictorial information sources are processed separately by the verbal system and pictorial system

PAGE 24

24 of human memory. Dual coding is superior to single coding of information in that pictorial information may elicit an enhanced memory of verbal informat ion and vice versa As meaningful learning is facilitated when referential connections between the mental representations ( e.g. verbal and pictorial representations) are built learning with both verbal and pictorial information is hypothesized to be supe rior to learning with verbal and pictorial information alone (Paivio, 1986). Cognitive Theory of Multimedia Learning Using the t heories of dual coding, cognitive load and working memory, Mayer ( 1997, 2014 ) developed Cognitive Theory of Multimedia Learning (CTML), which explains and predicts learning with multimedia. First, CTML suggests that two separate channels of working memory (i.e., visual and auditory channels) are responsibl e for holding and processing verbal and pictorial information Second, CTML indicates that each channel of the working memory has a limited ca pacity and processing is improved when both channels are engaged. Third, CTML posits that cognitively learning is an active process of filtering, selecting, organizing, and integrating information. Extending and pictorial information in three steps: a) select incoming verbal and pic torial information for further processing in working memory; b) organize selected verbal and pictorial information into mental representations (i.e., verbal model from selected words and pictorial model from selected pictures) ; and c) integrate the already organized verbal and pictorial models with prior knowledge that is retrieved from long term memory.

PAGE 25

25 Human Cognitive Architecture and Video Based Instruction Selection, organiz ation and integration of information can be vary based on the number and natu re of the media and modalities used in multimedia learning. Some multimedia formats, such as video, consist of multiple information sources using different media and modalities. As a widely used format of multimedia learning, video provides multiple source s of information in different modalities. Verbal information such as narration, captions, and text, a s well as pictorial information such as still and animated images, live video, visual effects. Some of this information is presented auditorily and may inc lude sound effects and background music, whereas other information is p resented in the visual modality. Information presented in video is also highly transient in that it is presented to the viewer dynamically and unless the video is paused, the learner ma metacognitive processes play an important role in video based instruction but are beyond the scope of this study. Although many parameters remain unknown as to what makes a video instructiona lly effective, affordances and constraints of human cognitive architecture render certain designs of video based instruction effective or not (Sweller et al., 1998) As working memory processes information via two channels (i.e., visual and audit ory) and dual coding is assumed to be superior to single coding (Paivio, 1986) effective video design should utilize this affordance of human cognitive architecture and allow information to be processed via two channels simultaneously. A typical design of video based instruction is where the instructor narrates the material while presenting the pictures and words on the screen. Narration (i.e., verbal information) is processed by

PAGE 26

26 the auditory channel and on screen information (i.e., visual information) is processed by the visual channel. As the learner simultaneously integrates verbal and visual information into representational structures in working memory, the construction of referential connections between verbal and pictorial information is facilitated. Dual coding and integration of narration and visual information on the screen could potentially support information processing in working memory resulting in enhanced comprehension of the material (Mayer & Moreno, 2003) Despite the affordances of working memory discussed abov e, the working memory system is also subject to the constraint of limited capacity. Cognitive load theory is especially useful in this regard as it influences the design of multimedia instruction to overcome the limited capacity of working memory, which is a major impediment to learning in general and multimedia learning in particular During the past three decades, Mayer and colleagues have conducted over 100 experimental tests and summarized their findings into several multimedia learning principles aimin g to manage cognitive load in multimedia learning (Mayer & Moreno, 2003) In other words multimedia learning should be designed to decrease extraneous cognitive load and allow more cognitive resources for germane processing of content imposing either high or low levels of intrinsic load. With a strong c onnection to the three types of cognitive load (i.e., intrinsic load, extraneous load and germane load), the twelve principles of CTML are grounded in the framework of three types of cognitive processing that is, essential processing, incidental processi ng, and generative processing. CTML principles based on these types of processing provid e guidance for the design of video

PAGE 27

27 based instruction and minimize the negative consequences potentially caused by the constraints of working memory. Video based instr uction designed to manage essential processing Essential processing refers to the process of selecting relevant words and images and organizing them as presented required to represent the essential material (Mayer & Pilegard, 2014 p. 317 ). Similar to the conceptualization of intrinsic load, t his type of proces sing is mainly related to the intrinsic complexity of material. Information with high intrinsic load has a higher possibility of affecting essential processing as it can be more cognitively demanding. Management of essential processing is proposed to be gu ided by t hree relevant multimedia learning principles : segmenting, pre training, and the modality principle (Mayer & Pilegard, 2014) The m odality principle states p eople learn better from animation and narration than from animation and on screen text (Mayer & Moreno, 2003) On screen text and pictures both constitute visual informa tion, and they compete for cognitive resources in the visual channel, potentially overloading the visual channel. Alternatively, conveying information through both visual and auditory channels by using on screen pictures (i.e., animation) and narration wil l utilize visual and auditory channels more efficiently, decreasing the possibility of either channel being overloaded. Several studies found that the performance on problem solving transfer was better when scientific explanations were presented in the for m of animation and narration compared to animation and on screen text (Mayer & Moreno, 1998; Moreno & Mayer, 1999; Moreno, Mayer, Spires, & Lester, 2001) The m odality principle easily extends to video based instruction. Instead of presenting on screen text video base d instruction can use narration to deliver verbal

PAGE 28

28 information auditorily while allowing the visual channel to process only on screen pictorial information. Besides allocating verbal and pictorial information to be processed by the two channels of working memory, multimedia instruction can also be designed to decrease the amount of information to be processed at a time. The segmenting principle states people learn better when multimedia material is presented in smaller segments rather than as a larger cont inuous unit (Mayer & Mo reno, 2003) This principle particularly applies to complex topics I ntrinsically complex material consumes a significant amount of cognitive resources at a time, potentially overloading the working memory. Instead, breaking the large presentation of cont ent down in to several manageable segments relieves the burden on working memory. Empirically Mayer & Chandler (2001) broke a narrated animation explaining lightning formation into 16 segments and found that compared to students who watched the whole presentation continuously, students who r eceived the segmented instruction performed better on the test of problem solving transfer. In terms of video based information, presenting information in shorter and manageable segments can facilitate learning and lead to deeper understanding of materials By examining videos in Massive Open Online Courses (MOOCs) Guo and colleagues (2014) found that shorter video s (e.g., within 6 minutes long) were more engaging than longer video s Both the modality principle and the segmenting principle aim to ensure work ing memory functions under its capacity. Another approach to prime essential processing is to allow the learner to receive training on components preceding instruction (Mayer & Moreno, 2003) dubbed as pre training principle. Pre training allows schema in the long

PAGE 29

29 term memory t o be activated prior to instruction to improve construction of mental models integrating new information and schema in long term memory which otherwise often overload the working memory. In previous studies, students were presented animations on how brake s and pumps work and it was found that students who had received pre training about the names and behavior of the components performed better on problem solving transfer tests (Mayer, Mathias, & Wetzell, 2002; Pollock, Chandler, & Sweller, 2002) than those who had not received any pre training Extending th ese findings to video based learning, preview of key information can be provided prior to presenting instruction. To summarize, t he modality principle and the segmenting principle take the limited capacity of working memory into account and aim to avoid overloading the cognitive system. The modality pri dual channel processing and maximizes essential processing. Pre training benefits essential processing in that pre training facilitates integrating information in working memory with existing prior knowle dge in long term memory. Video based instruction designed to reduce extraneous processing (Mayer & Fiorella, 2014 p. 281 ). For example, supplementing an animation with unnecessary background music can increase extraneous processing, as the learner needs to devote extra cognitive resource to process this type of extraneous information. Extran e ous processing of material will cause extraneous cognitive load to be increased. Extran e ous processing can be min imized using the CTML principles including the redundancy

PAGE 30

30 principle, the coherence principle, spatial contiguity, temporal contiguity, and the signaling principle (Mayer & Fiorella, 2014) The redundancy principle is derived from the m odality principle which states people l earn better when animation is accompanied by narration rather than on screen text (Mayer & Moreno, 2003) In the case of the redundancy principle, it is emphasized that people learn better from animation and narration than from animation, narration, and on on screen text becaus e on screen text serves as redundant information source given that the same information is already presented via narration (Mayer & Moreno, 2003) Studies found that students who learned from non redundant presentations performed better on problem solving transfer tests than th ose who were provided with the redundant version of presentation (Mayer, Heiser, & Lonn, 2001; Moreno & Mayer, 2002) A possible explanation is that the learner s who are presented with narration and on screen text need to compare and contrast information for each source to establish whether they convey t he same or different information and synthesize these two sources of verbal information as they process the learning materials This type of extraneous processing can overload the working memory, as working memory is also responsible for integrating verbal information with on screen pictures. According to the redundancy principle, video based instruction should eliminate redundant on screen text when it can be delivered by narration. The c oherence principle highlights that people learn better when interest ing but irrelevant words, pictures, and sounds are excluded (Mayer & Moreno, 2003) I nteresting but irrelevant visual and auditory stimuli can distract the learner from the focus of the instruction, and so they need to be excluded rather than included in

PAGE 31

31 instruction. S tudents w ho received a concise narrated animation performed better on problem solving transfer tests than those who watched an embellished narrated animation, which consisted background music and irrelevant video clips in the middle of the animation (Mayer, Heiser, & Lonn, 2001; Moreno & Mayer, 2000) According to the coherence principle, video based instruction should consider removing or minimizing all irrelevant and distracting information that may result in unnecessary extraneous processing. Unlike the redundancy principle and the coherence principle, which aim to minimize unnecessary information processing, contiguity principles keep extraneous processing to a minimum by aiding the learner in integrating verbal and pictorial information and building referential connections between the modalities (Mayer & Moreno, 2003) The contiguity principle requires ver bal and pictorial information to be presented together, either in proximity (mandated by spatial contiguity principle) or simultaneously (stated by temporal contiguity principle). The s patial contiguity principle suggests people learn better when correspon ding words and pictures are presented spatially near rather than far from each other on the page or screen (Mayer & Moreno, 2003) The t emporal contiguity principle states that people learn better when corresponding words and pictures are presented simultaneously at the same t ime, rather than successively (Mayer & Moreno, 2003) Presenting words and pictures far from each other either in time or space requires the learner to devote extra cognitive resources to search for verbal or pictorial information, while holding the other type of information in mind (i.e., pictorial or verbal). This often causes a negative consequence (Ayres & Sweller, 20 14) Mayer and

PAGE 32

32 colleagues (1999) found that students performed better on a problem solving transfer test when they le arned from integrated presentations (consisting of animation with integrated on screen text) than did students who learned from separated presentations (consisting of animation with separated on screen text ) The same study also found that students who lea rned from simultaneous presentations performed better on tests of problem solving transfer than those who learned from successive presentations. According to contiguity principles, video based instruction should consider presenting verbal and pictorial inf ormation in an integrated manner, either simultaneously or in close spatial proximity with each other. The s ignaling principle states that better transfer occurs when narrations and on screen animation are signaled (Mayer & Moreno, 2003) A signaled version of instruction can stressing key words in narration or providing visual cues (i.e., arrows) in on screen animation ( Mautone & Mayer, 2001) On the other hand, a non signaled version of instruction will increase extraneous processing in that learner will likely focus to a larger extent on nonessential aspects of the instruction ( Mautone & Mayer, 2001) To illustrate this principle empirically, Mautone and Mayer (2001) co nstructed a four min ute narrated animation explaining how airplanes achieve lift in a signaled and a non signaled version. S tudents who received the signaled version of the narrated animation performed better on a subsequent test of problem solving transfe r than did students who received the non signaled version. According to the signaling principle, video based instruction can employ

PAGE 33

33 during the video presentation. The presenter can also stress important words in the narration. Video based instruction designed to foster generative processing Multimedia instruction can be designed not only to manipulate the amount of essential processing and extraneous processing, but also foste r generative processing. Generative processing helps the learner to mentally organize new information into structures and integrate it with schema ta in long term memory (Moreno & Mayer, 2010) Generative processing results in germane load, which according to cognitive load theory, occurs when assimilation or accommodation of presented information is encouraged rather than hindered The evidenced based instructional strategies for fostering generative processing are based on the multimedia principle, personalization principle, voice principle and interactiv ity principle (Moreno & Mayer, 2010) First and f oremost, extending the dual coding theory, Mayer & Sims (1994) posited that people learn better from words and pictures than fr om words alone, referred to as the multimedia principle the core principle of CTML The memory enhancing effect of pictures encourages the assimilation of presented verbal information. When verbal and pictorial information are both presented, learners hav e the opportunity to build verbal and pictorial models and integrate them with each other. Numerous studies conducted by Mayer and other scholars have demonstrated the efficacy of the multimedia principle and supported integrative use of pictures with word s (Mayer & Sims, 1994) Mayer (1997) reviewed eight studies that had been conducted to compare multimedia instruction with single media instruction, and consistently found the positive effect of multimedia instruction. L earners who received multimedia instruction performed

PAGE 34

34 significantly better on subsequent problem solving transfer tests than those who received verbal instruction alone. Nowadays, video based instruction generally adopts multim edia presentation and provides both pictures and words, which allow assimilation and integration of verbal and pictorial information With the recent advancement s in media design and information technology, other instructional elements in multimedia instr uction can be manipulated to foster generative processing. For instance, p eople learn significantly better when the verbal information is presented in conversational style (e.g., first or second person) rather than formal style (e.g., third person), referr ed to as the personalization principle (Moreno & Ma yer, 2010) Moreover, people learn better when narration in multimedia instruction is spoken in a friendly human voice rather than a machine voice, referred to as voice principle. Video based instruction can utilize a real human friendly voice in a conver sational style to satisfy both personalization principle and voice principle. The i nteractivity principle states that people learn better when they have control over the pace of the presentation (Moreno & Mayer, 2010) Mayer & Chandler (2001) found that learners performed better on transfer tests when they had control over the pace of the narrated animation, compared to those who watched the same narration without any learner control. Video based instruction can easily fulfill the interactivity principle as most video players provide the affordances to pause, rewind, or fast forward the vid eo. These multimedia learning principles have offered important implications for the design of multimedia learning especially video based instruction. Taking the affordances and constraints of human cognitive architecture into consideration, video

PAGE 35

35 based i nstruction can be designed to foster essential processing and generative processing, while minimizing extraneous processing. Instructional Video with Instructor Presence Positive and Negative Influences of Instructor Presence The instructor embedded in t he video is a source of visual information that provides primarily nonverbal communication cues to the learner It is acknowledged that nonverbal communication plays an important role in interpersonal interaction (Argyle, 2013) and facilitates face to face learning (Goldin Meadow, & Alibali, 2013; Holler, Shovelton, & Beattie, 2009; Knapp, Hall, & Horgan, 2013) including language learning (e.g., Church, Ayman Nolley, & Mahootian, 2004) and mathematics learning (e.g., Alibali & Nathan, 2012) Th e utility of nonverbal communication also extends to online learning. In the context of instructional video, the image/video of the instructor may result in deeper engagement and cognitive processing of learning content due to the activation of social inte raction schema that people employ during traditional, face to face communication and learning (Clark & Mayer, 2016) The instructor typically provides such means of nonverbal communication as mutual gaze, gesturing, and facial expressions. These nonverbal cues can possibly support t he cognitive processing of verbal information that is narrated by the instructor, thus improving comprehension. As these nonverbal cues constitute visual information, processed primarily by the visuospatial sketchpad, it will not interfere with the process ing of auditory information (e.g., narration), which is handled by the phonological loop. In fact, based on the dual coding theory (Paivio, 1986), instructor should theoretically complement each other as they are processed by

PAGE 36

36 different channels (e.g., auditory and visual) and support information processing in two separate channels resulting in enhanced comprehension of the material. The presence of instructor should not only facilitate deeper processing o f learning content, but also improve students learning experience by encouraging deeper engagement with the content. Social agency theory suggests social cues in multimedia presentations lead learners to feel as if they are interacting with another person (Cui, Lockee, & Meng, 2013) The basic assumption is that social cues replicate the social aspects of human interaction, and this may induce beneficial socio emotional responses in the learner. Thus, s ocial cues provided by the instructor should result in enhanced positive affective responses externalized, for example, as high satisfaction or engagement ratings, while learners watch instructional videos featuring instructor presence In addition to can also attract a significant amount of attention (see Yee, Bailenson, & Rickertsen, 2007 for a review of research on the effects of faces in human computer interfaces). H uman attraction to faces has been shown to already develop in infants who attend favorably to faces and face like configurations (Johnson, Dziurawiec, Ellis, & Morton, 1991) As face relevant stimuli can assis t people in seeking social cues such as gaze, facial expression, and so on people are attracted to face as a source for information that facilitate social interaction and communication (Farroni et al., 2005) Consequently, it is substantial impact on the distributio n of visual attention. While it is understandable why

PAGE 37

37 provides other important instructional components on the screen such as text, diagrams, pencas ts and so may also serve as a distract or a nd hinder cognitive processing of the essential learning material Based on Baddeley working memory model, both the instructional components on the screen and the instructor ne ed to be processed by the visual spatial sketchpad structure of working memory, which has severe limitations in terms of both capacity and dura tion ( Paivio, 1991) The instructor frame will then compete for visuo spatial focused cognitive resources with o ther instructional information and possibly overload the visual channel. From a cognitive load perspective, the instructor video may result in increased extraneous cognitive load which will hinder the processing of the information o n the rest of the scree n. Emerging evidence for this hypothesis was provided by Djamasbi and colleagues (2012) who found images of human face had performance o n tasks that are based on the information located in close proxim ity to the face suggesting that face imagery was too distracting for many users to handle The t wo types of visual stimuli that is, the instructor and the text, diagrams and other content in the rest of the screen will likely result in split attention (Sweller, Ayres, & Kalyuga, 2011) d to split their attention between and mentally integrate several sources of physically or temporally disparate information, where each source of information is essential for understanding (Ayres & Sweller, 2014) If the learner needs to mentally integrate the two sources of information that are either temporally or spatially isolated, learning materials will impose high extraneou s cognitive load on the learner. Split attention is a particular issue in the case of content with high element interactivity (i.e., high intrinsic load),

PAGE 38

38 whereas its effect within content with low element interactivity may not be as pronounced (Kalyuga, Chandler, & Sweller, 2011) In the video based instruction with instructor presence, split attention could occur when attention is switched between the instructor and the rest of content in the video frame (Johnson, Ozogul, Moreno, & Reisslein, 2013; Schmidt Weigand et al., 2010) Empirically, Kizilcec and colleagues (2014) attention distribution. They observed that p articipants switched between the instructor frame and the rest of the screen every 3.7 seconds. Thus, i t is reasonable to hypothesize that this level of split attention could interfere with the effective processing of the important instructional inf ormation presented in the rest of the frame and impede learning. Bringing it all together, the presence of instructor in instructional video could elicit beneficial socio nonverbal modali ties of interaction (Clark & Mayer, 2016) On the other hand, the instructor presence also provides a group of complex visual stimuli that might distract lear ners especially when a high level of cognitive processing is involved due to high intrinsic load imposed by the complex lear ning content Limited empirical evidence suggests a tradeoff between the costs and benefits of presenting the instructor in video but the findings are not consistent. Several studies have examined the influence of instructor presence on learning and perce ptions; however, these findings are largely inconclusive. Evidence of positive effect was provided by Chen & Wu (2015) who used an experimental design and compared the influence of three types of videos on learning document writing : voice

PAGE 39

39 video in the upper left corner of the screen), lecture capture (i.e., a video recording of the lecture) and picture in picture video that includes an instructor frame in the bottom right hand corner The results indicated that performance on recall and transfer of learning with lecture capture and picture in picture formats were superior to that ass ociated with the voice over format. The intensity of positive and negative emotions elicited by the three video formats did not appear to significantly differ from each other. They also found the self reported cognitive load related to the lecture capture and picture in picture types of video was significantly lower than that of the voice over video Additional positive evidence was provided by Pi and Hong (2016) who examined the effects of video presentation types on learning a topic in developmental psychology. The researchers compared four different video types, PowerPoint slides alone, instructor video alone, PowerPoint slides with instructor video and classroom recording (including PowerPoint slides, instructor, and students in the classroom). Learning from the video was measured by recall and transfer. The findings showed although t he participants distributed a significant amount of attention to the instructor (62.3%) the PowerPoint slides with instructor condition resulted in significantly higher learning test scores compared to the other three video types. Besides using experimen tal designs, researchers have also mined Massive Open Online Course (MOOC) interaction logs and examined the influence of instructor presence on student engagement (Guo, Kim, & Rubin, 2014) In a large scale study of MOOC based videos i n 6.9 million video wat ching sessions across four courses on the edX MOOC platform, Guo and colleagues (2014) examined two proxies for engagement: engagement time (i.e., video watching session length) and problem

PAGE 40

40 attempt at follow up multiple choice problem. They found students we re engaged more with videos that included an instructor compared to videos with PowerPoint slides alone. At the same time other studies revealed no or negative effects of instructor presence in video on learning or cognitive load Homer, Plass, & Blake (2008) conducted an experiment in which undergraduate students viewed one of two versions of a computer based multimedia presentation on The Millennium Dialogue on Child Development: one in cluded slides with instructor and the other only slides with audio narration. They compared two conditions by measuring recall and transfer of knowledge, as well as using a social presence questionnaire. They found no significant difference in lea rning or social presence between the two conditions. However, t he study also found participants self reported a significantly higher cog nitive load in the condition with instructor present on the screen compared to the audio narration condition. Kizilcec and colleagues (2014 ) investigated how adding the instructor to instructional video influences in Organizational Sociology. Although learners strongly preferred video instruction with instructor presence and perceived it as more educational, they did not perform significantly better or worse o n knowledge recall tests compared to the control condition without instructor presence. While several studies have found positive influence of providing instructor presence in instructional video s other studies revealed no or negative effects of it on le arning or self reported cognitive load. The empirical evidence for the support of incorporating an instructor in the instructional video is limited and conflicting. The

PAGE 41

41 current study is designed to further examine the effect of instructor presence o n learn ing, visual attention distribution, cognitive dynamics, and learner perceptions in instructional videos. Factors That Influence t he Efficacy of Instructor Presence E mpirical evidence suggested that adding the instructor to instructional video is not alwa ys beneficial and potential positive effects of instructor presence depend on several factors The se factors include type of knowledge to be learned content difficulty, the frame size of the instructor, interactional style of the instructor, and whether i nstructor presence is permanent or intermittent Types of knowledge to be learned Hong and colleagues (2016) examined the inf luence of instructor presence in video on declarative knowledge (a topic in educational technology) and procedural knowledge ( perform an action in Adobe Photoshop ). The videos on these two topics were presented in two different styles: PowerPoint slides or PowerPoint slides with a picture in picture video of the instructor. The findings indicated that adding the instructor to the PowerPoint slides only facilitated declarative knowledge development In the video on a procedural knowledge topic the learn ing test scores were no different between the two conditions; however, learners who watched the video with instructor presence reported a significantly higher level of cognitive load. The authors explained that a higher level of cognitive resources was req uired when learning procedural knowledge than declarative knowledge and thus the instructor video could possibly interfere with the learning process and overload learners The findings were important as they suggested the positive effects of instructor pre sence in instructional

PAGE 42

42 video s varied depending on the type of knowledge being taught. It is possible learners who watched the video on declarative knowledge with instructor presence showed greater learning due to the benefits of non verbal communication pr ov ided by the instructor presence; however, this positive effect could be reduced by additional a procedural knowledge topic Content difficulty The type of knowledge is not the only variable that has been found to moderate the effects of instructor presence in instructional video A nother highly relevant factor that has been examined is content difficulty because it relates directly to the levels of intrinsic load imposed by the learning conte nt Wang and Antonenko (2017) found the effects of instruc perceptions were moderated by the content difficulty of the topics. Undergraduate p articipants were assigned to watch two videos on an easy mathematics topic (similar triangles) and a d ifficult topic (Trigonometry), each with the instructor present or absent in a counter balanced design The findings generally supported the use of instructor presence in the videos, but the influence of instructor presence varied for the two topics. S howi ng the instructor improved recall of information when it was present in the easy topic, but there was no difference in recall for the difficult topic with the instructor present or absent. The eye movement analysis also indicated instructor frame attracted significant amount of visual attention, especially for the easy topic video (25%) T he instructor presence version of the video was perceived as less demanding (lower mental effort ratings) for the difficult topic video, but this effect was not demonstrat ed for

PAGE 43

43 the easy topic when the instructor was present. This study provided some evidence that the decision of including an instructor in instructional video should also be made based on content difficulty. Fram e size of the instructor In addition to exp loring the moderating effect of type of knowledge and content difficulty researchers such as Pi and colleagues (2017) examined if there was any differential effect of the instructor perceptions of social presence, cogniti ve load, learning performance, and satisfaction. Learners watched video on adjusting a curve in Adobe Photoshop that included a small, medium or large image size of the instructor. The results demonstrated the small size of instructor facilitated learning compared to the other two conditions. Learners were also more satisfied with the instructional video s including a small size of the instructor The perceived level of social presence and cognitive load were the same across the three video s The re required fewer attentional resou rces compared to a larger image size, while preserving the non verbal cues and eliciting social responses. Interactional s tyle of the instructor The interacti onal style of the i nstructor style also played a role based on evidence from extant literature. Bhat and colleagues (2015) used clickstream data from one Coursera course to analyze the engagement (i.e., video wat ching time, discussion forum visits following a lecture view), motivation (i.e., certificate earner proportion, fraction of lectures and quizzes that the learner viewed and submitted) and navigational patterns of learners upon being presented with videos i ncorporating the

PAGE 44

44 instructor in two styles: first, where the instructor was situated right next to the slide and seamlessly interacted with the content and second, where the instructor appeared in a window in a fixed portion of the presentation window. The results showed that learners preferred to watch videos in the mode where the instructor seamlessly interacted with the course content. It was suggested that the video integrating the instructor seamlessly offered access to the instructor's eye gaze and ges tures in close proximity to the lecture content that resulted in a better learning experience for the learners via the availability of more realistic social cues (Bhat et al., 2015) Intermittent vs. permanent i nstructo r presence I t has been argued the instructor can be strategically presented to keep the non verbal social cues while not causing a significant amount of distraction ( Kizilcec, Bailenson, & Gomez, 2015) Kizilcec and colleagues (2015) conducted a 10 week field experience where they compared the permanent presentation of the instructor with the intermittent presentation of the instructor (where the ins tructor was not always shown on the screen). The results revealed learning did not differ between the permanent and intermittent presentation conditions. In fact, t he perceived mental effort was higher in the intermittent presentation condition than in the permanent presentation condition. Learner Cognitive Differences The previous studies all examined how the influence of instructor presence can be moderated by the video and its design. It is likely that the mixed findings of instructor ce may be explained by learner variabilities. So far, a limited number of stud ies that ha ve examined the effects of individual differences o n learning in the context of video with instructor presence. Considering the fact learners in general and

PAGE 45

45 undergradu ate students represent a variety of cognitive and attenti onal differences (Jonassen & Grabowski, 2012) and individual differences possibly moderate the influence of instructor presence in video it is important to examine the moderating effects of individual differences on the influence of instructor presence. Decades ago, r esearch within the aptitude tr eatment interaction (ATI) tradition revealed strong interactions between individual learner characteristics and instructional interventions (Cronbach & Snow, 1977; Snow, 1989) So, it is highly possible that individual differences may lead to different patterns of interaction and learning within videos that integrate instructor presence. There are very few studies on how individual differences among learners moderate the effect of the instructor presence in instructional video ( Kizilcec et al., 2014; Chen & Wu, 2015 both focus ing on learning styles). Individual differences in working memory capacity and i nhibitory control are theoretically important moderators of visual attention, cognition, and learning in this context. P icture in picture instructor video and the learning content in the rest of the frame in an instructional video represent two potentially competing sources of information on th e screen and they can control and working memory process es So, t he variability in these two individual differences will likely moderate tion to the stimuli on the screen their cognitive dynamics, and ultimately their learning of the content, especially when the content is difficult, imposing a high level of intrinsic load

PAGE 46

46 Working memory c apacity One relevant individual cognitive diff erence that could possibly moderate the effects of instructor presence is working memory capacity. Working memory capacity (Conway & Engle, 1994) maintain ing relevant information and ignoring irrelevant information (e.g., salient sensory stimuli). It has been found that individuals with high working memory capacity are better at directing attention when they are encoding information from the environment (Bleckley, Durso, Crutchfield, Engle, & Khanna, 2003; Kane, Conway, Bleckley, & Engle, 2001) In other words, people with high working memory capacity are better at utilizing executive control to focus on relevant stimuli and ignore distractions in th e environment (Sanchez & Wiley, 2006). Several studies have examine d the role of WMC related to multimedia learning conditions and these studies focused on the influence of seductive details in multimedia learning (Sanchez & Wiley, 2006) learning from animation with cueing (Skuballa, Schwonke, & Renkl, 2012) learning underlying conceptual relationships across mult iple documents (Banas & Sanchez, 2012) learning from s crolling and paginated tex t (Sanchez & Wiley, 2009) and so on. Sanchez & Wiley (2006) examined the role of WMC in relation to the seductive details effect which was operationalized as intriguing but irrelevant information in the form of illustrations The results indicated that as expected, learners with high WMC were less suscep tible to seductive information and performed b etter than low WMC groups in the seductive details condition on the inference verification and argumentative essay tasks. Interestingly, t hey also found that high WMC individ uals performed best under the seductive illustration condition

PAGE 47

47 compared to conditi ons which provided conceptually relevant illustration or no illust ration at all. The eye movement analysis indicated that low WMC learners allocated attention to seductive details more often and for a longer duration than high WMC learners and they perform ed more poorly on the tests of learning than the ir high WMC counterparts In another study by Skuballa and colleagues (2012) learners with high and low WMC w ere presented a narrated animation to learn how a par abolic trough power plant works. Participants were randomly assigned to one of the thr ee experimental conditions: one group supported by verb al instruction in the narration, one group supported by spotlight cues in the animation and one control group (no s upport). V erbal instruction and on screen important on screen information in the animation rning was measured b y open ended questions, following the Function Process Structure fram ework which describes knowledge about technical devices. The findings indicated the spotlight cues were impeded by the spotlight cues. The authors suggested that high WM C learners maintained their own control of attention to information and thus the spotlight cues became redundant and distracting resulting in expertise reversal These studies suggested the high WMC learners can better control their attention to relevant a nd important information in multimedia learning. The prior work all ows us to make hypotheses about how WMC might affect students process ing of instructor video and the rest of the screen in an instructional video. W hile attending to the instructor in the video, it is likely that learners with high WMC will overcome the split attention effect caused by the instructor video because

PAGE 48

48 they can manage their attention more effectively and allocate more cognitive resources to process instructional content in the rest of the frame. On the contrary, the amount of visual attention allocated to the instructor is expected to increase for participants with low WMC. This exacerbated split attention effect for learners with low WMC could be more detrimental to learning in a video that requires a higher level of cognitive processing (e.g., a video on a difficult topic with higher intrinsic complexity ). This hypothesis has yet to be tested empirically. Working memory capacity can be measured via complex span tas k s which dif fer from simple span task s in that they require simultaneous storage and manipulation of information (Bayliss, Jarrold, Gunn, & Baddeley, 2003) For example, in the Operation Span t ask learners need to verify the correctness of math operation s while memorizing letters. Commonly used complex span tasks include Operation Span (Turner & Engle, 1989) Reading Span (Daneman & Carpenter, 1980) and Counting Span (Case, Kurland, & Goldberg, 1982) The Operation Span t ask has been used widely in previous studies to measure working memory capacity (Conway et al., 2005) and has been shown to be a reliable measure to assess individual differences in WMC (Unsworth et al., 2005) The Operation Span task could predict verbal abilities and reading comprehension even though the subjects were solving mathematical problems. Engle and colleagues have argued that this implies a general pool of resources that is used in every type of working memory processing situation. As the proposed study examines learning from video s on Statistics topics which involve mathematical operations, the Operation Span t ask appears more appropriate in this context. This study will employ t he Automated Operation Span task (AOSPAN, Unsworth, Heitz, Schrock, & Engle, 2005)

PAGE 49

49 which is a computerized version of the original Operation Span t ask (Turner & Engle, 1989) Inhibitory c ontrol When discussing the ability to control attention during media rich information processing, another relevant individual difference is inhibitory control. Inhibitory control is the ability to selectively attend and suppress attention to irrelevant stimuli while focusing on the task goal ( Rothb art & Posner, 1985 ; Diamond, 2013) Empirically, Ophir and colleagues (2009) found that younger adults who are heavy media multitaskers (cf., Prensky, 2001) perform worse o n the inhibitory control test compared to light media multitaskers Heavy media multitaskers were found to be more susceptible to interference from irrelevant environm ental stimuli. In a recent study Homer and Plass (2014) investigated the interaction b etween instructional format and among high school students In this a color word was displayed either in a congr Participants were instructed to choose a color from a list of color words that matched the color of the displayed word. In incongr uent trials of the Stroop task, learners reacted slower and committed more errors as they had to inhibit their attention to the meaning of the color words and foc us their attention on the color attribute only (MacLeod, 1991). After degree of inhibition was measured, they were assigned to learn with a web based simulation of a n intrinsically complex topic (i.e., ideal gas law) that varied in instructional format (exploratory simulation or worked examples). In the simulation with

PAGE 50

50 worked examples learners navigated the simulation by following step by step instructions provided by an expert. Learning was measured by comprehension and transfer. Results indicated that the exploratory simulation facilitated transfer for students with higher levels o f inhibitory control whereas students with lower levels of inhibitory control benefited more from the guided simulation with worked examples in learning transfer It was suggested the participants with high inhibitory control use d their cognitive resources more efficiently thus resulting in enhanced learning with the exploratory simulation ; whereas the intensity of interaction in this condition may increase cognitive load for p articipants with low inhibitory control and poorer control of attention The find ings implied this negative effect may possibly be mitigated or reversed in the worked example simulation for this group of learners. Thus, i t is reasonable to hypothesize that inhibitory control is an important individual difference variable mediating col processing of information in an instructional video Instructional video i nclude s multiple sources of stimuli and requires inhibitory control to certain elements on the screen at some time. I t is possible that the visual of the instructor be comes irrelevant when attention should be focused on the instructional content in the main frame and the variable ability of the learners to block out irrelevant (or, rather less relevant) stimuli should play an important role in the efficient use of cogni tive resources in such a learning situation This hypothesis has yet to be tested empirically. Inhibitory control ca n be measured using the Flanker t est (Eriksen & Eriksen, 1974) Attention Network T est (Fan et al., 2002) Stroop T ask (MacLeod, 1991) G o/ N o Go tasks (Cragg & Nation, 2008) and S top S ignal tasks (Verbruggen & Logan, 2008)

PAGE 51

51 For the purpose of the proposed study, Flanker t est (via NIH toolbox) will be administered This measure ha s been used in multiple studies (Christ, Holt, White, & Green, 2007; Poarch & van Hell, 2012) to inhibitory control and has been validated to be a s ensitive measure of variations in inhibitory control (Zelazo et al., 2013) Me asurement Considerations Product Measures Prior research has relied on traditional product measures to identify the influence of i nstructor presence in videos. These product measures include learning outcome and self report measures, such as cognitive lo ad. Learning Retention and transfer have been two typical product measures of learning R etention tests measure ze information from the learning material (Tulving, 1968) Recall questions are typically presented as either cued recall where cues are provided to facilitate retrieval of information (e.g., fill in the blank questions) or as free recall where learners are asked to recall everything th ey remember from the lesson or learning materials without any cues to aid with their recall Previous studies in the context of instructor presence in videos employed r etention measure s in the form of both recall and recognition to test learners comprehen sion of key concepts covered in instructional video (Homer et al., 2008; Kizilcec et al., 2014; Pi & Hong, 2016) On the other hand, transfer measure s test ability to apply acquired knowledge to novel situations (Bransford & Schwartz, 1999) Transfer of knowledge is differentiated into far transfer or near transfer. Near transfer involves

PAGE 52

52 apply ing what was learned to a context that is closely related to the initial learning situation F ar transfer, however, refers to the ability to apply kno wledge acquired in a novel context that is very different from the original one (Mestre, 2002) Several studies adopted t ransfer test s to measure instructional video to novel situations (Homer et al., 2008; Pi & Hong, 2016) T he current study will address both retention and transfer aspects of learning. R etention items will include cued recall and recognition questions to ability to recall and recognize inf ormation presented in the video Near and far transfer questions will be used to measure apply the information from the two videos to new situa tions that are either close to or distinct from the situations covered in the video s Learner perceptions Perceived cognitive l oad Cognitive load has been traditionally assessed using subjective measures (Paas, T uovinen, Tabbers, & Van Gerven, 2003) relying on the assumption that learners can introspect on the amount of mental effort they expend on a task (Antonenko, Paas, Grabner, & van Gog, 2010). For example, National Aeronautics and Space Administration Task Load Index (NASA TLX) has been used to TLX includes ten workload related factors and consists of six subscales, including mental demand, physical demand, temporal demand, frustration, effort, and performance, which represent cognitive load in task performance (Hart & Staveland, 1988) Later Paas (1992) developed a one item 9 point rating scale, which allows learners to self report on w

PAGE 53

53 validated in many prior studies on instructor presence as a subjective measure of cognitive load (Kizilcec et al., 201 4; Wang & Antonenko, 2017) In the proposed study, the researcher will adopt this widely used subjective measure of cognitive load to measure overall load. Besides overall cognitive load, the proposed study will also measure intrinsic load, extraneous lo ad, and germane load as reflected by subjective measures of learner perceptions To measure intrinsic load, Ayres (2006) scale will be used. It asks participants to rate th e difficulty of the content they perceive in each video. To measure extraneous load, the researcher will use the scale from Cierniak (2009) and ask participants to rate how difficult it is to learn the material. To measure germane load, Salomon's (1984 ) question of how much learners concentrated during each video will be asked. At the end of each video, participants will respond to four questions assessing in trinsic load, extraneous load, and germane load). Perceived learning, satisfaction, and social presence As instructor presence in instructional video has been hypothesized to provide socio emotional cues and elicit social responses the researcher will e xamine the influence of instructor presence on satisfaction and social presence in the proposed study. For perceived learning, participants will self repo r t the amount of learning they acquire from the video on a Likert s cale For satisfaction, participants will be asked to rate their level of satisfaction with instructor presence on a Likert scale as well To measure the degree of soc ial presence with regard to the presence or absence of the instructor in the

PAGE 54

54 video, parti cipants will indicate the level of agreement wi th five statements that examine the degree of social presence perceived (Kizilcec et al., 2015) Process Measure s One limitation of previous research i n multimedia learning is the almost sole reliance on product measures (Sanchez & Wiley, 2009) While these measures may be useful and efficient in gauging global performance, they may not be sufficient to study the complex attentional and cognitive process ing that occurs during multimedia learning. Using the product measures, educational researcher s obtain little insight into the underpinning process es students engage in while they watch video and it is usually not very clear, at least from the empirical standpoint, why learners end up with a particular score on the learning test or why they report a high level of cognitive load As the current study adopts an individual differences perspective, it is especially important visual attention distribution and cognitive dynamics while they watch an instructional video with or without instructor presence Visual attention di stribution will be measured using eye tracking, and cognitive dynamics will be assessed using novel EEG methodology. The complementary use of EEG and eye tracking c a n reveal how learners with individual differences processed the videos with or without instructor presence learning easy and difficult topics. Visual attention distribution As learners with individual differences in working memory capacity and inhibitory control are hypothesized to display d ifferent patterns of interaction with the visual stimuli on the screen ( i.e., instructor frame and other on screen visual content, i t is important to

PAGE 55

55 understand the visual attention distribution exhibited by learners w ith individual differences in working memory capacity and inhibitory control. The eye mind hypothesis (Just & Carpenter, 1980) suggests that eye movement recordings can reveal where the processing E ye tracking can be used to understand the dynamics of visual attention during multimedia learning (e.g., Mayer, 2010) Visual attention distribution is typically inferred using gaze fixations and saccades. Fixation occurs when the eye focuses on a visual target for a short period of time (around 300 ms). Fixation duration is longer on a difficult or unfamiliar word during reading (Rayner, 1998) An aspect of problem that requires more cognitive processing will receive more and l onger fixations (Carpenter & Shah, 1998; Duchowski, 2007) Saccade is a rapid eye movement between two fixations and saccades range in amplitude from small movements to large ones. Empirically, eye tracking method ology has been used to explore the attentional and cognitive process during learning with multimedia Some of t hese studies helped determine that (a) a strong link exists between eye fixa tions and learning outcomes ( Boucheix & Lowe, 2010 (Boucheix & Lowe, 2010; de Koning et al., 2010) ; (c) prior knowledge guides visual attention (Canham & Hegarty, 2010; Jarodzka, Scheiter, Gerjets, & Gog, 2010) ; and (d) learners who view animation and on screen text must split their attention between graphics and printed words (Schmidt Weigand et al., 2010) Eye tracking has been shown to be a useful tool to study visual attention distribution (van Gog & Scheite r, 2010) and it is very suited to study differences in attentional processes evoked by different types of

PAGE 56

56 multimedia (Holsanova, Holmberg, & Holmqvist, 2009) Compared to tradition al outcome measures, eye tracking is a process measure that can shed light on the mechanism of split attention that occurs when attention is divided between the instructor and the rest of learning content in the video frame (Johnson, Ozogul Moreno, & Reisslein, 2013; Schmidt Weigand et al., 2010) So far, few empirical studies have used eye tracking to study the role of instructor presence in instructional videos to shed light on the mechanisms by which learners interact ed with this instr uctional medium. Kizilcec and colleagues (2014) investigated how adding the instructor to video instruction affect ed the distribution of visual attention, time looking at the instructor and switched between the instructor and slide every 3.7 seconds. In another study where an instructor was added to P slides ( Pi et al., 2017) it was found that the participants spent significantly more time fixating on the instructor (62.3%) than on the PPT slides (37.7%). In addition, t he mean fixation duration in the area of the instructor was shorter than in that of the PPT slides. Similar findings were reported in a context where human instructor was replaced by an animated pedagogical agent. Louwerse and colleagues (2008) found that even when pedagogical agents only make up around one fourth of the display, participants contributed 56% of visual attention to th e agent For this study, eye tracking will help understand the visual attention distribution that occurs as learners with individual differences in working memory capacity and inhibitory control learn from a video with and without instructor presence. In the study, the portion of the frame showing the instructor will be defined as one interest area and

PAGE 57

57 attention distribution will ber of fixations, average fixation duration, and dwell time in these two interest areas. A s visual attention is hypothesized to be split between the instructor and the content when the instructor is present, the number of transition s (saccades) between the two interest areas will also be calculated to infer this split attention effect caused by instructor presence. Cognitive dynamics It has been suggested that self reported measures of cognitive load are confounded (e.g., it is difficult to separate the ef fects of cognitive load from fatigue) and relatively insensitive to variations in cognitive load over time (Xie & Salvendy, 2000). The use of objective online measures of cognitive dynamics can provide greater insight into the cognitive load fluctuations t hat occur at different stages during a task, that is otherwise not evident using simple self report s of cognitive load (Antonenko, Paas, Graber, & van Gog, 2010) Among these objective online measures of cognitive load, neuroimaging tools can be used to pr ovide insights into the underlying cognitive dynamics while learners watch the instructional video s. Electroencephalography (EEG), compared to other neuroimaging techniques (e.g., fNIRS), can be used as a measure of cognitive dynamics that yields high temp oral resolution (Antonenko, van Gog, & Paas, 2014 ; Antonenko & Keil, in press ) The high temporal resolution of EEG enables it to a millisecond scale. EEG has been widely adopted as a measure of process in educational research on reading, mathematics and problem solving (Antonenko et al., 2014) For example, EEG has gnitive process in

PAGE 58

58 a hypertext learning environment (Antonenko & Niederhaus er, 2010). The findings showed page previews led to a higher decrease of brain wave activity associated with extraneous cognitive processing, compared to the condition without page previews. However, few empirical studies have used EEG to examine learners cognitive process in the context of multimedia learning (e.g., multimedia learning between typically developing and gifted students under text; text, narration, and image; and text, narration and video conditions. The results indicated the typically developing and gifted students demonstrated different patterns of cortical activity while processing the multimedia presentations. Gifted students displayed higher alpha power (i.e., less mental activity) during all th ree formats of presentation and the difference was most pronounced in the text, narration and video condition. The authors suggested that the strength of mental activity under cognitive load was negatively associated with intelligence. So far, few empiric al studies have used EEG to study the cognitive dynamics during learning with instructional video Chen and Wu (2015) used EEG to measure sustained attention while participants were placed in one of the three conditions of instructor presence in video (ins over lecture capture and picture in picture types of videos). contained three windows for the slides in the lower left p the screen. Using NeuroSky's EEG headset and adopting its algorithms for calculating the level of sustained attention t he researchers found that sustained attention induced

PAGE 59

59 by the voice over ty pe is markedly higher than that of the picture in picture type. It was suggested that the higher sustained attention was related to increased cognitive load in ion was split by the instructor fra me, the table of content, and the slides. The process of integrating information from thus resulted in increased cognitive load. Using EEG, Diaz and colle a gues (2015) examined the effect of instructor presence on cognitive load and emotional states. Each participant was exposed to three video conditions: instructor always present, only (mixed condition). They found a significantly higher event related desynchronizati on in alpha band power (a proxy of cognitive load ) in the mixed condition. It was suggested as the instructor disappeared after initial appearance on the screen in the mixed condition, participants had to look for a new media for attending the presentation thus causing the increment in cognitive load In the current study, EEG can noninvasively provide insights into the underlying cognitive dynamics that occur during the learning process, and thus inform the influence of instructor presence on learners wi th individual differences, that is otherwise not evident comparing learning test scores or self report responses.

PAGE 60

60 CHAPTER 3 METHODOLOGY T he current study is designed to explore the following two research questions. Question 1. To what extent does instru ctor presence in easy and difficult instructional videos influence learning, visual attention distribution, cognitive dynamics, and learner perceptions (cognitive load, perceived learning, satisfaction, situational interest social presence, and perception s of the instructor)? Question 2. How do individual differences in working memory capacity and inhibitory control moderate instructor presence effects in easy and difficult instructional videos? Research Design Given the theoretical perspectives and empiri cal evidence on instructor presence in instructional video s t he purpose of the current study is to explore how instructor presence in videos influences learning, visual attention distribution, cognitive dynamics, and learner perceptions and how these effe cts are moderated by individual differences in working memory capacity and inhibitory control. This study manipulated video design that differentiates content difficulty (easy topic vs. difficult topic) as a within subjects v ariable and instructor presence (instructor present vs. instructor absent) as a between subjects variable. Based on the results of working memory capacity and inhibitory control tests, participants were assigned to watch two videos in one of the two conditions: 1) the easy topic video w ith instructor present and the difficult topic video with instructor absent ; 2) the difficult topic video with instructor present and the easy topic video with instructor absent. This way, the balancing of the groups could be controlled to include both low WMC and low IC vs. high WMC and high IC in the sample for each condition (compared to random assignment). Half of the participants watched the easy topic video with instructor

PAGE 61

61 present and the difficult topic video with instructor absent. The other half of the participants watched the easy topic video with instructor absent and the difficult topic video with instructor present (Figure 3 1) To avoid the order effect participants watched the easy topic and the difficult topic videos in randomized order The variables in the current study are summarized in Table 3 1. Participants Using a priori power analysis for multiple linear regression (G Power) with an alpha level at = .05, an estimated medium effect size, three predictor variables (i.e., instructor presence, working memory capacity and inhibitory control), and a desired power of 0.8, the study required approximately 77 participants (Cohen, 1992, p. 158) Given the challenge s of recruit ing this many freshman students and using complicated data collection methods like E lectroencephalography (E EG ) and eye tracking, this study recruit ed 60 participants from the University of Florida The possibility was high that t hese participants would watch these instructional videos for their Statistics courses at UF. As the topics in the two videos are not covered in high schoo l Advanced Placement ( AP ) statistics, the recruited freshman students were presumed to have no prior knowledge on the topics selected for the current study Inclusion criteria were as follows: (a) all participants should be between 18 and 27 years old, (b) a ll participants should have normal or corrected to normal vision (c) participants should have no history of brain trauma or n eurological disorders (d) c andidates for the study cannot be using depression and anxiety medications as the y alter brain activ ity e) c andidates with Autism Spectrum Disorder were exclude d from the study as they respond to face differently. A pre screening instrument include d these questions.

PAGE 62

62 Materials The two videos cover ed an easy and a difficult topic in Statistics as dete rmined by Statistics Education experts Statistics is important to college students, as it can academic or professional discipline (Gal, Ginsburg, & Schau, 1997, p. 39) The easy topic video focu sed on the terminology associated with experiments and observational studies. The easy topic video covered information on the definition s of experiment, observational study, explanatory variable, response variable, factor, levels, treatments, subjects and replication. The details of the content covered in the easy topic are provided in Appendix A. The easy topic video last ed approximately 3 minutes. The difficult topic focuse d on the rationale for conducting the Analysis of V ariance (ANOVA) The instructor explained concepts such as the purpose of ANOVA, null hypothesis, alternative hypothesis, between group variance, within group variance, and F test statistics. The details of the content covered in the difficult topic are provided in Appendix B The diffic ult topic video last ed 4 minutes and 30 seconds. Both videos were within six minutes long which has been suggested as a median engagement length for online instructional videos ( Guo et al., 2014) Both the easy topic and difficult topic videos were designed in two versions: instructor present or instructor absent. The main frame of each video was devoted to a Khan Academy style pencast, whereas for the instructor present videos, the bottom right hand corner alway s showed a shoulder up view of the instructor. The in structor frame showed the instructor e xplaining the content and displaying non verbal cues (eye gaze, gesturing, and facial expressions). The same instruct or (a middle aged white male) was present in the easy and difficult topic videos with instructor pres ence and the

PAGE 63

63 amount of his non verbal cues (eye gaze, gest uring, and facial expressions) wa s kept consistent across these two videos. The instructor narrated the same script for the instructor absent videos as he did in the instructor present videos. In si tuations where pencast was produced on the screen, the instructor wa s writing with a purple pen on the screen in the two instructor prese nt videos whereas the pencast was produced using a tablet pen in the instructor absent videos without showing the hand of the instructor as students c ould be wondering where the hand came from. Screenshot s of the videos with and without instructor presence for the easy topic are provided in Figure 3 2 Screenshots of the videos with and without instructor presence for the difficult topic are provided in Figure 3 3 The two videos were displayed using Experiment Builder software (SR Research, Ontari o, Canada). The software captur ed is compatible with Data Viewer, the eye movement data anal ysis software from the same company (SR Research, Ontario, Canada). Apparatus Participants were seated in a comfortable chair approximately 55 cm from the computer screen in a laboratory with controlled lighting. The instructional video with or without i nstructor presence was displayed on an external 20 inch flat panel monitor, with a resolution of 1 600 by 9 00 pixels and a refresh rate of 60 Hz. Participants use d a chinrest (SR HDR) with a forehead bar to minimize head movement. Eye movement data were col lected with Eyelink 1000 Plus system (SR Research, Ontario, Canada) using a desktop mount at a sampling rate of 1000 Hz in the monocular mode EEG data were collected using wireless, dry electrode DSI 24 EEG system (Wearab le Sensing,

PAGE 64

64 California, USA) at a sampling rate of 300 Hz Figure 3 4 illustrates the experimental setup. Measures The study employ ed multiple measures First, pre intervention measures were used to collect pre screening data ( using an online survey), and individual differences data usin g a test of working memory capacity (i.e., Automated Operation Span t ask), and a test of inhibitory control (i.e., NIH Flanker In hibitory Control and Attention t est). Second, process measures of a) visual attention distribution ( interest area fixation and transition measures) and b) EEG based cognitive dynamics (e.g., alpha and theta power ) were used during the intervention. Finally, several product measures were used upon the completion of the intervention: a) a learning test of retention and transfer and b) a survey of learner perceptions (i.e., cognitive load, perceived learning, satisfaction, situational interest social presence, and per ceptions of the instructor). Details on each measure are provided below. Pre Intervention Measures Pre screening s urv ey Before participants watch ed the two instr uctional videos, they respond ed to a brief online pre screening survey (Appendix C) providing information on their age, gender, ethnicity, first language, major (if declared), familiarity with research design t erminology and ANOVA The p re screening survey also include d items regarding participant vision, history of neurological disorders and medication use. Inhibitory c ontrol Inhibitory control was measured using the Flanker In hibitory Control and Attention t est in NIH Toolbox (Akshoomoff et al., 2013) iPad application The test

PAGE 65

65 required the participant to focus on a given visual stimulus while inhibiting attention to stim uli (arrows) flanking it. Sometimes the middle stimulus was pointing in the same (incongruent). Participants were asked to press either the left key or the right key on the iPa d TM keyboard that corresponded to the direction in which the middle arrow was pointing. Scoring was based on a combination of accuracy and reaction time. Figure 3 5 shows an incongruent example of the Flanker Inhibi tory Control and Attention t est. This mea sure has been shown to have excellent reliability and validity and appears sensitive to variations in inhibitory control (Zelazo et al., 2013) Working memory c apacity Working memory capacity w as measured by the Au tomated Operation Span t ask (Unsworth et al., 2005) An alternative test is a rea ding span test but because participants in this study would learn s tatistical concepts that involve mathematics in the easy and dif ficult topic video, the Operation Span t ask was used. The Automated O peration S pan task has been shown to be a reliable and v alid indicator of WMC (Unsworth et al., 2005) The test was cond ucted online at http://www.millisecond.com/download/library/ospan/ In the Oper ation Span t ask, participants were h operation (e.g., F). After a series of math operations and letters had been presented, participants were asked to recall the letters that follow each math operation. The number of operation letter strings in a sequence (i.e., set size) were increased and decreased to measure the participant's operation span. Figure 3 6 shows an example of the Automated Operation Span t ask.

PAGE 66

66 Measures Used During Intervention Visual attention d istribution In order to examine the influence of instructor presence on visual a ttention distribution in the instruct ional videos, this study defined the portion of the frame showing the instructor in each video as the instructor interest area and the rest of the frame including images and pencast as the content interest area. Visual attention distribution was duration, and dwell time on the two interest areas. As visual attention is hypothesized to split between the instructor frame and the rest of the screen wh en the instructor is present, the number of transitions between fixating on the instructor interest area and the content interest area was calculated to examine the potential for and magnitude of split attention caused by the instructor presence in the vid eo. Cognitive d ynamics While participants watch ed the two videos, the DSI 24 EEG system record ed EEG data in the following four frequency bands: delta (0 4 Hz), theta (3 7 Hz), alpha (8 12 Hz), and beta (13 30 Hz) (Basar, 1999) EEG was recorded at a sampling ra te of 300 Hz for each of the electrodes located at Fp1, Fp2, F3, Fz, F4, F7, F8, P3, Pz, P4, T3, T4, T5, T6, C3, Cz, C4, O1 and O2, of the 10 20 international system ( Figure 3 7 Jasper, 1958) with respect to mastoid electrodes (A1 and A2) in common reference. Post Intervention Measures Learning t est After participants finish ed watching the two videos, their learning from the videos was assessed by a learning test that include d 12 retention questions and 5 transfer qu estions for the easy topic and 8 retention questions and 5 transfer questions for the

PAGE 67

67 difficult topic. Thus, each participant complete d a total o f 30 learning test items. The difference in the number of retention questions for the two videos was due to the fact that the difficult video was slightly lo nger. The easy topic video lasted approximately 3 minutes whereas the difficult topic video last ed 4 minutes. For the video on terminology associated with experiments and observational studies (easy topic), the retentio n and transfer questions assess ed definition of experiment, observational study, explanatory variable, response variable, factor, levels treatments, subjects and replication. For the video on rationale of the ANOVA (difficult topic), the retention and transfer questions assess ed purpose of ANOVA, null hypothesis, alternative hypothesis, between group variance, within group variance, and F test statistics. The retention questions were recognition based question s, which assess ed retention question for the difficult topic of rationale for The purpose of running a one factor ANOVA on three groups is to _________. A. Compare the means between three groups; B. Find the similarities between three groups; C. Characterize the overlapped portions between three groups; D. Find which two groups have different means; The transfer questions provide d new scenarios that were not pr ovided in the videos and examine d video in to a novel context An example transfer question for the easy t opic of terminology associated with experim ents and observational studies wa researcher carried out an experimental study to examine the effectiveness of GRE prep programs (Powerscore or Kaplan) and the method of delivery (in person,

PAGE 68

68 online, or blended ) on GRE verbal score. In the study, there are ____ factors. A. 1; B. The retention and transfer questions were both multiple choice questions. The learning test items were reviewed and optimized by three experts (e.g., fa culty members and graduate students) in Statistics E ducation to ensure content validity. In order to reduce variance from guess work participants were they did not know the correct answer. No time limit was imposed on pa rticipants for completing the learning test. The learning test was delivered using Qualtrics The learning test questions for the two t opics are provided in Appendix D Learner p erceptions Several questions were asked to elicit of the videos with and without instru ctor presence, and these include d questions on the perceived cognitive load, perceived learning, satisfaction, situational interest social presence, and p erceptions of the instructor. It is important to examine learnin g effects but it is also necessary to understand how the intervention influenced which could be reflected in the constructs such as perceived learning, satisfaction, and social presence. The perceptions q uestions were deliver ed using The questions included in the learner perception survey are provided in Appendix E Cognitive l oad. Immediately after viewing each video with or without instructor presence, participants self report ed four well recognized types of cogn itive load (Leppink et al., 2013) they experience d while watching the video Overall load (Paas, 1992) Participants rate d the perceived amount of mental effort they in vested while watching the video, using a 9 point Likert scale that rang e d from very, very low mental effort (1) to very, very high mental effort (9).

PAGE 69

69 Intrinsic load (Ayres, 2006) Participants rate d the how easy the topic covered in the video was, using a 9 point Likert scale t hat range d from very, very easy (1) to very, very difficult (9). Extraneous load (Cierniak, 2009) Participants rate d how easy it was to learn from the v ideo using a 9 point Likert scale that range d from very, very easy (1) to very, very difficult (9). Germane load ( Salomon, 1984) Particip ants rate d how concentrated they were while they watch ed the video, using a 9 point Likert scale that range d from very, very little (1) to very, very much (9). Perceived learning After reporting the four types of cognitive load, participants indicate d ho w much they learn ed from each video, using a 9 point Likert scale that range d from did not learn anything (1) and learned a great deal (9 ). This measure has been used in a related previous study (Wang & Antonenko, 2017) and it appears sensitive Satisfaction After reporting perceived learning, participants rate d their satisfaction regarding learning with each video, using a 9 point Likert scale that range d from extremely dissatisfied (1) to extremely satisfied (9). This measure has been used in a related previous study (Wang & Antonenko, 2017) and it appears sensitive to S ituational interest Participants respond ed to a question (Lathrop, 2011) that examine d the effect s of each video on increasing situational interest with or without instructor presence on a five point scale: I am willing to watch more video s like this because it is exciting and relevant. ( Strongly di sagree ; Disagree ; Neither agree nor

PAGE 70

70 disagree ; Agree ; Strongly Agree ). This measure has been reported sensitive to situational interest with regard to learning environment stimulation (Lathrop, 2011) Social p resence Afterwards, for each video participants indicate d the level of agreement with five statements that examine social presence (Kizilcec et al., 2015) on a five point scale ( Strongly disag ree ; Disagree ; Neither agree nor disagree ; Agree ; Strongly Agree ). The statements include d : I felt like the instructor was in the same room as me ; I felt that the instructor was very detach ed in his interactions with me; I felt that the instr ucto r was aware of my presence; I felt th at the instructor was present; I felt that the instructor remained focused on me throughout our interaction. This measure has been social pr esence (Kizilcec et al., 2015) Perceptions of the instructor. Each participant watch ed one video with instructor presence and the other video without instructor presence. A t the end of the instructo r present video, participants respond ed to two additional questions to provide more information regarding their perceptions of the instructor. They first respond ed to an open ended question: Please explain what you think about seeing the instructor in the video, compared to not seeing the instructor. Then, participants were asked to provide feedback on their perceptions of instructor presence by sele cting all adjectives that characterize d were also allowed to write down other adjectives that were not provided on the list. This measure has been reported sensitive

PAGE 71

71 Da ta Collection Procedures Potential particip ants were approached and they complete d the online pre screening survey and the researcher decide d if they would be invited to participate in the experiment. I f the participants did not qualify after pre screeni ng, they were thanked and were not invited to participate in the study. The qualified participants were invited to participate in the study. All testing took place at the Neuroscience Applications for Learning (NeurAL) Laboratory located at Norman Hall G52 5 in the UF College of Education. After signing the informed consent (Appendix F ), memory capacity was assessed by the Automated Operation Span task. After the working memory capacity test, p was assessed by the Flanker In hibitory Control and Attention t est Afterwards, participants were fitted with the DSI 24 EEG headset (Wearable Sensing, California, USA). The DSI 24 EEG headset did n ot require any additional skin preparation or applying conduct ive gel to the scalp or electrodes. The EEG headset was non invasive and had been used in multiple research studies in higher education. After the headset was donned appropriately, participants were asked to perform several simple tasks such as rapid eye b links, coughing, and clenching jaw, to check if the EEG responses were accurate. After that, the gaze of each participant was point calibration algorithm. Then, participants watch ed the videos on the two topi cs given the experimental condition they were assigned to Participants watch ed the video s without pauses and were not allowed to take notes while watching the v ideos. While participants watch ed the videos with or without instructor presence, simultaneous EEG data and eye movement data were collected to capture their cognitive dynamics and

PAGE 72

72 visual attention distribution respectively. Immediately after watching each instructional video, participants report ed four types of cognitive load, perceived learning, s atisfaction, situational interest social presence for the video they just watched. For the video with instructor presence participants also respond to two additional questions eliciting about their perceptions of the instructor After watching both video s, participants complete d a learning test that mea sured retention and transfer of knowledge from the two videos. The entire data collect ion session took approximately 6 0 minutes. The task s for each phase of the study and the approximate length for each tas k were summarized in Table 3 2. Data Cleaning and Scoring Procedures Inhibitory C ontrol Test D ata The NIH Toolbox calculate d one 2 vector score (reaction time and accuracy) and a total normed score (fully corrected for age and other demographic variables ) Flanker In hibitory Control and Attention t est. For data analysis, the fully corrected score was used to represent each Working M emory C apacity Test D ata Th e OSPAN score was based on the traditional "absolute OSPAN" scoring method, which produce d the sum of all perfectly recalled sets. For example, if a participant recalled correctly 2 letters in a set size of 2 (i.e., 2 math operations to process and 2 lette rs to recall), 3 letters in a set size of 3, and 3 letters in a set size of 4, then would be 5 (2 + 3 + 0). This OSPAN score was used to represent

PAGE 73

73 Cognitive Dynamics Data Custom scr ipts in EEGLAB (Delorme & Makeig, 2004) were written to decontaminate EEG data and pre process EEG data. The researcher adopt ed Independent Component Analysis ( ICA) approach to decontaminate EEG artifacts due to body movement and eye blinks. Wavelet analysis was conducted in EEGLAB using custom scripts to infer the cognitive dynamics while participants watch ed the two videos and the resulting time frequency repre sentations were achieved for each participant. This method had been used in multiple cognitive neuroscience studies focusing on the mechanisms of attention (Keil et al., 2003) and cognition (Keil et al., 2001) Visual Attention Distribution Data The eye movement data ( interest area fixation s and transition s ) were processed in Data Viewer (SR Research, Ontario, Canada). Data Viewer export ed several variables in respect to the interest areas, including number of fixations, percentage of fixations, average fixation duration, dwell time and percentage of dwell time regarding the instructor interest area and the content interest area and transitions between the two interest areas. Learning Test Data Learning test scores for the easy and difficult topics were computed by awarding Each participant had four scores from the learning test: a) score for easy topic retention test ; b) score for easy topic transfer test ; c) score for difficult topic retention test ; d) score for difficult topic transfer test

PAGE 74

74 Learner Perception Data For me asures of cognitive load, perceived learning, satisfaction, situational interest and social presence, responses on the scale s were directly used. For responses to the question on the perceptions of the instructor (i.e., response to the open ended question ), responses were treated as qualitative data. For the characterization of feelings toward the instructor, frequency of selecting each adjective was examined for the easy topic and the difficult topic respectively. All the cleaned data were entered into S PSS. Table 3 3 shows an example of data format for two participants who watch the difficult topic with and without instructor presence respectively (shown in transposed manner due to page size limit). Data Analysis Analyses of variance (ANOVA) was used to address Research Question 1, that is, the effect s of instructor presence on dependent variables including learning, cognitive dynamics, visual attentio n distribution, and learner perceptions. Multiple regression was used to answer Research Question 2, tha t is, to examine how influences of individual differences moderate the effect s of instructor presence on these DVs. Step 1: Checking Assumptions for Statistical Tests First, assumptions for specific statistic models were checked. For ANOVA tests, assumpti ons of parametric tests (normality, independence, homogeneity of variance) were checked and if not met, non parametric tests (e.g., Mann Whitney test) was adopted as an alternative. For multiple regression tests, assumptions including independence, homogen eity, linearity, normality, and collinearity were checked. A

PAGE 75

75 criterion of .05 was used for determining the level of significance in all statistical tests. The data analysis was conducted in several steps. Step 2: Ex ploring Group Homogeneity To determine w hether the experimental groups are homogeneous relative to working memory capacity and inhibitory control one way MANOVA was performed across the experimental conditions. As the balancing of the groups was controlled to include both low WMC and low IC vs. high WMC and high IC in the sample for each condition, t he results should indicate that there is no significant difference across the experimental conditions on either of the two individual difference variables and the two experimental conditions should b WMC and IC scores. Step 3: Examine the Effect s of Instructor Presence First, ANOVA was performed to test the effects of instructor presence ( i ndependent v ariable) f or easy and difficult topics respectively on each of the f ollowing dependent variable s : Learning ( retention, transfer ); Cognitive d ynamics ( average alpha and theta power ); Visual a ttention d istribution ( number of fixations and dwell time in the two interest areas ) ; Self reported cognitive load (overall CL, int rinsic CL, extraneous CL, germane CL); Perceived learning; Satisfaction ; Situational interest ; Social presence

PAGE 76

76 Step 4 : Examine the Moderating Effect s of Individual Differences To examine how individual differences in WMC and IC moderate d the effect s of i nstructor presence on each dependent variable for the easy and difficult topics I utilize d a battery of multiple regression analyses with instructor presence variable ( p resent or a bsent), WMC score, and IC score as predictors of each of the following outc ome variable s: Learning (retention, transfer ); Cognitive dynamics ( average alpha and theta power ); Visual attention dis tribution (number of fixations and dwell time in each interest area; number of transitions between the two interest areas) ; Self report ed cognitive load (overall CL, intrinsic CL, extraneous CL, germane CL); Perceived learning; Satisfaction; Situational interest ; Social presence I ran regressions with WMC and IC variables separately with instructor presence variable (one regression invol ving instructor presence and WMC and the other one regression involves instructor presence and IC for each dependent variable ) The regression analyses were conducted for each topic respectively. H ierarch ical linear regression analysis was also conducted for the two IVs (instructor presence and WMC ). This model aim ed to identify the added variance of the DVs the interaction between main I V and moderator IVs (WMC, IC) could contribute in addition to the variance explained by the IVs alone.

PAGE 77

77 I will enter in structor presence and WMC into block 1, interaction between instructor presence and WMC into block 2. Results from the first block could indicate whether instructor presence or WMC has a significant influence on DVs. Results from the second block could ind icate whether the addition of the interaction term in the second block can account for a significant portion of additional variance. To further evaluate the influence of WMC in instructor present/absent condition, separate linear regressions were conducte d. If WMC is significantly related to any DVs, I could conclude a moderating effect of WMC in the instructor present/absent condition. Assuming in the instructor presence condition, there is a significant positive relationship between WMC and learning tran sfer I could conclude that instructor presence has a positive effect on transfer of learning, and this effect is most pronounced in participants with higher WMC Step 5 : Examine the Effect of Individual Differences on Attentional Dynamics A ll the foll ow ing analyses (steps 5 and 6 ) focus ed on the data from the instructor present videos. As only the instructor present videos would prod u ce data on number of fixations and dwell time on instructor interest area as well as number of transitions between the tw o interest areas (instructor interest area and content interest area ), I ran regressions analyses for the data from the instructor present videos, to examine the moderating effect of WMC and IC on sev eral dependent variab les (i.e., number of fixations and dwell time on instructor interest area, as well as number of transitions between the two interest areas) in instructor present videos. Similar procedure (as in step 4 ) were used here except only WMC and IC scores were be entered as IVs in the hierarchical linear regression model.

PAGE 78

78 Step 6 : Examine the Relationship between Attentional Dynamics and Other Dependent Variables Furthermore, I am interested in examining whether number of fixations, average fixation duration, and dwell time on instructor IA and num ber of transition s between two IAs can predict learning, learner perceptions, and cognitive dynamics in the instructor present condition. Multiple regression models with four IVs (i.e., number of fixations, average fixation duration, and dwell time on inst ructor IA and number of transitions between two IAs ) on learning, learner perception variables and cognitive dynamics were tested.

PAGE 79

79 Table 3 1 Variables in the current study Variable Definition Main IV Instructor present vs. in structor absent Moderator IV Working memory capacity Inhibitory control DV Learning (retention and transfer) Visual attention distribution (fixations and transitions) Cognitive dynamics (alpha and theta power) Perceptions (cognitive load, perce ived learning, satisfaction, situational interest social presence, and perceptions of the instructor) Table 3 2 Phases of the study Phases of Study Tasks Pre intervention Informed consent (2 min) Automated Operation Span task (10 min) Flanker Inhibitory Control and Attention test (5 min) EEG headset setup (3 min) Eye tracker setup (2 min) Intervention Cognitive dynamics (8 min) Visual attention distribution Post intervention Learner perceptions survey (5 mi n) Learning test (20 min)

PAGE 80

80 Table 3 3 Data format for difficult topic Variable Participant 1 (Condition: Instructor present) Participant 2 (Condition: Instructor absent) Instructor presence 1 0 OSPAN 40 70 Inhibitory control 20 30 Retention 6 8 Transfer 2 5 Overall load 5 7 Intrinsic load 5 7 Extraneous load 5 7 Germane load 5 7 Perceived learning 8 6 Satisfaction 7 6 S ituational interest 4 3 Social presence statement 1 2 2 Social presence statement 2 2 2 Social presence stateme nt 3 2 2 Social presence statement 4 2 2 Social presence statement 5 2 2 Average alpha power at parietal lobe 1.1 1.3 Average theta power at central lobe 1.3 1.5 Percentage of fixations on the instructor IA 20% 1% Percentage of dwell time on the in structor IA 25% 2% Percentage of fixations on the content IA 80% 99% Percentage of dwell time on the content IA 75% 98% Number of transitions between IAs 30 1

PAGE 81

81 Figure 3 1 Group assignment for the study. Figure 3 2 Screenshots of the videos on the easy topic: Terminology associated with E xperiments and O bservational S tudies. Figure 3 3 Screenshots of the videos on the difficult topic: Rationale for the Analysis of Variance (ANOVA).

PAGE 82

82 Figure 3 4 Experimental setup. Photo courtesy of author. Figure 3 5 E xample s of the Flanker Inhibitory Control and Attention test. In the upper example (congruent) the par ticipant is expected to press the right pointing key as the middle arrow in the array is poi nting right I n the bottom example (incongruent) the participant is expected to press the left pointing key as the middle arrow in the array is poi nting left.

PAGE 83

83 F igure 3 6 An example of the Automated Operation Span task In this example, the participant is expected to check if the answer to the math problem is correct (Figure a; b) memorize lette r R (Figure c), check if the answer to anot her math probl em is correct (not shown here), memorize letter S (not shown here) then check if the answer to another math probl em is correct (not shown here), memorize letter L (not shown here). At the end, the participant is expected to select letters R, S and L (Figure d). Figure 3 7 EEG 10 20 system

PAGE 84

84 CHAPTER 4 RESULTS This chapter presents the findings of the current study. I first reported the demographics information. Next, I presented the data analyses on the effects of instructor presence on learning, learner perceptions, visual attention distribution, and cognitive dynamics in the easy and difficult topic video conditions Then I provided the findings related to the predic tors of learning and learner perceptions in the two experimental conditions Finally I reported the findings relative to the moderating effects of individual differences in working memory capacity and inhibitory control in each condition Participants Demographics Demographics inf ormation for the participants is provided in Table 4 1. Out of the six ty participants, t wenty one participants were male, and 39 of the participants were female. The participants had an average age of 18.36 ( SD = 0.66). Twenty five participants identified themselves as White or Caucasian, and 20 were Hispanic or Latino, 7 were Black or African American, and 8 were Asian or Pacific Islander. Influence s of Instructor Presence on Learning, Learner Perceptions, Visual Attention Distribution, a nd Cognitive Dynam ics I explored the influence of instructor presence on learning, learner perceptions, visual attention distribution, and cognitive dynamics. In the following section, I will discuss the influence of instructor presence on each of these dependent variable s Learning In the current study, the influence of instructor presence on learning was examined using retention and transfer measures.

PAGE 85

85 Retention of information from the videos was measured for both topic s using a test of 12 quest ions for the easy topic and 8 questions for the difficult topic Table 4 2 shows the descriptive statistics for the retention test scores when instructor was present or absent in the videos ANOVA test indicated that instructor presence did not influence r etention positively or negatively for either the easy topic video, F (1, 58) = .849, p = .361 or the difficult topic video, F (1, 58) = .090, p = .766 (see Figure 4 1 ). Transfer of information from the two videos was measured using a test of 5 transfer questions for both the easy and difficult topic videos T able 4 3 shows the descriptive statistics for the transfer test scores when instructor was present or absent in the videos The finding indicated that instructor presence did not influence transfer score positively or negatively for the easy topic video F (1, 58) = .670 p = 416 However, p articipants who watched the difficult topic video with the instructor present d id significantly bet ter on the transfer test, F (1, 58) = 3.779, p < .05 (see Figure 4 2 ). Learner Perception s T he influences of instructor presence on learner perceptions were examined using measures of cognitive load, satisfaction, perceived learning, situational interest and social presence. Cognitive load Partic ipants self reported four types of cognitive load: overall load (Paas, 1992) intrinsic load (Ayres, 2006) extraneous load (Cierniak, 2009) and germ ane load ( Salomon, 1984) at the end of each video on a 9 point Likert scale The measures of

PAGE 86

86 cognitive load have been used and validated in previous studies ( e.g., Antonenko & Niederhauser, 2010; Leppink et al., 2013) Table 4 4 shows the descriptive statistics for the four types of cognitive load when the instructor was present or absen t in the easy topic video F or the easy topic video, instructor presence did not increase or decrease perceived overall load, F (1, 58) = 1.048 p = .310 ; intrinsic load, F (1, 58) = 1.844 p = .180 ; e xtraneous load F (1, 58) = 2.311 p = .134, or germane load, F (1, 58) = .471 p = .495. Figure 4 3 shows the comparison of self perceived cognitive load when instructor was pr esent or absent in the easy topic vid eo Table 4 5 shows the descriptive statistics for the four types of cognitive loa d when instructor was present or absent in the difficult topic video. For the difficult topic video, instructor presence did perceived overall load, F (1, 58) = .731 p = .396, o r germane load, F (1, 58) = 3.598 p = .063. Howev er, instructor presence was found to decrease i ntrinsic load, F (1, 58) = 4.095 p < .05, and e xtraneous load F (1, 58) = 23.960 p < .001. Figure 4 4 shows the comparison of self perceived cognitive load when the instructor was pr esent or absent in the dif ficult topic vid eo Satisfaction Participants rated th eir satisfaction level with instructor presence using a 7 point Likert scale ranging from extremely dissatisfied (1) to extremely satisfie d (7) This measure has been used in a related previous study (Wang & Antonenko, 2017) and it appear ed sensitive to the action. Table 4 6 shows the ratings when the instructor was present or absent. The fin d ings showed that participants reported

PAGE 87

87 significantly higher satisfact ion when the instructor was present both in th e easy topic video, F (1, 58) = 7.562 p < .001, and in the difficult topic video, F (1, 58) = 99.076, p < .001. The fin di ngs are summarized in Figure 4 5 Perceived learning Participants indicated how much they learn ed from each video using a 5 point Like rt scale that ranged from did not learn anything (1) to learned a great dea l (5) This measure has been used in a related previous study (Wang & Antonenko, 2017) and it also appear ed sensitive to the self perceptions of learning. Table 4 7 shows the reported perceived learning when the in structor was present or absent in the two videos. The analysis showed that p articipants reported a significantly higher level of perceived learning when the instructor was present for the difficult topic, F (1, 58) = 20.354 p < .00 1 However, the instructo topic video, F (1, 58) = .104 p = .748 (Figure 4 6 ). Situational interest Participants reported their situational interest by rating their agreement with the statement I am willin g to watch more video s like this because it is exciting and relevant on a five point Likert scale that ranged from strongly disagree (1) to strongly agree (5) This measure has been reported sensitive to the of sit uational interest with regard to learning environment stimulation (Lathrop, 2011) Table 4 8 shows the reported situational interest level when the instructor was present or absent in the two videos. The findings indicated participants had significantly higher situational interest when the instructor was

PAGE 88

88 present in both the easy topic video, F (1, 58) = 6.795, p < .05, and the difficult topic video, F (1, 58) = 110.481, p < .001 (Figure 4 7 ). Social presence For each video participants indicate d their level of agreement with five statements that examine d the social presence on a five point scale that ranged from strongly disagree (1) to strongly agree (5) The statements include d : I felt like the instructor was in the same r oom as me; I felt that the instructor was very detached in his interactions with me; I felt that the instructor was aware of my presence; I felt that the instructor was present; I felt that the instructor remained focused on me throughout our interaction. This measure has been reported sensitive to the perceptions of social presence in a previous study (Kizilcec et al., 2015) For the current study, i nternal consistency rel iability was calculated using = .7 71 for the five social presence items For the easy topic video, T able 4 9 shows the descriptive statistics for items when the instructor was present or absent in the easy topic video When the instructor was present in the easy topic video F (1, 58) = 5.266, p < .0 5 F (1, 58) = 17.263, p < .001 F (1, 58) = 19.189, p < .001 Also, when the instructor was present in the easy topic video that F (1, 58) = 16.143, p < .001 However the statement, F (1, 58) = 3.258, p

PAGE 89

89 = .076 Overall, participants perceived a higher level of social presence when the instructor was present in the easy topic (Figure 4 8 ). For th e difficult topic video, Table 4 10 shows the descriptive statistics for social p resence responses when the instructor was present or absent in the difficult topic video When the instructor was present in the difficult topic video participants agreed more with the following statement s the same room F (1, 58) = 160.616, p < .001 F (1, 58) = 65.006, p < .001 F (1, 58) = 119.885, p < .001 F (1, 58) = 73.025, p < .001. Also, when the instructor was present in the difficult topic video F (1, 58) = 31.849, p < .001 ; Overall, participants perceived a higher level of social presence when the instructor was present in the difficult topic video (Figure 4 9). eelings toward the instructor At the end of watching the easy and difficult topic videos with instructor presence, p articipants were asked to report their feelings towards the instructor by selecting any of the following adjectives that characterized their feeling s toward the instructor (e.g., frustrati were also encouraged to write down any adjectives that were not provided on the list. The findings indicated that p articipants generally had a positive feeling toward the instructor. They thought the instru both the easy and difficult topic videos. Only a few participants thought Figure 4 1 0 represents the

PAGE 90

90 frequencies of participa nt s responses, for the easy and difficult topic videos respectively. erception s of instructor presence In addition to selecting the adjectives that characterized their feelings toward the instructor, p articipants also responded to an open ended question : Please explain what you think about seeing the instructor in the video, compared to not seeing the instructor. The data analysis followed Creswell's ( 2013) guidelines for data analysis in qualitative research. Thematic analysis was used to search for repeated patterns or themes in responses to this open ended question. Based on analysis of 60 t he following two themes were found. Theme 1: Instructor provided n on verbal cues First, participants appreciated the presence of non verbal cues provided by the instructor One participant commented, d hand gestures reinforces the content being taug ht Another participant wrote, I am able to see his hand movements and gestures, making it easier for me to maintain my focus and attention on what he is saying and less distracted. It also helps me feel mo re connected and learn better than to try to follow along with a voice that I cannot see. Participants also expressed satisfaction with the personal feeling the instructor It makes it seem more personal and like I am in a real, active classroom Having the instructor in the video made the learning experience seem more like a true interaction, however, it wasn't by much due to the special effects of the video. It was still obvious that the instru ctor wasn't aware of my presence. However, I felt that it was easier to focus on the video because I could watch the instructor's body language and movements; it's like watching a professor in a lecture hall. The learning experience is slightly more person alized than by not seeing the instructor in the video at all. When the instructor isn't in the video, it's hard to focus because seeing

PAGE 91

91 just words and material on the screen isn't that interesting to watch. Overall, seeing the instructor makes it easier to pay attention because it creates a mor e engaging learning experience. Theme 2: Instructor h elped focus on learning Moreover, participants reported seeing the instructor on the screen helped them focus on processing the material better. One participant me ntioned, was learning. I think that if I didn't see him, I would not have been as focused on the material This positive influence of instructor presence was also reflected in another response I think that the instructor being visible in this study made it easier for me to focus, rather than just looking at a screen full of words and having to read it for myself. He gave many examples that would help better understand the concepts. Ho wever two participants thought including the instructor was unnecessary. One I don't think there will be too different either with or without the instructor in the video. As long as the voice is being played, I will be able to stay focus It didn't make much of a difference whether he was on screen or not While not seeing him I concentrate more on the paper Overall participants ha d a positive perception of the instructor on the screen for both the easy and the difficult topic videos. Ninety five percent of the participants provided a positive response about their perception of instructor presence. Visual Attention Distribution Par for each video (i.e., easy topic video with the instructor present; easy topic video with the instructor absent; difficult topic video with the instructor present; difficult topic video with the instructor absen t). To

PAGE 92

92 ttention distribution, was examined with regard to two areas of interest: the instructor interest area (IA) and the content interest area (IA). As the instructor absent condition did not a ctually have an instructor IA, I created a corresponding interest area that ha d the equivalent size and Based on the aggregated fixation maps ( Figure 4 11 for easy topic videos and Figure 4 1 2 for difficult topic videos) the participants distributed a significant amount of attention to the instructor on the screen for both video s particularly focusing on the face part of the instructor. ntage of fixations in each interest area. Also, as fixations are different in length, I also examined the percentage of dwell time (i.e., total length of fixations) in each interest area. Last, as visual attention wa s hypothesized to split between the inst ructor frame and the content frame when the instructor is present, I also examined the number of transitions between the two interest areas to understand the magnitude of split attention when instructor was present in the video. For the easy topic video, t he statistics on instructor IA_f ixations (%) i nstructor IA_ dwell time (%) content IA_f ixations (%) content IA_ dwell time (%) n umber of transitions between the two IAs are provided in Table 4 11 For the easy topic video, w hen instructor was present, l earners devoted 22% of the fixations and 35% of the dwell time to the instructor. Compared to the instructor absent videos, the same area of interest around the instructor attracted significantly more fixations, F (1, 58) = 251.570, p < .001 and a longer dw ell time, F (1, 58) = 334.181, p < .001. Also, when the instructor

PAGE 93

93 was present, less attention was distributed to the content interest area compared to the instructor absent condition. Moreover participants made 22 transitions on average between the two in terest areas. For the difficult topic video s the statistics on the instructor IA_f ixations (%) i nstructor IA_ dwell time (%) content IA_f ixations (%) content IA_ dwell time (%) n umber of t ransitions between the two IAs are provided in Table 4 12 For t he difficult topic video, w hen the instructor was present, learners devoted 27% of the fixations and 37% of the dwell time to the instructor. Compared to the instructor absent videos, the same area of interest around the instructor attracted significantly more fixations, F (1, 58) = 323.131, p < .001 and a longer dwell time, F (1, 58) = 242.273, p < .001. Similar to the finding for the easy topic videos when the instructor was present, less attention was distributed to the content interest area compared to t he instructor absent condition. Also, participants made an average of 47 transitions between the two interest areas. Cognitive Dynamics using their EEG data for the two videos when the instructor was present or absent. As it was a n authentic learning task, there were artifacts in the EEG data such as eye blinks and body movements Therefore, b efore the time frequency analysis of EEG data, EEG artifacts due to body movement and eye blinks were decontaminated us ing the Independent Compo nent Analysis (ICA) approach (Delorme & Makeig, 2004) example (see Figure 4 1 3 ) components 1, 2, 15 are typical examples of vertical eye blink and the components got rejected befo re the time freq uency analysis of the EEG data for this specific participant.

PAGE 94

94 After decontaminati ng EEG data, similar to the approached used by Stevens and colleagues ( 2013) the data for the two videos were broken down into 2 second segments for each video and the noise free segments were retained for further power spectrum analysis. The easy t opic video was 170 seconds long and so each EEG file was divided into 85 segments The difficult topic vid e o was 240 seconds long and so each EEG file was divided into 120 segments. A high percentage of noise free segments were retained for furt her analysis for easy topic video with instructor present (90%), easy topic video with instructor absent (97%), difficult topic video with instructor present (97%), difficult topic video with instructor absent (95%). Table 4 13 summarizes the number of int erpretable EEG data and the number of noise free segments for each video condition. P ower S pectr al A nalysis was then conducted on the EEG data using Fast Fourier Transform in EEGLAB (Delorme & Makeig, 2004) I n this study, the analysis focused on examining alpha and theta wave activity as alpha and theta wave activity are most prevalent when a person is cognitively engaged, such as in a learning task (Klimesch, Schack, & Sauseng, 2005) As the alpha rhythm is most pronounced in the parietal l obe I focused the analyses of the alpha wave data acquired from three parietal lobe sensors P3, P4, and Pz. As the theta rhythm is most pronounced in the central and frontal lobe s, I focused the analyses of the theta wave data acquired from three central lobe sensors C3, C4, and C z as well as the three frontal lobe sensors F3, F4, and Fz. Absolute power in theta frequency band (4 8 Hz) and alpha frequency band (8 13 Hz) was computed and compared between the instructor present and instructor absent ver sions of the easy and difficult topic videos.

PAGE 95

95 Table 4 14 shows p EEG power in theta and alpha bands for the easy topic video. Before group comparison, EEG absolute power data was transformed using L og transformation to obtain normal ly distributed data (van Albada & Robinson, 2007) The findings indicated that for the easy topic video, h igher theta absolute power values in C4 were found in t he instructor present condition, F (1, 37) = 5.283, p < .05. Figure 4 1 4 at C4 for the easy topic video when the instructor was present and absent. Table 4 15 summarizes p EEG power in theta and alpha bands for the difficult topic video. The findings indicated that for the difficult topic video, higher theta absolute power values at C4 were found in t he instructor absent conditi on, F (1, 42) = 8.918, p < .01. Also a higher average theta absolute power values at C3, C4, and Cz were found in the instructor absent condition, F (1, 42) = 5.587, p < .05. Figure 4 1 5 4 for the difficult topic video when instructor was present or absent. Figure 4 1 6 average absolute power at C3, C4, and Cz for the difficult topic video when the instructor was present or absent. Predictors of Learning and Learner Perceptions I also conducted a seri es of regression analyses to explore if visual attention distribution and cognitive dynamics, as process measures, predict ed learning and learner perceptions. In the following section, I presented the findings on v isual attention distribution as a predic to r of learner perception as well as cognitive dynamics as a predictor of learning and learner perception. Visual Attention Distribution Predicted Learner Perception To explore the effects of visual attention distribution on learner perception in the instr uctor present easy and difficult topic videos, I conducted a series of linear

PAGE 96

96 regression analyses with learner perception variables (i.e., satisfaction, situational interest perceived learning, cognitive load types level of agreement with social presence statements ) as outcome variables and the visual attention distribution variables ( i.e., fixation count on the instructor IA, percentage of fixation count on the instructor IA, dwell time on the instructor IA, percentage of dwell time on the instructor IA, number of transitions between the instructor IA and the content IA) as predictor variables. Easy topic video with instructor present For the instructor present easy topic video, t he regression analysis indicated f ixation count on the instructor IA signi ficantly predict ed participan F (1, 28) = 4.984, p < .05 R 2 = .156 T he more fixations were allocated to the instructor IA, the more satisfied the pa rticipants were with the instructor present easy topic video, b = .015, t (28) = 2.2 33, p < .05. D ifficult topic video with instructor present For the instructor present difficult topic video, t h e regression analysis indicated p ercentage of fixation count on the instructor IA significantly positively predict ed ating with the video, F (1, 28) = 7.309, p < .05 R 2 = .207 The higher percentage of fixation count were allocated to the instructor IA, the more satisfied the p articipants were with the instructor present difficult topic video, b = 5.660, t (28)= 2.704, p < .05. Also, p ercentage of dwell time significantly positively predicted satisfaction rating with the video, F (1,28) = 10.10, p < .05 R 2 = .265 The higher percentage of dwell time were allocated to the instructor IA, the more satisfied the

PAGE 97

97 p articipants were with the instructor present difficult topic video, b = 4.645, t (28)= 3.178, p < .01. To summarize the findings, for both the instructor present easy and difficult topic videos, the amount of attention allocated to the in structor IA posit ively predicted ratings The more participants paid attention to the instructor on the screen, the m ore satisfied they were with the instructor present videos However, the amount of attention to the instructor was not found to p redict other learner perception variables, which included situational interest perceived learning, cognitive load types or level of agreement with social presence statements I found no effects of increased attention to the instructor on improving learni ng test scores (i.e., retention and transfer test scores), either. Cognitive Dynamics Predicted Learning To explore the effects of cognitive dynamics on learning, I conducted a series of multiple regression analys e s with each one of the learning test scor es (i.e., retention and transfer test scores) as outcome v ariable and each one of cognitive dynamics variables (i.e., Alpha power at P3; Alpha power at P4; Alpha power at Pz; Average Alpha power at P3 & P4 & Pz; Theta power at C3; Theta power at C4; Theta power at Cz; Average t heta power at C3, C4, and Cz; Theta power at F3; Theta power at F4; Theta power at Fz; Average t heta power at F3, F4 and Fz ) and instructor presence as predictor variables. The analys e s w ere conducted for easy and difficult topic vid eos, respectively. E asy topic video For the easy topic video, t he multiple regression model including theta power at F4 and instructor presence as predictor variables and retention test score as outcome

PAGE 98

98 variable was significant at F (2, 35 ) = 3.439 p < 05, R 2 = .164. Theta power at F4 was a significant negative predict or of test scores, b = 1.676 t (35 ) = 2.552 p < .05 controlling for instructor presence Besides, the multiple regression model incl uding average theta power at F3, F4 and Fz and instructor presence as predictor var iables, and retention test score as outcome variable was significant at F (2, 35) = 2.687, p < .05, R 2 = .133. A verage theta power at F3, F4, and Fz was a significant negative predict or of participants test scores, b = 1.498, t (35) = 2.241, p < .05, controlling for instructor presence. Difficult topic video For the difficult topic video, the findings indicated n one of the cognitive dynamics variables (i.e., Alpha power at P3; Alpha power a t P4; Alpha power at Pz; Average Alp ha power at P3, P4 and Pz; Theta power at C3; Theta power at C4; Theta power at Cz; Average t heta power at C3 C4 and Cz; Theta power at F3; Theta power at F4; Theta power at Fz; Average t heta power at F3, F4 and Fz) along with the instructor presence variable significantly predict ed either the retention or transfer test scores. C ognitive Dynamics Predict ed Learner Perception To explore the effects of cognitive dynamics on le arner perception I conducted a series of m ultiple regression analys e s with each one of the learner perception variables (i.e., satisfaction, situational interest perceived learning, cognitive loads, level of agreement with social presence statements ) as outcome v ariable and each one of cognitive dynamics variables (i.e., Alpha power at P3; Alpha power at P4; Alpha power at Pz; Average a lpha power at P3 P4 and Pz; Theta power at C3; Theta power at C4; Theta power at Cz; Average t heta power at C3, C4 and Cz; Theta power at F3; Theta

PAGE 99

99 power at F4; Theta power at Fz; Average t heta power at F3, F4 and Fz) and instructor presence as predictor variables. Easy topic video For the easy topic video, t he multiple regression model including theta power at Fz and instructor presence as predictor vari ables, and level of agreement with social presence statement #2 (i.e., I felt that the instructor was very detached in his interactions with me ) as outcome variable was significant at F (2, 35) = 7.368 p < .01 R 2 = .296 Theta power at F4 was a significant nega tive predict or of level of agreement with social presence statement #2 (i.e., I felt that the instructor was very detached in his interactions with me ) b = .812 t (35) = 2.097 p < .05, controlling for instructor presence. Besides, the multiple regression model inclu ding average theta power at F3, F4 and Fz and instructor presence as predictor variables and level of agreement with social presence statement #2 (i.e., I felt that the instructor was very detached in his interactions with m e ) as outcome variable was significant at F (2, 35) = 6.666 p < .01 R 2 = .276 Average theta power at F3 F4 and F z significantly negatively predict ed level of agreement with social presence statement #2 (i.e., I felt that the instructor wa s very detached in his interactions with me ), b = .746 t (35) = 1.812 p < .05, controlling for instructor presence. Difficult topic video N one of the cognitive dynamics variables (i.e., Alpha power at P3; Alpha power at P4; Alpha power at Pz; Average A lpha power at P3, P4 and Pz; Theta power at C3; Theta power at C4; Theta power at Cz; Average t heta power at C3, C4 and Cz; Theta power at F3; Theta power at F4; Theta power at Fz; Average t heta power at F3, F4 and

PAGE 100

100 Fz) along with the instructor presence variable significantly predict ed the learner perceptio n variables (i.e., satisfaction, situational interest per ceived learning, cognitive load types level of agreement with social presence statements). Moderating Effects of Individual Differences The s econd goal of the study was to explore the moderating effects of individual differences in easy and difficult topic videos with and without the instructor In the following section, for working memory capacity and inhibitory control respectively, I present moderating effects of individual differences in easy and difficult topic videos. Working Memory Capacity Individual differences in working memory capacity Working memory capacity a ffects the ability to Engle, 1994) In this study, working memory capacity was m easured by the Automated Operation Sp peration Sp an (O Span) s cores range d from 13 to 75 where the maximum score was 75. The mean score equals 49.21 ( SD = 14.96) s cores is presented in Figure 4 17 Moderating effects of individual differences in working memory capacity To explore the moderating effects of working memory capacity on learning in easy and difficult topic videos with and without the instructor I conducted a series of multiple linear regression analys e s with each one of the learning test scores (i.e., retention and transfer test scores) as outcome variable, and using instructor presence and working memory capacity as predictor variables. Same procedure was followed to examine the moderating effects of working memory capacity on learner perception in

PAGE 101

101 easy and difficult topic videos with and w ithout the instructor except using each one of the learner perception variables (i.e., satisfaction, situational interest per ceived learning, cognitive load types level of agreement with social presence statements) as outcome variable. I also explored t he moderating effects of working memory capacity on cognitive dynamics and visual attention distribution in easy and difficult topic videos with and without the instructor Easy topic video. T he multiple regression model including working memory capacity score and instructor presence as predictor variables and retention test score for the easy topic video as outcome variable was significant at F (2, 56) = 6.063 p < .01, R 2 = .181 Working memory capacity score was a significant positive predictor of parti retention test score for the easy topic video, b = .051, t (56 ) = 3.358, p < .001, controlling for instructor presence. However, t he multiple regression model including working memory capacity score and instructor presence as predictor variables, and transfer test score or learner perception variables or cognitive dynamics variables, or visual attention distribution variables for the easy topic video as outcome variable was not significant at p < .05. Difficult topic video. T he multiple regressi on model including working memory capacity score and instructor presence as predictor variables, and retention test score or transfer test score, or learner perception variables or cognitive dynamics variables, or visual attention distribution variables for the difficult topic video as outcome variable was not significant at p < .05.

PAGE 102

102 Inhibitory C ontrol Individual differences in inhibitory c ontrol ability to inhibit attention to irrelevant stimuli while focusing on the task goal (Rothbart & Posner, 1985). In the current study, inhibitory control was measured by the Flanker Test (Eriksen & Eriksen, 1974). demographic variables using corrected inhibitory control scores ranged from 21 to 70 where a higher score indicated a stronger ability to inhibit attention to irrelevant stimuli while focusing on the task goal. The m ean fully corrected inhibitory control score was 40.83 ( SD = 10.03) Figure 4 18 Moderating effects of individual differences in inhibitory c ontrol To explore the moderating effe cts of inhibitory control in easy and difficult topic videos with and without the instructor I conducted a series of multiple linear regression analyses with each of the learning test scores (i.e., retention and transfer test scores) as outcome variable, and using instructor presence and inhibitory control as predictor variables. Same procedure was followed to examine the moderating effects of inhibitory control on learner perception in easy and difficult topic videos with and without the instructor excep t using each one of the learner perception variables (i.e., satisfaction, situational interest per ceived learning, cognitive load types level of agreement with social presence statements) as outcome variable. I also explored the moderating effects of inh ibitory control on cognitive dynamics and visual attention distribution in easy and difficult topic videos with and without the instructor

PAGE 103

103 Easy topic video. T he multiple regression model including inhibitory control score and instructor presence as predic tor variables, and retention test score or transfer test score, or learner perception variable s or cognitive dynamics variables, or visual attention distribution variables for the easy topic video as outcome variable was not significant at p < .05. Diff icult topic video. T he multiple regression model including inhibitory control score and instructor presence as predictor variables and transfer test score for the difficult topic video as outcome variable was significant at F (2, 56 ) = 4.986 p < .01, R 2 = .151 Inhibitory control score was a significant positive predictor of transfer test score for the difficult topic video, b = .045, t (55) = 2.423 p < .05 controlling for instructor presence. However, t he multiple regression model including inhibitory control score and instructor presence as predictor variables, and retention test score or learner perception variable s or cognitive dynamics variables, or visual attention distribution variables for the difficult topic video as outcome variab le was not significant at p < .05. Summary of Findings The study has two main purposes. The first purpose was to study the influences of instructor presence on the products of learning (i.e., learning and learner perceptions ) and processes of learning (i.e ., visual attention distribution and cognitive dynamics) in easy and difficult topic videos respectively The second purpose was to examine the moderating effects of individual differences in working memory capacity and inhibitory control in instructional videos with and without instructor presence. Regarding the first purpose of the current study, the data suggested that t he instructor on the screen did not affect retention positively or negatively in either topic,

PAGE 104

104 er information from the difficult topic. Regarding the influence s on learner perceptions, instructor presence in the video produced a situational interest for both topics and led to higher level of perceived l earning for the difficult topic. Instructor presence also helped lower the intrinsic load and the extraneous load for the difficult topic video while not influencing any type of cognitive load for the easy topic video. The instructor on the screen was larg ely perceived as helpful, entertaining, us eful, engaging for both topics and a high percentage of participants preferred to see the instructor compared to not seeing it for both easy and difficult topic videos. Also, p rocess measures (i.e., visual attenti on distribution and cognitive dynamics) indicated that the instructor on the screen attracted significant amount of vi sual attention for both topics and when the instructor was present, less attention was distributed to the content interest area Regarding cognitive dynamics, instructor presence led to higher theta power at C4 sensor for the easy topic video and lower theta power at C4 senso r for t he difficult topic video, where a lower theta power indicated a lower level of working memory load. Also, the current study suggested the process measures (i.e., visual attention distribution and cognitive dynamics) were able to predict product measures (i.e., learning and learner perception). First, t he more participants paid attention to the instructor on the sc reen, the more satisfied they were with the instructor present videos. For instructor present easy topic video, the fixation count on the instructor IA positively And for the instructor present difficult topic v ideo, the percentage of fixation count and the percentage of dwell time on the

PAGE 105

105 video. Second, cognitive dynamics was found to predict learning. Specifically, f or th e easy topic video, theta power at F4 sensor and the average theta power at F3, F4, and Fz (i.e., three frontal lobe sensors) were both significant negative predictors of Regarding the second purpose of the study, we did not identify any moderating effects of individual differences in working memory capacity and inhibitory control. In fact, participants who had higher WMC scores performed significantly better on the retention test for the easy topic, controlling for i nstructor presence. Also, those participants who had higher inhibitory control scores excelled on the transfer test for the difficult topic.

PAGE 106

106 Table 4 1 Participant c haracteristics Variables Statistics Gender 21 Male, 39 F emale Age M = 18.36 ( SD = 0.6 6 ) Race 25 White/Caucasian 20 Hispanic/Latino 7 Black / African American 8 Asian / Pacific Islander Table 4 2. Mean accuracy (%) and standard deviations on retention test Topic Instructor Present Instructor Absent Mean SD Mean SD Easy 0.82 0.16 0.85 0.15 Difficult 0.53 0.21 0.51 0.21 Table 4 3. Mean accuracy (%) and standard deviations on transfer test Topic Instructor Present Instructor Absent Mean SD Mean SD Easy 0.84 0.20 0.79 0.19 Difficult 0.71 0.26 0.55 0.32

PAGE 107

107 Table 4 4. Mean s and standard deviations on four types of cognitive load for the easy topic video Cognitive Load Instructor Present Instructor Absent Mean SD Mean SD Overall load: In the video I just finished watching, I invested (1: very, very low mental effort; 9: ve ry, very high mental effort) 4.72 1.46 4.33 1.47 Intrinsic load: The video I just finished watching was (1: very, very easy; 9: very, very difficult) 3.17 1.26 3.60 1.16 E xtraneous load : Learning from this video was (1: very, very easy; 9: very, very d ifficult) 3.00 1.13 3.53 1.53 Germane load: When watching this video, I concentrated (1: very, very little; 9: very, very much) 5.45 1.38 5.20 1.40 Table 4 5. Mean s and standard deviations on four types of cognitive load for the difficult topic video C ognitive Load Instructor Present Instructor Absent Mean SD Mean SD Overall load: In the video I just finished watching, I invested (1: very, very low mental effort; 9: very, very high mental effort) 6.07 1.23 5.76 1.53 Intrinsic load: The video I just finished watching was (1: very, very easy; 9: very, very difficult) 5.97 1.27 6.62 1.21 E xtraneous load : Learning from this video was (1: very, very easy; 9: very, very difficult) 4.90 1.77 6.97 1.45 Germane load: When watching this video, I concentra ted (1: very, very little; 9: very, very much) 6.30 1.29 5.45 2.08

PAGE 108

108 Table 4 6. Mean and standard deviations on satisfaction ratings with the easy and difficult topic videos. 1: Extremely dissatisfied; 7: Extremely satisfied Topic Instructor Present Instr uctor Absent Mean SD Mean SD Easy 5.10 1.40 4.10 1.37 Difficult 5.63 1.00 2.48 1.40 Table 4 7. Mean and standard deviations on perceived learning for the easy and difficult topic videos Topic Instructor Present Instructor Absent Mean SD Mean SD E asy 2.72 0.96 2.80 0.85 Difficult 3.13 0.73 2.14 0.95 Table 4 8. Mean and standard deviations on self reported situational interest level for the easy and difficult topic videos Topic Instructor Present Instructor Absent Mean SD Mean SD Easy 3.10 0.94 2.47 0.94 Difficult 3.77 0.77 1.66 0.77

PAGE 109

109 Table 4 9. Mean and standard deviations on level of agreement with the social presence statements for the easy topic video Social Presence Statements 1: Strongly disagree; 5: Strongly agree Instructor Present Instructor Absent Q1. I felt like the instructor was in the same room as me. 2.86 (0.92) 2.33 (1.30) Q2. I felt that the instructor was very detached in his interactions with me. 2.10 (1.01) 3.17 (1.02) Q3. I felt that the instruct or was aware of my presence. 3.14 (0.88) 2.50 (1.22) Q4. I felt that the instructor was present. 3.86 (0.83) 2.70 (1.26) Q5. I felt that the instructor remained focused on me throughout our interaction. 3.72 (0.92) 2.47 (1.25) Table 4 10. Mean and st andard deviations on level of agreement with the social presence statements for the difficult topic video Social Presence Statements 1: Strongly disagree; 5: Strongly agree Instructor Present Instructor Absent Q1. I felt like the instructor was in the sa me room as me. 3.77 (0.90) 1.31 (0.54) Q2. I felt that the instructor was very detached in his interactions with me. 2.20 (0.92) 3.76 (1.18) Q3. I felt that the instructor was aware of my presence. 3.47 (0.78) 1.72 (0.88) Q4. I felt that the instruct or was present. 4.03 (0.72) 1.76 (0.87) Q5. I felt that the instructor remained focused on me throughout our interaction. 3.83 (0.75) 1.83 (1.04)

PAGE 110

110 Table 4 11 Visual attention distribution statistics for the easy topic video Parameters Instructor Pres ent Instructor Absent Instructor IA_Fixations (%) (average percentage of all fixations on instructor IA) 22% 1% Instructor IA_Dwell time (%) (average percentage of trial time spent on instructor IA) 35% 2% Content IA_Fixations (%) (average percentage of all fixations on content IA) 78% 99% Content IA_Dwell time (%) (average percentage of trial time spent on content IA) 65% 98% Number of transitions between IAs (average number of times the attention was switched from the instructor IA to the conte nt IA, or from the content IA to the instructor IA) 22 1 Table 4 12 Visual attention distribution statistics for the difficult topic video Parameters Instructor Present Instructor Absent Instructor IA_Fixations (%) (average percentage of all fixations on instructor IA) 27% 1% Instructor IA_Dwell time (%) (average percentage of trial time spent on instructor IA) 37% 2% Content IA_Fixations (%) (average percentage of all fixations on content IA) 73% 99% Content IA_Dwell time (%) (average percentage of trial time spent on content IA) 63% 98% Number of transitions between IAs (average number of times the attention was switched from the instructor IA to the content IA, or from the content IA to the instructor IA) 47 3

PAGE 111

111 Table 4 13. Number of interpre table EEG data and number of noise free segments for each video condition Number of total segments for easy topic video EEG data 85; number of total segments for difficult topic video EEG data 120 Video # of interpretable EEG data Mean (SD) of noise f ree segments Easy Topic Instructor Present 22 76.73 (9.69) Easy Topic Instructor Absent 17 82.65 (2.57) Difficult Topic Instructor Present 23 115.83 (3.76) Difficult Topic Instructor Present 21 113.57 (5.95) T able 4 14 solute power in theta and alpha bands for the easy topic video Instructor Present Instructor Absent Mean SD Mean SD Alpha power at P3 1.33 0.76 1.00 0.40 Alpha power at P4 1.37 0.78 1.06 0.33 Alpha power at Pz 1.26 0.75 1.03 0.32 Average Alpha power at P3, P4, and Pz 1.32 0.72 1.03 0.34 Theta power at C3 1.26 0.54 1.07 0.30 Theta power at C4 1.39 0.68 0.99 0.27 Theta power at Cz 1.90 2.18 1.39 0.70 Average theta power at C3, C4, and Cz 1.51 0.94 1.15 0.33 Theta power at F3 1.31 0.55 1.25 0.92 Theta power at F4 1.27 0.47 1.17 0.32 Theta power at Fz 1.34 0.54 1.14 0.23 Average theta power at F3, F4, and Fz 1.31 0.42 1.19 0.41 indicates significant difference in absolute power between instructor present and instructor absent conditions at p < .05.

PAGE 112

112 Table 4 15 topic video Instructor Pre sent Instructor Ab sent Mean SD Mean SD Alpha power at P3 0.95 0.33 1.27 0.84 Alpha power at P4 0.97 0.30 1.33 1.10 Alpha power at Pz 1.10 0.38 1.23 0.68 Average Alpha power at P3, P4 and Pz 1.01 0.32 1.28 0.85 Theta power at C3 0.94 0.23 1.05 0.32 Theta power at C4 0.90 0.30 1.29 0.68 Theta power at Cz 1.04 0.25 1.42 1.41 Average theta power at C3, C4, and Cz 0.96 0.21 1.25 0.66 Theta power at F3 1.41 1.72 1.20 0.42 Theta power at F4 1.09 0.33 1.18 0.29 Theta power at Fz 1.03 0.20 1.17 0.23 Average theta power at F3, F4, and Fz 1.18 0.64 1.18 0.27 indicates significant difference in absolute power between instruct or present and instructor absent conditions.

PAGE 113

113 Figure 4 videos in the instructor present and instructor absent conditions. Figure 4 percentage for the easy and difficult topic videos in the instructor present and instructor absent conditions.

PAGE 114

114 Figure 4 reported cognitive load for the easy topic video. Figure 4 reported cognitive load for the difficult topic video.

PAGE 115

115 Figure 4 Figure 4 reported perceived learning. 1. Not at all; 2. A little; 3. A moderate amount; 4. A lot; 5. A great deal.

PAGE 116

116 Figure 4 reported situational interest level. Figure 4 8 responses for the easy topic video.

PAGE 117

117 Figure 4 9. responses for the difficult topic video. Figure 4 videos. Figure 4 color represents a higher amount of visual attention; Yellow colo r medium amount of visual attention; Green color lower amount of visual attention.

PAGE 118

118 Figure 4 color represents a higher amount of visual attention; Yellow color med ium amount of visual attention; Green color lower amount of visual attention. Figure 4 13. Example Independent Component Analysis output from one participant. Components 1, 2, 15 (in red boxes) are typical examples of vertical eye blink and these c omponents were rejected before the time frequency analysis of the EEG data for this participant.

PAGE 119

119 Figure 4 and absent (perforated line) in the easy topic video. Theta wave i s in the 4 8 Hz range indicated by the dotted vertical lines. Figure 4 and absent (perforated line) in the difficult topic video. Theta frequency band is in the 4 8 H z indicated by the dotted vertical lines.

PAGE 120

120 Figure 4 topic video when instructor was present (solid line) and absent (perforated line). Theta frequency band is in the 4 8 Hz in dicated by the dotted vertical lines. Figure 4

PAGE 121

121 Figure 4

PAGE 122

122 CHAPTER 5 DISCUSSION This chapter discusses the findin gs of the current study. I first discussed the findings with regard to the influences of instructor presence on learning, learner perceptions, visual attention distribution, and cognitive dynamics in easy and difficult topic videos. I also discussed the pr edictors of learning and learner perceptions (i.e., process measures such as visual attention distribution and cognitive dynamics) in easy and difficult topic videos with or without instructor presence. Next, I discussed the moderating effects of individua l differences in working memory capacity and inhibitory control in easy and difficult topic videos when the instructor was present or absent. Influences of Instructor Presence on Learning, Learner Perception, Visual Attention Distribution, and Cognitive D ynamics The first goal of this study was to examine the influences of instructor presence on the products of learning (i.e., learning and learner perception s ) and processes of learning (i.e., visual attention distribution and cognitive dynamics) in easy an d difficult topic videos respectively The data generated from the product measures (i.e., learning and learner perception) showed that t he instructor on the screen did not affect retention bility to transfer information from the difficult topic. Regarding the influence s on learner perceptions, situational interest for both topics and led to higher leve l of perceived learning for the difficult topic. Instructor presence helped lower the intrinsic load and the extraneous load for the difficult topic video while not influencing any type of cognitive load for the easy topic video. The instructor on the scre en was largely perceived as helpful,

PAGE 123

123 entertaining, us eful, engaging for both topics and a high percentage of participants preferred to see the instructor compared to not seeing it for both easy and difficult topic videos. Next, the findings based on the pr ocess measures (e.g ., visual attention distribution) indicated that the instructor on the screen attracted significant amount of vi sual attention for both topics and when the instructor was present, less attention was distributed to the content interest ar ea Regarding cognitive dynamics, instructor presence led to higher theta power at C4 sensor for the easy topic video and lower theta power at C4 sensor for the difficult topic video. Learning In the current study, p articipants who viewed difficult top ic video with the instructor present performed significantly better on the transfer test compared to those who viewed the instructor absent version video A possible explanation for this result is that the instructor in this study used a variety of nonverb al communication means including mutual gaze, facial expressions, and gestures, which attracted a significant for the difficult topic video ). As discussed in the introduction to this s tudy, nonverbal cues are very important in everyday social interactions (Argyle, 2013) and in mathematics learning and instruction specifically (e.g., Alibali & Nathan, 2012) The nonverbal cues provided by the instructor in our study likely served an important signaling function (Mayer, 2014; Van Gog, 2014) and direct ed to the most important and relevant aspects of the instructional content, thus resulting in improved comprehension of the difficult material and enhanced transfer performance. The finding regardi ng the positive influence on the transfer of information is aligned with the results of a previous study conducted by Chen and colleagues (2015). The researchers used an experimental

PAGE 124

124 design and compared the influence of three types of videos on learning do cument capture (i.e., a video recording of the lecture) and picture in picture video that included an instructor frame in the bottom right hand corner. T he researchers found the transfer performance was improved by instructor presence, that is, p ar ticipants who watched the video with instructor presence in either the picture in p icture or the lecture capture format outperformed those who received the instruction in the voice over format The finding from the current study provides empirical evidence for the assumption that the instructor in the video could help direct and possibly maintain materials when the topic is difficult However, the positive effect of instructor presence on transfer was not replicated in the case of the easy topic video. The nonverbal cues did not produce quite the same magnitude of effect for the easy topic video l ikely because learners had more cognitive resources at their disposal as they were learning easier content. So, i t is reasonable to assume that in this situation, when the instructor was absent in the easy topic video, the transfer performance was already high considering the information was less difficult for learners Therefore, it is possible that the ceiling effect prevented us from observing the additional benefits of instructor presence on transfer perfor mance for the easy topic video. In addition to transfer performance t he present study examined if instructor presence could help improve retention of information for the easy and difficult topic videos. However, the study did not identify any evidence that instructor presence enhanced retention of in formation for either topic. The finding is consistent with the

PAGE 125

125 study conducted by Kizilcec and c olleagues (2014 ) who investigated how adding the instructor t o instructional video influenced learning on a topic in Organizational Sociology. The ir findings indicated that learners who received the video instruction with instructor presence did not perform significantly better or worse on knowledge recall tests compared to the control condition wit hout instructor presence. However, in another study that examined the influence of instructor presence on retention in eas y and difficult topic videos (i.e., Similar Triangles and Trigonometry) Wang & Antonenko (2017) found that instructor presence helped from the easy topic video. The se mixed findings regarding the influence of instructor presence on retention of information can possibly be explai ned by the differences in the nature of materials and topics used in the videos. Learner Perceptions Unlike many studies that focus solely on learning outcome s ( e.g., retention and tra nsfer of information) of the efficacy of the educational intervention that is, instructor presence in instructional video. P ognitive load (i.e., ov erall load, intrinsic load, extraneous load, and germane load) suggested that in the difficult topic reported intrinsic load and extraneous load, although not affecting any type of cognitive load p articipants self perceived for the easy topic video P erceived learning, situational interest and satisfaction scales, as well as participant responses to the open ended questions indicated a strong preference for the videos with instructor presence for b oth easy and difficult topics. This is important because student perceptions moderate their experience

PAGE 126

126 with learning tools and materials and so it is important to understand where learners stand with regard to their preferences. C ognitive load In the pres ent study, instructor presence was found to decrease self reported intrinsic load and extraneous load for the difficult topic video, while not influencing any type of self reported cognitive load for the easy topic video. F or the easy topic video instruc tor presence did not increase or decrease any type of self reported cognitive load, which included overall load (Paas, 1992) intrinsic load (Ayres, 2006) extraneous load (Cierniak, 2009) or germane load (Salomon, 1984) The result on overall load is con sistent with the finding from Wang & Antonenko (2017), where participants who watched an easy topic video (i.e., Similar Triangle) with or without instructor presence reported the same level of overall load. Similar to the findings on retention and transfe r of information for the easy topic video, instr uctor presence did not influence overall load, intrinsic load, extraneous load or germane load Based on the statistics regarding cognitive load for the easy topic video, it is reasonable to speculate the fo ur types of cognitive load are relatively low for the easy topic video when the instructor was absent resulting in a floor effect (i.e., around 3 on 9 point Likert scale s ). So, this study fail ed to show that instructor presence decrease d the already low co gnitive load for the easy topic video For the difficult topic video, the findings showed that instructor presence did not overall load or germane load. However, instructor presence was found to decrease self reported intrinsic load and extraneous load. The finding is important as it points out the usefulness of instructor presence in lowering the intrinsi c load of the content, especially when the content itself is difficult. T he instructor on the

PAGE 127

127 screen helped the participants with l ittle knowledge about the difficult topic in this study (i.e., Analysis of Variance) perceive a lower level of intrinsic load. A possible explanation for this verbal cues such as body language and facial expression were instrumental in making the content more understandable, and thus helping decrease the intrinsic load of the content Th is explanation can also be corroborated ended question, Please explain what you think about seei ng the instructor in the video, compared to not seeing the instructor liked being able to see him so that I could 'maintain' eye contact with him and watch his hand motions. These two factors aided in m y compreh ension of the material Another two participant s expressed the same feeling in the ir response s The instructor allowed me to fully understand the concept because he was present and his gestured helped ely added a visual that helped me understand the lecture better than the video when he was not present. Moreover the extraneous load was lower when the instructor was present in the difficult topic video This finding is quite interesting, as the instruc tor present video actually include d additional information on the screen (i.e., the instructor) as compared to instructor absent version. This additional inform ation, in turn, helped decrease the extraneous load participants perceived for the difficult top ic video Relevant information regarding this interesting finding was provided by ended question, compared to not seeing the instructor For examp le, o ne participant pointed out that the instructor on the screen helped him/her focus on the material better Seeing the

PAGE 128

128 instructor helped me keep my focus on what I was learning. I think that if I didn't see him, I would not have been as focused on the material So, i t appears seeing the instructor on the screen help ed at least some learners give their attention to the most important information on the screen, and avoid atten ding to extraneous components on the screen, thus lowering extraneous load. Un like most previous studies on instructor presence that just examined the overall load (e.g., Wang & Antonenko, 2017; Homer & Plass, 2008), the present study explored the influence of instructor presence on cognitive load in a more detailed manner. B y exami ning three individual types of cognitive load (i.e., intrinsic load, extraneous load, and germane load), we were able to identify the positive influence of instructor presence on lowering extraneous load and intrinsic load for the difficult topic video It is important that future research on instructor presence examines effects on diffe rent types of cognitive load instead of merely gauging the overall load. Caution has to be exercised, however, because the extent to which the existing measures tap these d ifferent types of cognitive load still needs to be explored. S atisfaction The findings from the current study showed that participants reported significantly higher level of satisfaction when the instructor was present in both easy and difficult topic vid eos. The results of this study align with the findings of Wang & Antonenko (2017). In that study, undergraduate participants were assigned to watch two videos on an easy mathematics topic ( S imilar T riangles) and a difficult topic (Trigonometry), each with the instructor present or absent in a counter balanced design. The ir findings the two instructor present videos on an easy and a difficult topic.

PAGE 129

129 The satisfaction re sults of this study echo the findings of Fredericksen and colleagues (2000) who reported that student instructor interaction is the most significant factor that contributes to a higher level of satisfaction with the course. Similarly, empirical data on the affective aspects of animated pedagogical agents in multimedia learning suggest agents, especially when the level of embodiment is high (e.g., Baylor & Kim, 2005 ; Kim & Baylor, 2016 ). P erceived learning P articipants reported a significantly higher level of perceived learning when the instructor was present in the difficult topic. learning were not correlated with their actual learning perf ormance on either the retention or the transfer test. This result confirms findings from other relevant studies that tried to compare subjective, self reported learning performance or judgments of learning with the results of objective learning tests. For example, in a study published in the journal Science Karpicke & Blunt (2011) found that students could not accurately predict what learning strategy was the most effective for restudy for test The strategy they iden tified in their metacognitive predictions (repeated study) was significantly less effective than retrieval practice, which resulted in the highest learning outcomes on recall and inference tests. Other relevant studies have reported the following problems with self reports of learning: students typically overestimate how well they understand, fail to recognize their own states of impasse in problem solving, and persist with unproductive strategies (Anderson & Beal, 1995; Gobert, Sao Pedro, Baker, Toto & Montalvo, 2012; Markman 1977; Stevens & Thadani, 2007) Thus, an unintended contribution of the present study is empirical evidence that individuals may not be able

PAGE 130

130 to provide reliable data regarding their own learning and that studies should not be designed to rely only on se lf reported measures when i t comes to assessing learning Situational interest The findings from the present study indicated participants had significantly higher situational interest when the instructor was present in both the easy topic video and the di fficult topic video. Generally, participants who watched the easy and difficult topic videos with instructor presence were willing to watch more videos like the ones they saw in the study s ituational interest intent to watch more videos. This finding is important, as including situational interest and potentially even result in longer term changes to their situational in terest in the autonomous and self regulated online learning as a result of watch ing videos that they find en gaging S ocial presence The findings from the current study also indicated having an instructor on the screen contributed to ced perception of the presence, a nd this finding applied to both the easy and difficult topic videos. When the instructor was present in the videos participants agreed more with the social presence the instructor was presen The high level of perceived

PAGE 131

131 s ocial presence when instructor was present could also help explain the enhanced perceived learning and s atisfaction ( Kizilcec et al., 2015) In this study, v ideos with instructor presence elicited positive responses to the social presence statements from the participants possibly due to th e social cues in the instructor present videos. The non verbal socia l cues provided by the instructor such as interaction with the content likely contributed to an improved socio emotional reaction in the learners (Bhat et al., 2015; Cui et al., 2013; Krmer & Bente, 2010) and thus in t he video This explanation ended question that elicited their perception toward instructor presence. For example, one participant reported the usefulness of non I am ab le to see his hand movements and gestures, making it easier for me to maintain my focus and attention on what he is saying and I feel less distracted. It also helps me feel more connected and learn better than to try to follow along with a voice that I can not see Other participants positively commented on the social connection established by the Seeing the instructor gave a better connection between myself and the instructor, and therefore promoted more concentratio n on the topic and a better learning experience the learning experience resemble one that occurs in a classroom, based on two participants The instructor on the screen made it feel more personal and like I was sitting in a cl When I saw the instructor, I

PAGE 132

132 felt like I was in a regular classroom. It helped me stay focused on the material as I saw the teacher and what he was pointing at. Perceptions of instructor p resence I nstructor presence was found in this study to have a significant positive effect on lowering two of the four cognitive load types learning, satisfaction, and situational interest all of which are essential factors that influence learning engagement and interest (e.g., Bandura & Cervo ne, 1986; Kim, Kim, & Wachter, 2013) According to the responses to the two qualitative questions, participants for both the easy and difficult topic videos. itive perception of the instructor was s to the open ended question, Having said this, seeing the in structor was not unanimously preferred by all learners in our study. Although a high percentage of positive feelings toward t he instructor was identified, a few participants thought the instructor was frustrating and distracting in both the easy and diffic ult topic videos and commented the instructor was unnecessary or they preferred not to see the instructor Based on this finding, instructional video designers could consider providing the option of hiding the instructor in the video for whom instructor pr esence is not favorable, instead of offering a video of instructor in instructional videos to all learners. Visual Attention Distribution Predicted Learner Perception The current study is perhaps one of the first studies that used eye tracking technology to explore the effect of instructor presence on visual attention distribution in instructional videos. Eye tracking measures such as dwell time, fixation count, and

PAGE 133

133 number of transitions between areas of interest in the visual field were able to illustrat e the attentional dynamics as the learner s distributed attention between the content portion of the video and the picture in picture video of the instructor. In the current on in important ways, as evidenced by a relatively high percentage of fixation count s and dwell time o n the instructor interest area, as compared to the instructor absent version of the videos. Also, the content area received less attention when instructor was present. The findings on the considerable amount of attention to the instructor is consistent with the visual attention distribution data reported in previous studies, such as Wang & Antonenko (2017) and Kizi lcec and colleagues (2015) both of which d emonstrated that participants allocated significant amount of attention to the instructor on the screen. The high amount of visual attention given to the instructor can possibly be explained by the well known finding that face relevant stimuli can assist p eople in seeking social cues such as gaze, facial expression, and so on, and so people are attracted to face as a source for important information that facilitate s social interaction and communication (Farroni et al., 2005) The current study also indicated that t he more participants paid attention to the instructor on the screen, the more satisfied they were with the instructor present videos. For instructor present easy topic video, the fix ation count on the instructor interest area And for the instructor present difficult topic video the percentage of fixation count and the percentage of dwell time on the instructor interest area with the video. The response to the open ended question Please explain what you

PAGE 134

134 t hink about seeing the instructor in the video, compared to not seeing the instructor also indicated the relationship between visual attention allocation to the instructor and eeing the instr uctor on the screen made it feel like he could actually see me. It made me pay attention more because it was as if he was watching me. Also, it made the vi deo more entertaining and This finding implies the visual attention instructor on the screen received from the learners and it suggests includi ng instructor presence based on the fact that paying attention to the instructor on the screen predicts partic current study is the first study that identifies this relationship between visual attention given to the instructor and enhanced satisfaction with the video. Cognitive Dynamics Predicted Learning To our knowledge, t he current study is the first study that examined the influence of instruct or presence on EEG based cognitive dynamics in instructional videos with and without instructor presence they viewed instruct ional videos on easy and difficult content. EEG findings from th is study indicated that instructor presence led to significantly different cognitive dynamics (mainly manifested in the theta frequency band) in easy and difficult to pic videos. The EEG data i ndicated that i nstructor present video resulted in higher theta power at C4 sensor compared to the instructor absent video for the easy topic; and for the difficult topic video, having the instructor on the screen decreased the theta power at C4 senso r amo ng t he participants, as compared to the instructor absent version of the video.

PAGE 135

135 Theta activity is most prevalent when a person is cognitively engaged (Klimesch et al., 2005) It has been demonstrated that brain oscillations in the theta frequency band are involved in active maintenance and recall of working memory represe ntations ( Jensen & Tesche, 2002) Using different experimental paradigms in previous studies, t het a activity has been found to be associated with cognitive load, that is, theta frequency band power increases as cognitive load increases (Gevins et al., 1998; Jensen & Tesche, 2002; Antonenko & Niederhauser, 2010; Kumar & Kumar, 2016) For example, Gevins and colleagues (1998) had eight parti cipants perform high moderate and low load working memory tasks, and found t heta activity increased with increasing working memory load. In a subsequent study, Jensen and colleagues (2002) recorded neuromagnetic responses from 10 subjects performing the Sternberg task. Subjects were required to retain a list of 1, 3, 5 or 7 visually presented digits during a 3 s retention period. Similar finding was obtained, that is, activity in the theta frequency band increased parametri cally with the number of items retained in working memory. The findings from the current study showed theta frequency band power increase d in the easy topic video when instructor was present, which could indicate an increase in working memory load when th e instructor was present in the easy topic video. Th is study also showed that a process measure such as EEG based cognitive dynamics can predict learning. For the easy topic video, theta power at F4 sensor and the average theta power at F3, F4, and Fz (i.e ., three frontal lobe sensors) were both In other words, participants who experienced higher increase in theta power in the frontal lobe exhibited lower scores on the retention test. I t is possible the participants who watched the

PAGE 136

136 instructor present easy topic video experienced higher extraneous cognitive load and thus poorer retention performance. Also, it needs to be pointed out the self report cognitive load measure did not identify any difference in cognitive load when the instructor was present or absent in the easy topic video. Thus, cognitive dynamics measure based on theta EEG activity can be a more reliable measure of extraneous load that is the type of cognitive load that hind ers rather than facilitates learning On the other hand, theta frequency band power decreased in the difficult topic video when instructor was present, which could indicate a decrease in cognitive load when the instructor was present in the difficult topi c video as compared to the version with instructor absent It has also been found when the instructor was present in the difficult topic video, participants self perceived a lower intrinsic load a nd extraneous load The learning measure also indicated part icipants who watched the difficult topic video with instructor present performed better on the transfer test. It is possible that learners who watched the instructor present difficult topic video experienced lower level of intrinsic load and extraneous loa d which resulted in better transfer performance. Taken together, for the easy topic video, instructor presence led to higher level of theta activity which could be associated with higher cognitive load. Higher theta power in the frontal lobe was associa ted with poorer performance on the retention test for the easy topic For the difficult topic, instructor presence resulted in lower theta power, which could be an indicator of lower cognitive load. Actually, participants reported lower intrinsic and extra neous cognitive load and they performed better on the transfer test when instructor was present in the difficult topic video. Based on these findings, cognitive dynamics based on EEG theta activity is a more reliable measure of cognitive

PAGE 137

137 load compared to s elf report measure of cognitive load, and it is useful in helping us understand why learners end up with a higher or lower scores on the tests of learning (i.e., retention and transfer). Moderating Effects of Individual Differences The second goal of th is study was to investigate the moderating effects o f individual differences in working memory c apacity and inhibitory control in easy and difficult topic videos when instructor was present or absent Despite the hypotheses, the current study did not identif y any moderating effects of individual differences in working memory capacity and inhibitory control on learner perceptions, visual attention distribution, or cognitive dynamics, in easy and difficult topic videos with or without instructor presence. Howe ver, the working memory capacity score was found in this study to predict the retention test score for the easy topic. Participants with higher working memory capacity scores performed better on the retention test for the easy topic. Besides, the inhibitor y control score was found to predict the transfer test scores for the difficult topic. People with higher inhibitory control scores outperformed those with lower inhibitory control scores on the transfer test for the difficult topic. Working Memory Capaci ty Score Predicted Retention Performance Working memory capacity inf (Conway & Engle, 1994) Despite the hypothesis, in this study, we did not identify any mediating effect of working memory capacity on learning a nd learner perception in instructional videos with and without instructor presence. Based on the eye tracking data generated from the current study, we did not see that working memory capacity influenced tion to the instructor interest area and

PAGE 138

138 content interest area, either by more attention to the instructor interest area or more transitions between the two interest areas (i.e., a proxy for split attention). The findings from the current study indicated that working memory capacity score score for the easy topic video controlling for instructor presence From a C ognitive T heory of M ultimedia L earning (CTML) perspective, successful mult imedia learning requires three process es: selecting, organizing and inte grating relevant information (Mayer, 2014) results suggest ed that one or more of these processes ma y be more easily disrupted in learners with lower working memory capacity In the current study, t he participants were expected to maintain many conceptual information from the instructional videos, in order to do well on the retention tests. For example, the easy topic video focused on the terminology associated with experiments and observational studies, and conceptual information such as the definition of factor, level of treatment needed to be organized by the working memory and integrated with the long term memory while participants watched the easy topic video. As a result, presenti ng lower working memory capacity learners with mul tiple sources of visual information in the instructional v ideo may impair one or mor e of these essential processes, and the refore leading to a lower retention performance for the easy topic video. This finding on the positive influence of working memory capacity on retention from instructional video is consistent with findings from at least two previous studies. Lusk and colleagues (2009) explored the influe nce of individual differences in working memory capacity on learning from a multimedia historical inquiry unit in segmented or unsegmented condition Data generated from the recall measure showed working

PAGE 139

139 memory capacity had a significant, positive effect in the unsegmented condition Moreover, in the same study, the researchers found that segmentation of the unit specifically improved recall performance for participants with low working memory capacity score. Similarly, Sanchez & Wiley (2009) examined the influence s of working memory capacity in learnin g from paginated text and non paginated scrolling text. In their study, participants who varied in WMC read a complex illustrated text about ice age Learning was measured by response to one free recall question hat causes ice age Results showed lear ners with higher working memory capacity outperformed lower WMC individuals in the scrolling format and higher WMC individuals learned equivalently across scrolling or paginated presentations, suggesting a resiliency to competing demands in the scrolling c ondition. So, it is reasonable to speculate that in the current study, participants with higher working memory capacity were better at processing information and organizing it into existing mental models, and thus excel led at recall ing and recognizing more information from the easy topic video. To further explore this hypothesis, future studies can test if breaking down the video into smaller segments can improve retention of information among learners with lower working memory capacity, as compared to the unsegmented version of the video. Inhibitory Control Score Predicted Transfer Performance Inhibitory control represents the ability to selectively attend to relevant information and suppress attention to irrelevant stimuli while focusing on the task goal ( Rothbart & Posner, 1985 ; Diamond, 2013) Despite the original hypothesis in the literature review in the current study, we did not identify any mediating effect of inhibit ory control on learning and learner perception in instructional videos with and

PAGE 140

140 without instructor presence. Based on the eye tracking data generated from the current study we did not see that inhibitory con trol influence d patterns of visuals attention distribution to the instructor interest area and content interest area either by more attention to the instructor interest area or more transitions between the two interest areas (i.e., a proxy f or increased split attention) Instead, data from the present study showed that inhibitory control ability significantly positively sfer information for the difficult topic video, controlling for instructor presence This finding is not very surprising, considering working memory capacity positively predicted retention performance for the easy topic video. It is likely that participants with higher inhibitory control used their cognitive resources more efficiently, t hus resulting in enhanced comprehension of the difficult material, whereas learners with lower inhibitory control may have f ou nd it more challenging to process a high amount of complex information from the difficult topic video. Similar findings were obtai ned from a previous study conducted by Homer and Plass (2014) who investigated the inte raction between instructional format and Participants were assigned to learn with a web based simulation of an intrinsically complex topic (i.e., ideal gas law) that varied in instructional format (exploratory simulation or worked examples). In the simulation with worked examples, learners navigated the simulation by following step by step instructions provided by an expert. Results indicated that the exploratory simulation fac ilitated transfer for students with higher levels of inhibitory control whereas students with lower levels of inhibitory control benefited more from the guided simulation with worked examples in learning transfer.

PAGE 141

141 T he current study implie s that learners wi th high er inhibitory control benefit more from the instructional video on difficult content. Thus, it is reasonable to hypothesize that inhibitory control is an important individual difference variable mediating college in an instructional video Thus, additional instructional support can be provided to learners with lower inhibitory control to better facilitate the processing of complex information For example, learners can be provided with some pre training (Mayer et al., 2002) at the beginning of the video to as sist the learners in the knowledge construction process It is also suggested the videos can include visual cues to important concepts in the instructional video, which can serve as a cognitive guide in facilitating the learners in understanding the inform atio n (Mautone & Mayer, 2001) Taken together, in the current study, learning from the instructional videos is significa ntly predic ted by individual differences in working memory capacity and inhibitory control. Working memory capacity score was found to be a positive predictor of retention performance for the easy topic video while inhibitory c ontrol score positively predi cted transfer performance for the difficult topic video. Th is study emphasized the importance of considering individual differences in working memory capacity and inhibitory control when designing instructional videos. The instructional videos are largely used by college students who vary in working memory capacity and inhibitory control. Thus, additional instructional support can be provided for learners who have low working memory c apacity and inhibitory control, for example, segmenting the video into sma ller units, in order to allow those learners to process the information in the instructional video more efficiently.

PAGE 142

142 Implications First, t he study provided evidence that supports including the instructor in online instructional video. Instructor presence of information from the difficult topic video. Also, instructor presence enhanced situational interest perceived learning, and their perceptions nce. Instructor presence was also found to lower self reported intrinsic and extraneous load for the difficult topic video. The practical implication of the current study is that instructional video designers could consider including the instructor in inst ructional videos, especially for the videos that deal with difficult content. The practitioners also need to realize that merely including an instructor in the video will not necessarily produce the positive effects on learning and learner perception. Inst ructors in online videos should consider replicating the practices of the instructor in the current study, for example, delivering a sufficient amount of non verbal cues such as facial expression, body gesture, and eye contact which help engage the learner s. Future research could also test the efficacy of instru ctor presence in other subjects besides Statistics. It would also be useful to examine the effect of adaptive instructor presence, specifically when the instructor frame is presented not all of the t ime but only during the times when the instructor provided nonverbal cues and signaling can enhance the processing of learning content. Second, individual differences in working memory capacity and inhibitory control have been found to affect learning fro m video. The practical implication is instructional videos can be designed to accommodate individual differences in working memory capacity and inhibitory control. To be specific, as processing information from the instructional video is more difficult for students who have low inhibitory control and

PAGE 143

143 working memory capacity, additional instructional support such as signaling and pre training can be provided to these learners. Besides, i n addition to working memory capacity, there are other cognitive differe nces that vary among learners that could have As millions of learners representing a variety of individual differences use instructional videos today, it is also imperati ve to understand how learners with individual differences respond to instructor presence in instructional videos and learn with such videos and how the design of videos can be improved to accommodate the needs of a wider range of learners. Future research should consider examining the influence of other individual differences and examine how these individual difference variables impact learning from instructional videos. The findings will offer important information on how individual differences affect lear ning and provide implications for future personalized instructional content, which is a common practice with adaptive systems today. Last but not least, neurocognit ive and psychophysiological tools such as EEG and eye tracking are useful in understanding the process of learning during watching an instructional video with or without instructor presence An implication of the current study is that e ducational researche rs can apply these neurocognitive and psychophysiological tools to multiple contexts of learning, including but not limited to instructional videos, to understand the underlying process of learning and better explain the outcome measures such as learning a nd learner perception. For future studies, i t could also be helpful to investigate the cognitive dynamics on the temporal scale from

PAGE 144

144 the start of the video to the end, in order to explore the influence of instructor presence as time goes. Limitations T he re were certain limitations to the study design that may have influenced the reliability and generalizability of the findings. First, the study was conducted in a highly controlled lab setting with the use of an eye tracker, which could have influenced par on the materials after watching the videos, and this possibly increased their engagement with the video content. Moreover, the participants in the current study were not a llowed to take notes or pause videos during the viewing session which is not representative of authentic video viewing contexts. Last ly the majority of the participants were female and the instructor in the video was a male. It is possible we will identi fy different findings on the influence of instructor presence if recruiting a higher percentage of male participants, or including a female instructor in the video among a group of female participants. Therefore, f uture studies could consider having a more balanced participant sample with males and females or designing video conditions that include both male and female instructors. Conclusion This study explored the influence of instructor presence on learning, learner perception, visual attention distrib ution, and cognitive dynamics, when viewing instructional videos at easy and difficult levels of content complexity. Findings suggest that while the picture in picture video of the instructor attracted significant levels of visual attention across the enti re viewing session, instructor presence did not result in increased retention for easy or difficult learning content but increased transfer for

PAGE 145

145 difficult content. Moreover, instructor presence produced a significant positive effect on increasing participan situational interest as well as lowering intrinsic load and extraneous load for the difficult topic video all of which are essential factors that contribute to learner engagement in the autonomous and self regula ted online learning environment.

PAGE 146

146 APPENDIX A EASY TOPIC VIDEO DETAILS Terminology associated with experiments and observational studies Experiment (e.g. compare effects of anti depressants, compare effects of brand of gas, etc.) A design in whi ch the treatments are being actively imposed on the experimental units. Observational study A design in which only the researcher only _____________ the response variable. Response variable (RV) Th e response variable is the outcome of interest of a study. Explanatory variables (EV, a.k.a. factors) The variables that we suspect affect the response variable. Levels The levels are the subcategories of the factors. Treatments

PAGE 147

147 One factor: The le vels and treatments are the same. Two factors: The treatments are the factor level combinations. Subjects The experimental units are the people or things in the study. Replications The number of replications equals the number of times each treatmen t is being applied.

PAGE 148

148 APPENDIX B DIFFICULT TOPIC VIDEO DETAILS Rationale of the Analysis of Variance (ANOVA) If the variation BETWEEN the sample means ( MSG ) is ___________________________ than the variation WITHIN each sample ( MSE ), i.e. if it implies that the population means significantly differ.

PAGE 149

149 APPE NDIX C PRE SCREENING SUR V EY What is your age? ________ years old I identify my gender as _______________ What is your undergraduate major (if declared)? ___________________ _____ How do you identify your race/ethnicity? White/Caucasian Black/African American Asian/Pacific Islander Native American Hispanic/Latino Other Is English your first language? Yes No

PAGE 150

150 Do you have normal or corrected to normal vision? Yes No Do you h ave a history of brain trauma or neurological disorders? Yes No Do you have autism? Yes No Are you taking depression and anxiety medications or have you taken these medications in recent 3 months? Yes No How familiar are you with terminology associated with experiments and observational studies on a scale of 1 5? 1: N ot familiar at all 5: V ery familiar How familiar are you with Analysis of Variance (ANOVA) on a scale of 1 5? 1: Not familiar at all

PAGE 151

151 5: Very familiar

PAGE 152

152 APPENDIX D LEARNING TESTS Easy t opic Terminology associated with experiments and observational studies Retention questions: 1. Which statement is true about an experiment? A. In an experiment, the researcher observes a group of subjects without actually doing anything to the subject s B. In an experiment, treatments are being actively imposed on the experimental units C. In an experiment, no data is collected from the subjects D. All of the above 2. Which statement is true about an observational study? A. In an observ ational study, treatments are actively imposed on the experimental units B. In an observational study, C. In an observational study, the researcher only observes the experimental units and records the r esponse variable D. All of the above 3. What is the relationship between explanatory variable and response variable?

PAGE 153

153 A. Response variable may affect explanatory variable B. Explanatory variable may affect response variable C. Explanator y variable and response variable are independent of each other 4. In an experiment with two explanatory variables, each explanatory variable is known as a ________. A. Level B. Factor 5. A researcher examined the influ ence of exercise (low, moderate, or high) on health score. How many treatment(s) does this study have? A. 1 B. 2 C. 3 6. In an experiment, there are two explanatory variables and each explanatory variable has two subcategories. Each sub category is known as a ________ of an explanatory variable.

PAGE 154

154 A. Level B. Factor The following set of questions (7 12) will focus on the following scenario: A researcher examined the influence of diet (healthy or unhealthy) and exercise (low, moderate, or high) on health score. Ninety UF students were randomly assigned to each cell. 7. Specify the subjects of this study. UF students or 90 UF students 8. Specify the explanatory variable(s). Diet and exercise 9. Specify the response variable(s). Health score 10. Specify the number of levels for the factor(s). Diet has two levels and exercise has three levels. 11. Specify the number of treatments. 6

PAGE 155

155 12. Specify the number of replications. 15 Transfer questions: 1. Students are ran domly assigned to one of two statistics classes. One class was exposed to an applied hands on approach, whereas the other one used a traditional lecture approach. The outcome consisted of scores on the final exam. This is an example of an ______ study. A. Experimental B. Observational 2. A study in California showed that students who learn a musical instrument have higher GPAs than students who do not, 3.59 to 2.91. Of the music students, 16% had all As, compared with only 5% among t he students who did not learn a musical instrument. This is an example of an______ study. A. Experimental B. Observational

PAGE 156

156 3. A researcher is curious if the number of hours spent doing homework has an effect on the grade students earn o n an exam. In this case, the number of hours spent doing homework is the ______ variable, and the grade on the exam is the ______ variable. A. explanatory; response B. explanatory; explanatory C. response; explanatory D. response; response ow 4. A researcher carried out an experimental study to examine the effectiveness of GRE prep programs (Powerscore or Kaplan) and the method of delivery (in person, online, or blended) on GRE verbal score. In the study, there are ____ factors. A. 1 B. 2 C. 3 D. 6 E 5. A researcher carried out an experimental study to examine the effectiveness of GRE prep programs (Powerscore or Kaplan) and the method of delivery (in person, online, or blended) on GRE verbal score. In the study, there are ____ treatments.

PAGE 157

157 A. 1 B. 2 C. 3 D. 6 Difficult topic Rationale of the ANOVA Retention questions: 1. What does ANOVA stand for? A. Analysis of Variance B. Analysis of Variation C. Analyzing Ordinal Variation D. At Nor thern Virginia 2. The purpose of running a one factor ANOVA on three groups is to _________. A. Compare the means between three groups

PAGE 158

158 B. Find the similarities between three groups C. Characterize the overlapped portions between three g roups D. Find which two groups have different means 3. In a one factor ANOVA of four samples, the null hypothesis asserts that _________. A. All the means are equal among the four groups B. The means of two specific groups are equal C. Al l the means are different among the four groups D. At least one group mean is different from another group mean 4. In a one factor ANOVA of four samples, the alternative hypothesis asserts that _________. A. All the means are equal among the four groups B. The means of two specific groups are equal C. All the means are different among the four groups D. At least one sample mean is different from another sample mean

PAGE 159

159 5. The following plot shows the average weight loss of fou r different exercise programs. The way to decide the variation between the sample means is to _________. A. Compare the means of each sample which are shown by the red dots on the bars B. Examine the overlapping area of the bars C. Compare the height o f bars D. Compute the length of the connected line E. 6. The following two plots show the average weight loss of four different exercise programs. Which one of the following statements is true? Plot A Plot B

PAGE 160

160 A. The data in plot A have a higher variation within each sample B. The data in plot B have a higher variation within each sample C. The data in plot A and B have the same variation within each sample D. Insufficient information to determine E. 7. The F test statistic is defined as _________. A. The variation between the sample means divided by the variation within each sample B. The variation within each sample divided by the variation between the sample means C. The variation between the sample means plus by the var iation within each sample D. The variation between the sample means minus by the variation within each sample 8. The following two plots show the average weight loss of four different exercise programs. Assuming the variation between the s ample means is the same for these two plots, which one of the following statements is true?

PAGE 161

161 Plot A Plot B A. The data in plot A have a higher F test statistic than the data in plot B B. The data in plot A have a lower F test statistic than the d ata in plot B C. The data in plot A have the same F test statistic as the data in plot B D. Insufficient information to determine E. Transfer questions: 1. When will a one factor ANOVA be appropriate? A. When assessing if three instructi onal methods are associated with different mean test scores B. When assessing if students in grade levels 6, 7, and 8 have significantly different average reading comprehension levels

PAGE 162

162 C. When assessing if different ethnicity groups display significant mean differences on the GRE quantitative section D. All of the above E. 2. Data for three samples (training program A, training program B, training program C) were analyzed using a one factor ANOVA and the null hypothesis was rejected. Which one of the following statements is true? A. The group mean of training program A is different from that of training program B B. The group mean of training program B is different from that of training program C C. The group means of training program A, B, and C are different from each other D. At least one group mean is different from another group mean 3 Data for three samples are analyzed using a one factor ANOVA. If means of the three samples are equal, what is the value of the variation between the sample means? A. 0 B. 1 C. 2

PAGE 163

163 D. 3 4. Which one of the following statements is true for the dataset below? A. There is a high variation within each sample, and no variation between the sample means B. There is a high variat ion between the sample means and no variation within each sample C. There is a high variation between the sample means and a high variation within each sample D. There is a no variation between the sample means and no variation within each sample E. I d 5 Which one of the following statements is true for the dataset below?

PAGE 164

164 A. There is a high variation within each sample, and no variation between the sample means B. There is a high variation between the sample means and no variation within each sample C. There is a high variation between the sample means and a high variation within each sample D. There is a no variation between the sample means and no variation within each sample E.

PAGE 165

165 APPENDIX E LEA R NER PERCEPTIONS SUR V EY 1. In the video I just finished watching, I invested 1. very, very low mental effort / 2. very low mental effort / 3. low mental effort / 4. rather low mental effort / 5. neither low nor high mental effort / 6. rather high mental effort / 7. high mental e ffort/ 8. very high mental effort / 9. very, very high mental effort 2. The topics covered in the video I just finished watching was 1. very, very easy / 2. very easy / 3. easy / 4. Rather easy / 5. neither easy nor difficult / 6. rather difficult / 7. di fficult / 8. very difficult / 9. very, very difficult 3. Learning from the video that I just finished watching was 1. very, very easy / 2. very easy / 3. easy / 4. Rather easy / 5. neither easy nor difficult / 6. rather difficult / 7. difficult / 8. very difficult / 9. very, very difficult 4. When watching this video, I concentrated 1. very, very little / 2. very little / 3. little / 4. Rather little / 5. neither little nor much / 6. rather much / 7. much / 8. very much / 9. very, very much 5. On a scal e from 1 to 5, please indicate how much you have learned from the video 1. Did not learn anything; 5. Learned a great deal. 6. On a scale from 1 to 7, rate your satisfaction regarding the instructor presence in the video

PAGE 166

166 1. Extremely dissatisfied; 7. Ex tremely satisfied. For questions 7 12, please indicate how much you agree or disagree with each statement: 7. I am willing to watch more video s like this because it is exciting and relevant. Strongly disagree; Disagree; Neither agree nor disagree; Agree ; Strongly Agree. 8. I felt like the instructor was in the same room as me. Strongly disagree; Disagree; Neither agree nor disagree; Agree; Strongly Agree. 9. I felt that the instructor was very detached in his interactions with me. Strongly disagree; D isagree; Neither agree nor disagree; Agree; Strongly Agree. 10. I felt that the instructor was aware of my presence. Strongly disagree; Disagree; Neither agree nor disagree; Agree; Strongly Agree. 11. I felt that the instructor was present. Strongly disa gree; Disagree; Neither agree nor disagree; Agree; Strongly Agree. 12. I felt that the instructor remained focused on me throughout our interaction. Strongly disagree; Disagree; Neither agree nor disagree; Agree; Strongly Agree. Question 13 14 apply to t he video with instructor presence.

PAGE 167

167 13. Please explain what you think about seeing the instructor in the v ideo, compared to not seeing him. ______________________________________________________________________ _____________________________________________ _________________________ ______________________________________________________________________ 14. Do you think the instructor in the video is _______ (Select all adjectives that apply)? Helpful Entertaining Useful Engaging Frustrating Distracting Ann oying Other: ____________

PAGE 168

168 APPENDIX F INFORMED CONSENT Protocol Title : Examining Influence of Instructor Presence in Instructional Videos: An Individual Difference Perspective Please read this consent document carefully before you decide to partici pate in this study Purpose of the research study : The purpose of the study is to evaluate the influence of instructor presence in Study Edge instructional videos on learning, cognitive dynamics, visual attention distribution, and learner perceptions and also how these effects are moderated by individual differences such as working memory capacity and inhibitory control. What you will be asked to do in the study : You will be first asked to complete two tasks that assess cognitive process such as WMC (via the Automated Operation Span t ask) and inhibitory control (Flanker In hibitory Control and Attention t est) You will be then asked to watch two Study Edge videos (Terminology associated with experiments and observational studies, and rationale of the ANOVA ) with or without instructor present on the screen. The two videos last 8 minutes in total. After each video, you will respond to a few questions on your perceptions of the video and the instructor. While you watch the two videos, you will be asked to wear an electroencephalography ( EEG ) headset that records your brain activity It doesn t require any additional skin preparation or applying conductive gel to the scalp or electrodes The EEG equipment is non invasive and has been used in multiple research st udies in higher education and K 12 education settings During the study you will also rest your chin on a chin rest and the researcher will collect

PAGE 169

169 eye movement data which will reveal where you focus on the screen As you finish watching the tw o videos, you will respond to a learning test to assess your retention and transfer of knowledge. Time required : Approximately 60 minutes. Risks and Benefits : No risks are anticipated While there are no direct benefits for you, the data collected may have signi ficance for the design of Study Edge videos which are used by thousands of students at the University of Florida Compensation : You will receive $15 for participating in this study Confidentiality : Your identity will be kept confidential to the extent provided by law 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 any time without consequence You do not have to answer any questions you do not want to answer Whom to contact if you have questions about the study : Principal Investigator : Jiahui Wang, G518F Norman Hall, School of Teaching and Learning, Un iversity of Florida. Phone: 352 222 0636. Email : jwang 01 @ufl edu

PAGE 170

170 Whom to contact about your rights as a research participant in the study : UFIRB Office Box 112250, University of Florida Gainesville FL 32611 2250; ph 392 0433. I have read the procedure outlined above I voluntarily agree to participate in this study and have received a copy of this description ______________________________ __________________________________ Participant Date Researcher Date

PAGE 171

171 LIST OF REFERENCES Heaton, R. K. (2013). NIH toolbox cognitive function battery (cfb): composite scores of crystallized, fluid, and overall cognition. Monographs of the Society for Research in Child Development 78 (4), 119 132. Alibali, M. W., & Nathan, M. J. (2012). Embodiment in mathematics teaching and s. Journal of the Learning Sciences 21 (2), 247 286. Allen, I. E., Seaman, J., Poulin, R., & Straut, T. T. (2016). Online report card: Tracking online education in the United States. Sloam Consortium 1 4. ognition of inconsistencies in science texts: Multiple measures of comprehension monitoring. Applied Cognitive Psychology 9 261 272. Antonenko, P. D., & Niederhauser, D. S. (2010). The influence of leads on cognitive load and learning in a hypertext envi ronment. Computers in Human Behavior 26 (2), 140 150. https://doi.org/10.1016/j.chb.2009.10.014 Antonenko, P., van Gog, T., & Paas, F. (2014). Implications of neuroimaging for educational research. In Handbook of research on educational communications and technology (pp. 51 63). New York: Springer. Argyle, M. (2013). Bodily Communication Routledge. Atkinson, R. C., & Shiffrin, R. M. (1968). Human memory: A proposed system and its control processes. Psychology of Learning and Motivation 2 89 195. Ayres, P ., & Sweller, J. (2014). The split attention principle in multimedia learning. In R. E. Mayer (Ed.), The Cambridge Handbook of Multimedia Learning (pp. 206 226). Cambridge University Press. Ayres, P. (2006). Impact of reducing intrinsic cognitive load on l earning in a mathematical domain. Applied Cognitive Psychology 20 (3), 287 298. https://doi.org/10.1002/acp.1245 Baddeley, A. D., & Hitch, G. (1974). Working memory. The Psychology of Learning and Motivation 8 47 89. Baddeley, A. (1992). Working memory. Science 255 (5044), 556. Banas, S., & Sanchez, C. A. (2012). Working memory capacity and learning underlying conceptual relationships across multiple documents. Applied Cognitive Psychology 26 (4), 594 600. https://doi.org/10.1002/acp.2834

PAGE 172

172 Bandura, A., & C ervone, D. (1986). Differential engagement of self reactive influences in cognitive motivation. Organizational Behavior and Human Decision Processes 38 (1), 92 113. https://doi.org/10.1016/0749 5978(86)90028 2 Basar, E. (1999). Brain function and oscillati ons. Berlin: Springer. Bayliss, D. M., Jarrold, C., Gunn, D. M., & Baddeley, A. D. (2003). The complexities of complex span: explaining individual differences in working memory in children and adults. Journal of Experimental Psychology: General 132 (1), 71 92. https://doi.org/10.1037/0096 3445.132.1.71 Baylor, A. L., & Kim, Y. (2005). Simulating instructional roles through pedagogical agents. International Journal of Artificial Intelligence in Education 15 (1), 95 115. https://doi.org/10.1007/BF02504991 Bha t, S., Chinprutthiwong, P., & Perry, M. (2015). Seeing the instructor in two video styles: Preferences and patterns. In Proceedings of the 8th International Conference on Educational Data Mining (pp. 305 312). Bleckley, M. K., Durso, F. T., Crutchfield, J. M., Engle, R. W., & Khanna, M. M. (2003). Individual differences in working memory capacity predict visual attention allocation. Psychonomic Bulletin & Review 10 (4), 884 889. https://doi.org/10.3758/BF03196548 Boucheix, J., & Lowe, R. K. (2010). An eye t racking comparison of external pointing cues and internal continuous cues in learning with complex animations. Learning and Instruction 20 (2), 123 135. https://doi.org/10.1016/j.learninstruc.2009.02.015 Bransford, J. D., & Schwartz, D. L. (1999). Rethinki ng transfer: A simple proposal with multiple implications. Review of Research in Education 24 (1), 61 100. Canham, M., & Hegarty, M. (2010). Effects of knowledge and display design on comprehension of complex graphics. Learning and Instruction 20 (2), 155 166. https://doi.org/10.1016/j.learninstruc.2009.02.014 Carpenter, P. A. & Shah, P. (1998). A model of the perceptual and conceptual processes in graph comprehension. Journal of Experimental Psychology: Applied 4 (2), 75 100. Case, R., Kurland, D. M., & Go ldberg, J. (1982). Operational efficiency and the growth of short term memory span. Journal of Experimental Child Psychology 33 (3), 386 404. https://doi.org/10.1016/0022 0965(82)90054 6 Chen, C., & Wu, C. (2015). Effects of different video lecture types o n sustained attention, emotion, cognitive load, and learning performance. Computers & Education 80 108 121. https://doi.org/10.1016/j.compedu.2014.08.015

PAGE 173

173 Christ, S. E., Holt, D. D., White, D. A., & Green, L. (2007). Inhibitory control in children with a utism spectrum disorder. Journal of Autism and Developmental Disorders 37 (6), 1155 1165. https://doi.org/10.1007/s10803 006 0259 y Church, R. B., Ayman Nolley, S., & Mahootian, S. (2004). The role of gesture in bilingual education: does gesture enhance le arning. International Journal of Bilingual Education and Bilingualism 7 (4), 303 319. https://doi.org/10.1080/13670050408667815 Cierniak, G., Scheiter, K., & Gerjets, P. (2009). Explaining the split attention effect: Is the reduction of extraneous cognitiv e load accompanied by an increase in germane cognitive load? Computers in Human Behavior 25 (2), 315 324. https://doi.org/10.1016/j.chb.2008.12.020 Clark, R. C., & Mayer, R. E. (2016). E learning and the science of instruction: Proven guidelines for consum ers and designers of multimedia learning John Wiley & Sons. Cohen, J. (1992). A power primer. Psychological Bulletin 112 (1), 155 159. https://doi.org/10.1037/0033 2909.112.1.155 Conway, A. R., Kane, M. J., Bunting, M. F., Hambrick, D. Z., Wilhelm, O., & Engle, R. Psychonomic Bulletin & Review 12 (5), 769 786. https://doi.org/10.3758/BF03196772 Conway, A. R. A., & Engle, R. W. (1994). Working memory and retrieval: a resource de pendent inhibition model. Journal of Experimental Psychology General 123 (4), 354 373. https://doi.org/10.1037/0096 3445.123.4.354 Cowan, N. (2001). The magical number 4 in short term memory: A reconsideration of mental storage capacity. Behavioral and Bra in Sciences 24 (1), 87 114. Cragg, L., & Nation, K. (2008). Go or no go? Developmental improvements in the efficiency of response inhibition in mid childhood. Developmental Science 11 (6), 819 827. https://doi.org/10.1111/j.1467 7687.2008.00730.x Creswell, J. W. (2013). Research design: Qualitative, quantitative, and mixed methods approaches. Sage publications. Cronbach, L. J., & Snow, R. E. (1977). Aptitudes and instructional methods: A handbook for research on interactions. Irvington. Crook, C., & Schofie ld, L. (2017). The video lecture. Internet and Higher Education 34 (May), 56 64. https://doi.org/10.1016/j.iheduc.2017.05.003

PAGE 174

174 review of social presence theory and its instructional design implications for future trends. Education and Information Technologies 18 (4), 661 685. https://doi.org/10.1007 /s10639 012 9192 1 Annual Review of Psychology 66 115 42. https://doi.org/10.1146/annurev psych 010814 015031 Daneman, M., & Carpenter, P. A. (1980). Individual differ ences in working memory during reading. Journal Of Verbal Learning And Verbal Behavior 19 (4), 450 466. https://doi.org/10.1016/S0022 5371(80)90312 6 de Koning, B. B., Tabbers, H. K., Rikers, R. M. J. P., & Paas, F. (2010). Attention guidance in learning f rom a complex animation: Seeing is understanding? Learning and Instruction 20 (2), 111 122. https://doi.org/10.1016/j.learninstruc.2009.02.010 Delorme, A., & Makeig, S. (2004). EEGLAB: An open source toolbox for analysis of single trial EEG dynamics includ ing independent component analysis. Journal of Neuroscience Methods 134 (1), 9 21. https://doi.org/10.1016/j.jneumeth.2003.10.009 Diamond, A. (2013). Executive functions. Annual Review of Psychology 64 135 168. Diaz, D., Ramirez, R., & Hernandez Leo, D. (2015). The effect of using a talking head in academic videos: An EEG study. Proceedings IEEE 15th International Conference on Advanced Learning Technologies: Advanced Technologies for Supporting Open Access to Formal and Informal Learning, ICALT 2015 3 67 369. https://doi.org/10.1109/ICALT.2015.89 Djamasbi, S., Siegel, M., & Tullis, T. S. (2012). Faces and viewing behavior: An exploratory investigation. AIS Transactions on Human Computer Interaction 4 (3), 190 211. Duchowski, A. T. (2007). Eye tracking m ethodology London: Springer. https://doi.org/10.1145/1117309.1117356 Eriksen, B. A., & Eriksen, C. W. (1974). Effects of noise letters upon the identification of a target letter in a nonsearch task. Perception & Psychophysics 16 (1), 143 149. https://doi. org/10.3758/BF03203267 Fan, J., McCandliss, B. D., Sommer, T., Raz, A., & Posner, M. I. (2002). Testing the efficiency and independence of attentional networks. Journal of Cognitive Neuroscience 14 (3), 340 7. https://doi.org/10.1162/089892902317361886

PAGE 175

175 F arroni, T., Johnson, M. H., Menon, E., Zulian, L., Faraguna, D., & Csibra, G. (2005). relevant stimuli: effects of contrast polarity. Proceedings of the National Academy of Sciences of the United States of America 102 (47), 17 245 50. https://doi.org/10.1073/pnas.0502205102 Fredericksen, E., Pickett, A., Shea, P., Pelz, W., & Swan, K. (2000). Student satisfaction and perceived learning with on line courses Principles and examples from the SUNY learning network. Journal of Asyn chronous Learning Networks 4 (2), 7 41. Gal, I., Ginsburg, L., & Schau, C. (1997). Monitoring attitudes and beliefs in statistics education. In The assessment challenge in statistics education (pp. 37 51). rences in cognitive processes observed with EEG. Educational Technology Research and Development 47 (3), 5 14. https://doi.org/10.1007/BF02299630 Gevins, A., Smith, M. E., Leong, H., McEvoy, L., Whitfield, S., Du, R., & Rush, G. (1998). Monitoring Working Memory Load during Computer Based Tasks with EEG Pattern Recognition Methods. Human Factors: The Journal of the Human Factors and Ergonomics Society 40 (1), 79 91. https://doi.org/10.1518/001872098779480578 Gobert, J., Sao Pedro, M., Baker, R., Toto, E., & Montalvo, O. (2012). Leveraging educational data mining for real time performance assessment of scientific inquiry skills within microworlds. Journal of Educational Data Mining 4 (1), 111 143. Goldin speaking, learning, and creating language. Annual Review of Psychology 64 257 283. https://doi.org/10.1146/annurev psych 113011 143802 Guo, P. J., Kim, J., & Rubin, R. (2014). How video production affects student engagement: An empirical study of mooc vi deos. In Proceedings of the first ACM conference on Learning@ scale conference (pp. 41 50). ACM. Hart, S. G., & Staveland, L. E. (1988). Development of NASA TLX (Task Load Index): Results of empirical and theoretical research. Advances in Psychology 52 1 39 183. Herold, J., Stahovich, T., Lin, H. L., & Calfee, R. C. (2011). The effectiveness of American Society for Engineering Education. American Society for Engineering Education. Holler, J., Shovelton, H., & Beatt ie, G. (2009). Do iconic hand gestures really contribute to the communication of semantic information in a face to face context? Journal of Nonverbal Behavior 33 (2), 73 88. https://doi.org/10.1007/s10919 008 0063 9

PAGE 176

176 Holsanova, J., Holmberg, N., & Holmqvis t, K. (2009). Reading information graphics: The role of spatial contiguity and dual attentional guidance. Applied Cognitive Psychology 23 (9), 1215 1226. Homer, B. D., & Plass, J. L. (2014). Level of interactivity and executive functions as predictors of l earning in computer based chemistry simulations. Computers in Human Behavior 36 365 375. https://doi.org/10.1016/j.chb.2014.03.041 Homer, B. D., Plass, J. L., & Blake, L. (2008). The effects of video on cognitive load and social presence in multimedia le arning. Computers in Human Behavior 24 (3), 786 797. https://doi.org/10.1016/j.chb.2007.02.009 Hong, J., Pi, Z., & Yang, J. (2016). Learning declarative and procedural knowledge via video lectures: cognitive load and learning effectiveness. Innovations in Education and Teaching International 3297 (June), 1 8. https://doi.org/10.1080/14703297.2016.1237371 Jarodzka, H., Scheiter, K., Gerjets, P., & Gog, T. Van. (2010). In the eyes of the Learning a nd Instruction 20 (2), 146 154. https://doi.org/10.1016/j.learninstruc.2009.02.019 Jasper, H. A. (1958). The ten twenty electrode system of the international federation. In Electroencephalography and clinical neurophysiology (pp. 371 375). Jensen, O., & Te sche, C. D. (2002). Frontal theta activity in humans increases with memory load in a working memory task. Eur J Neurosci 15 (8), 1395 1399. https://doi.org/10.1046/j.1460 9568.2002.01975.x Johnson, A. M., Ozogul, G., Moreno, R., & Reisslein, M. (2013). Ped agogical agent signaling of multiple visual engineering representations: The case of the young female agent. Journal of Engineering Education 102 (2), 319 337. https://doi.org/10.1002/jee.20009 Johnson, M. H., Dziurawiec, S., Ellis, H., & Morton, J. (1991) tracking of face like stimuli and its subsequent decline. Cognition 40 (1 2), 1 19. https://doi.org/10.1016/0010 0277(91)90045 6 Jonassen, D. H., & Grabowski, B. L. (2012). Handbook of individual differences, learning, and instruct ion Routledge. Just, M. A., & Carpenter, P. A. (1980). A theory of reading: From eye fixations to comprehension. Psychological Review 87 (4), 329 354. Kalyuga, S., Chandler, P., & Sweller, J. (2011). Managing split attention and redundancy in multimedia i nstruction. Applied Cognitive Psychology 13 (4), 351 371. https://doi.org/10.1002/acp.1773

PAGE 177

177 Kane, M. J., Conway, A. R. A., Bleckley, M. K., & Engle, R. W. (2001). A controlled attention view of working memory capacity. Journal of Experimental Psychology: G eneral 130 (2), 169 183. https://doi.org/10.1037//0096 3445.130.2.169 Karpicke, J. D., & Blunt, J. R. (2011). Retrieval practice produces more learning than elaborative studying with concept mapping. Science 331 (6018), 772 5. https://doi.org/10.1126/scien ce.1199327 Keil, A., Stolarova, M., Heim, S., Gruber, T., & Mller, M. M. (2003). Temporal stability of high frequency brain oscillations in the human EEG. Brain Topography 16 (2), 101 110. Keil, A., Mller, M. M., Gruber, T., Wienbruch, C., & Elbert, T. ( 2001). Human large scale oscillatory brain activity during an operant shaping procedure. Cognitive Brain Research 12 (3), 397 407. https://doi.org/10.1016/S0926 6410(01)00094 5 Kim, Y., & Baylor, A. L. (2016). Research based design of pedagogical agent rol es: A review, progress, and recommendations. International Journal of Artificial Intelligence in Education 26 (1), 160 169. https://doi.org/10.1007/s40593 015 0055 y Kim, Y. H., Kim, D. J., & Wachter, K. (2013). A study of mobile user engagement (MoEN): En gagement motivations, perceived value, satisfaction, and continued engagement intention. Decision Support Systems 56 (1), 361 370. https://doi.org/10.1016/j.dss.2013.07.002 Kizilcec, R. F., Papadopoulos, K., & Sritanyaratana, L. (2014). Showing face in vid eo instruction: effects on information retention, visual attention, and affect. In Proceedings of the 32nd annual ACM conference on Human factors in computing systems (pp. 2095 2102). ACM. Kizilcec, R. F., Bailenson, J. N., & Gomez, C. J. (2015). The instr instruction: evidence from two large scale field studies. Journal of Educational Psychology 107 (3), 724 739. https://doi.org/10.1037/edu0000013 Klimesch, W., Schack, B., & Sauseng, P. (2005). The functional significance of theta and upper alpha oscillations. Experimental Psychology 52 (2), 99 108. https://doi.org/10.1027/1618 3169.52.2.99 Knapp, M. L., Hall, J. A., & Horgan, T. G. (2013). Nonverbal communication in human interaction Cengage Learning. Krmer, N. C., & Bente, G. (2010) Personalizing e learning. The social effects of pedagogical agents. Educational Psychology Review 22 (1), 71 87. https://doi.org/10.1007/s10648 010 9123 x

PAGE 178

178 Kumar, N., & Kumar, J. (2016). Measurement of cognitive load in HCI systems using EEG power spectr um: An experimental study. Procedia Computer Science 84 70 78. https://doi.org/10.1016/j.procs.2016.04.068 Lathrop, A. (2011). Impact of student motivation in online learning activities. Leppink, J., Paas, F., Van der Vleuten, C. P. M., Van Gog, T., & Va n Merrinboer, J. J. G. (2013). Development of an instrument for measuring different types of cognitive load. Behavior Research Methods 45 (4), 1058 1072. https://doi.org/10.3758/s13428 013 0334 1 Louwerse, M. M., Graesser, A. C., McNamara, D. S., & Lu, S. (2008). Embodied conversational agents as conversational partners. Applied Cognitive Psychology 23 (9), 1244 1255. Lusk, D. L., Evans, A. D., Jeffrey, T. R., Palmer, K. R., Wikstrom, C. S., & Doolittle, P. E. (2009). Multimedia learning and individual dif ferences: Mediating the effects of working memory capacity with segmentation. British Journal of Educational Technology 40 (4), 636 651. https://doi.org/10.1111/j.1467 8535.2008.00848.x MacLeod, C. M. (1991). Half a century of research on the Stroop effect : an integrative review. Psychological Bulletin 109 (2), 163. Child Development 48 986 992. Mautone, P. D., & Mayer, R. E. (2001). Signaling as a cognitive guide in mul timedia learning. Journal of Educational Psychology 93 (2), 377 389. https://doi.org/10.1037//0022 0663.93.2.377 Mayer, R. E., & Fiorella, L. (2014). Principles for reducing extraneous processing in multimedia learning: Coherence, signaling, redundancy, sp atial contiguity, and temporal contiguity principles. In R. E. Mayer (Ed.), The Cambridge handbook of multimedia learning (pp. 279 315). Cambridge University Press. Mayer, R. E., & Moreno, R. (2003). Nine ways to reduce cognitive load in multimedia learnin g. Educational Psychologist 38 (1), 43 52. https://doi.org/10.1207/S15326985EP3801 Mayer, R. E., & Pilegard, C. (2014). Principles for managing essential processing in multimedia learning: Segmenting, pre training, and modality principles. In R. E. Mayer ( Ed.), The Cambridge handbook of multimedia learning (pp. 316 344). Cambridge University Press. Mayer, R. E. (1997). Multimedia learning: Are we asking the right questions? Educational Psychologist 32 (1), 1 19. https://doi.org/10.1207/s15326985ep3201

PAGE 179

179 Maye r, R. E. (2009). Multimedia learning Cambridge University Press. Mayer, R. E. (2010). Unique contributions of eye tracking research to the study of learning with graphics. Learning and Instruction 20 (2), 167 171. https://doi.org/10.1016/j.learninstruc.20 09.02.012 Mayer, R. E. (2014). Cognitive theory of multimedia learning. In R. E. Mayer (Ed.), The Cambridge Handbook of Multimedia Learning (pp. 43 71). Cambridge University Press. Mayer, R. E., & Chandler, P. (2001). When learning is just a click away: Do es simple user interaction foster deeper understanding of multimedia messages? Journal of Educational Psychology 93 (2), 390 397. https://doi.org/10.1037/0022 0663.93.2.390 Mayer, R. E., Heiser, J., & Lonn, S. (2001). Cognitive constraints on multimedia le arning: When presenting more material results in less understanding. Journal of Educational Psychology 93 (1), 187 198. https://doi.org/10.1037/0022 0663.93.1.187 Mayer, R. E., Mathias, A., & Wetzell, K. (2002). Fostering understanding of multimedia messag es through pre training: Evidence for a two stage theory of mental model construction. Journal of Experimental Psychology: Applied 8 (3), 147 154. https://doi.org/10.1037/1076 898X.8.3.147 Mayer, R. E., & Moreno, R. (1998). A split attention effect in mult imedia learning: Evidence for dual processing systems in working memory. Journal of Educational Psychology 90 (2), 312 320. https://doi.org/10.1037//0022 0663.90.2.312 Mayer, R. E., & Moreno, R. (2003). Nine ways to reduce cognitive load in multimedia lear ning. Educational Psychologist 38 (1), 43 52. https://doi.org/10.1207/S15326985EP3801_6 Mayer, R. E., Moreno, R., Boire, M., & Vagge, S. (1999). Maximizing constructivist learning from multimedia communications by minimizing cognitive load. Journal of Educ ational Psychology 91 (4), 638 643. https://doi.org/10.1037/0022 0663.91.4.638 Mayer, R. E., & Sims, V. K. (1994). For whom is a picture worth a thousand words? Extensions of a dual coding theory of multimedia learning. Journal of Educational Psychology 8 6 (3), 389 401. https://doi.org/10.1037/0022 0663.86.3.389 Mestre, J. (2002). Transfer of learning: Issues and research agenda. Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psycho logical Review 63 (2), 81 97.

PAGE 180

180 Moreno, R., & Mayer, R. E. (2000). A coherence effect in multimedia learning: The case for minimizing irrelevant sounds in the design of multimedia instructional messages. Journal of Educational Psychology 92 (1), 117 125. Mo reno, R., & Mayer, R. E. (2010). Techniques that increase generative processing in multimedia learning: Open questions for cognitive load research. In Cognitive load theory (pp. 153 177). Moreno, R., & Mayer, R. E. (1999). Cognitive principles of multimedi a learning: The role of modality and contiguity. Journal of Educational Psychology 91 (2), 358 368. https://doi.org/10.1037/0022 0663.91.2.358 Moreno, R., & Mayer, R. E. (2002). Verbal redundancy in multimedia learning: When reading helps listening. Journa l of Educational Psychology 94 (1), 156 163. https://doi.org/10.1037/0022 0663.94.1.156 Moreno, R., Mayer, R. E., Spires, H. A., & Lester, J. C. (2001). The case for social agency in computer based teaching: Do students learn more deeply when they interact with animated pedagogical agents? Cognition and Instruction 19 (2), 177 213. https://doi.org/10.1207/S1532690XCI1902 Ophir, E., Nass, C., & Wagner, A. D. (2009). Cognitive control in media multitaskers. Proceedings of the National Academy of Sciences 106 (37), 15583 15587. https://doi.org/10.1073/pnas.0903620106 Paas, F., Tuovinen, J. E., Tabbers, H., & Van Gerven, P. W. (2003). Cognitive load measurement as a means to advance cognitive load theory. Educational Psychologist 38 (1), 63 71. https://doi.org/1 0.1207/S15326985EP3801 Paas, F. (1992). Training strategies for attaining transfer of problem solving skill in statistics: A cognitive load approach. Journal of Educational Psychology 84 (4), 429 434. https://doi.org/10.1037/0022 0663.84.4.429 Paivio, A. ( 1986). Mental representations: A dual coding approach Oxford, England: Oxford University Press. Paivio, A. (1991). Dual coding theory: Retrospect and current status. Canadian Journal Of Psychology 45 (3), 255 287. https://doi.org/10.1037/h0084295 Pi, Z., & Hong, J. (2016). Learning process and learning outcomes of video podcasts including the instructor and PPT slides: a Chinese case. Innovations in Education and Teaching International 53 (2), 135 144. https://doi.org/10.1080/14703297.2015.1060133 Pi, Z., learning outcomes? Journal of Computer Assisted Learning 33 (4), 347 354. https://doi.org/10.1111/jcal.12183

PAGE 181

181 Plass, J. L., Moreno, R., & Brnken, R. (2010). Cognitive load t heory Cambridge University Press. Poarch, G. J., & van Hell, J. G. (2012). Executive functions and inhibitory control in multilingual children: Evidence from second language learners, bilinguals, and trilinguals. Journal of Experimental Child Psychology 113 (4), 535 551. https://doi.org/10.1016/j.jecp.2012.06.013 Pollock, E., Chandler, P., & Sweller, J. (2002). Assimilation complex information. Learning and Instruction 12 61 86. Prensky, M. (2001). Digital natives, digital immigrants part 1. On the Horiz on 9 (5), 1 6. https://doi.org/10.1001/archderm.142.11.1499 Rayner, K. (1998). Eye movements in reading and information processing: 20 years of research. Psychological Bulletin 124 (3), 372 422. https://doi.org/10.1037/0033 2909.124.3.372 Rothbart, M. K., & Posner, M. I. (1985). Temperament and the development of self regulation. In The neuropsychology of individual differences: A developmental perspective (pp. 99 123). New York: Plenum. ferential investment of mental effort in learning as a function of perceptions and attributions. Journal of Educational Psychology 76 (4), 647 658. https://doi.org/10.1037/0022 0663.76.4.647 Sanchez, C. a., & Wiley, J. (2009). To Scroll or Not to Scroll: S crolling, Working Memory Capacity, and Comprehending Complex Texts. Human Factors 51 (5), 730 738. https://doi.org/10.1177/0018720809352788 Sanchez, C. A., & Wiley, J. (2006). An examination of the seductive details effect in terms of working memory capaci ty. Memory & Cognition 34 (2), 344 355. https://doi.org/10.3758/BF03193412 Sanchez, C. a, & Wiley, J. (2009). To scroll or not to scroll: Scrolling, working memory capacity, and comprehending complex texts. Human Factors 51 (5), 730 738. https://doi.org/10 .1177/0018720809352788 Schmidt Weigand, F., Kohnert, A., & Glowalla, U. (2010). A closer look at split visual attention in system and self paced instruction in multimedia learning. Learning and Instruction 20 (2), 100 110. https://doi.org/10.1016/j.learni nstruc.2009.02.011 Skuballa, I. T., Schwonke, R., & Renkl, A. (2012). Learning from narrated animations with different support procedures: Working memory capacity matters. Applied Cognitive Psychology 26 (6), 840 847. https://doi.org/10.1002/acp.2884

PAGE 182

182 Snow R. E. (1989). Aptitude treatment interaction as a framework for research on individual differences in learning efficiency and effectiveness. Technology, Instruction, Cog nition, and Learning 5 325 337. Stevens, R., Gorman, J. C., Amazeen, P., Likens, A., & Galloway, T. (2013). The organizational neurodynamics of teams. Nonlinear Dynamics, Psychology, and Life Sciences 17 (1), 67 86. Sweller, J., Ayres, P., & Kalyuga, S. (2011). Cognitive load theory. New York: Springer. Sweller, J., Merrienboer, J. J., & Paas, F. (1998). Cognitive architecture and instructional design. Educational Psychology Review 10 (3), 251 296. Sweller, J. (1988). Cognitive load during problem solving : Effects on learning. Cognitive Science 12 (2), 257 285. Sweller, J. (2010). Element interactivity and intrinsic, extraneous, and germane cognitive load. Educational Psychology Review 22 (2), 123 138. https://doi.org/10.1007/s10648 010 9128 5 Tulving, E. (1968). When is recall higher than recognition?. Psychonomic Science 10 (2), 53 54. Turner, M. L., & Engle, R. W. (1989). Is working memory capacity task dependent? Journal of Memory and Language 28 (2), 127 154. https://doi.org/10.1016/0749 59 6X(89)90040 5 Unsworth, N., Heitz, R. P., Schrock, J. C., & Engle, R. W. (2005). An automated version of the operation span task. Behavior Research Methods 37 (3), 498 505. https://doi.org/10.3758/bf03192720 van Albada, S. J., & Robinson, P. A. (2007). Tra nsformation of arbitrary distributions to the normal distribution with application to EEG test retest reliability. Journal of Neuroscience Methods 161 (2), 205 211. https://doi.org/10.1016/j.jneumeth.2006.11.004 van Gog, T., & Scheiter, K. (2010). Eye trac king as a tool to study and enhance multimedia learning. Learning and Instruction 20 (2), 95 99. https://doi.org/10.1016/j.learninstruc.2009.02.009 Verbruggen, F., & Logan, G. D. (2008). Automatic and controlled response inhibition: associative learning in the go/no go and stop signal paradigms. Journal of Experimental Psychology: General 137 (4), 649 672. https://doi.org/10.1037/a0013170

PAGE 183

183 Wang, J., & Antonenko, P. D. (2017). Instructor presence in instructional video: Effects on visual attention, recall, an d perceived learning. Computers in Human Behavior 71 79 89. https://doi.org/10.1016/j.chb.2017.01.049 Yee, N., Bailenson, J. N., & Rickertsen, K. (2007). A meta analysis of the impact of the inclusion and realism of human like faces on user experiences i n interfaces. In Proceedings of the SIGCHI conference on Human factors in computing systems (pp. 1 10). ACM. https://doi.org/10.1145/1240624.1240626 Zelazo, P. D., Anderson, J. E., Richler, J., Wallner Allen, K., Beaumont, J. L., & Weintraub, S. (2013). II NIH Toolbox Cognition Battery (CB): Measuring executive function and attention. Monographs of the Society for Research in Child Development 78 (4), 16 33.

PAGE 184

184 BIOGRAPHICAL SKETCH Jiahui Wang is a doctoral student and research fellow in curriculum and I nstruction with emphasis on educational technology. She completed her b degree in elementary education with an emphasis on m ath ematics e ducation at Ningbo University during 2008 2012 and J iahui participated in an exchange program at the University of Wisconsin Oshkosh during 2010 2011 She graduated from the U niversity of Virginia in 2013 with a Master of Education d egree in curriculum and i nstruction and she taught Chinese at elementary and secondary schools in Fairfax County, VA after graduation. on learning technologies and cognition including multimedia and online learning, individual differences, and educational neuroscience.