<<

Video Modeling and Matrix Training 1

Video Modeling and Matrix Training to Teach Pretend Play in

Children with Autism Spectrum Disorder

A Thesis Presented

by

Lauren M. Dannenberg

In partial fulfillment of the requirements for the degree of

Master of Science

In the field of

Applied Behavior Analysis

Northeastern University

Boston, Massachusetts

August 2010

Video Modeling and Matrix Training 2

NORTHEASTERN UNIVERSITY

Bouvé College of Health Sciences Graduate School

Thesis Title: A replication of Video Modeling and Matrix Training to teach pretend play in children with autism.

Author: Lauren M. Dannenberg

Masters of Science in Applied Behavior Analysis

Committee members:

______

Rebecca MacDonald Date

______

William Ahearn Date

______

Chata Dickson Date

Video Modeling and Matrix Training 3

Video Modeling and Matrix Training to Teach Pretend Play in

Children with Autism Spectrum Disorder

Lauren Dannenberg

Northeastern University

Submitted In partial fulfillment of the requirements for the degree of

Master of Science in Applied Behavior Analysis

in the Bouvé College of Health Sciences Graduate School

of Northeastern University, August 2010

Video Modeling and Matrix Training 4

Acknowledgements

Thanks and appreciation to Rebecca MacDonald, the research supervisor of this Master’s Thesis, for her guidance and support throughout the data collection and writing of this Thesis. Special thanks are also offered to Cara Grieco for her assistance with data collection and to Cormac

MacManus for his assistance throughout the process of this project.

Video Modeling and Matrix Training 5

Abstract

The purpose of this study was to combine video modeling with matrix training to teach play skills in young children with autism. Three children with autism were taught scripted vocalizations and actions for three play sets. Scripted play scenarios were developed using a 3 x

3 x 3 matrix involving characters, vehicles and objects. A within subject multiple probe design across play sets was used to demonstrate experimental control with each participant.

Additionally, a multiple probe design across participants was used to demonstrate experimental control across participants. Baseline data were collected for each participant with each play set.

During training the participant watched a video model consisting of an adult engaging in the scripted scenario. After the participant watched the video twice they had five minutes to play with the toys. Once mastery criteria were achieved, untrained stimuli were presented to the participant to assess the emergence of recombinative play behaviors. Results showed that after training on at least one of the play sets, 1 of the 3 participants demonstrated emergence of script recombination. Recombination was performed by the second and third participant after a brief recombination training session. The use of a matrix was a beneficial way to systematically teach pretend play skills to the three participants of this study.

Video Modeling and Matrix Training 6

Table of Contents

A. Abstract…………………………………………………………………………………....5 B. Introduction……………………………………………………………………………….7 C. Method Participants……………………………………………………………………………....12 Setting………………………………………………………………………………...... 13 Materials……………………………………………………………………………....…13 Independent Variable…………………………………………………………...... 15 Dependent Variable……………………………………………………………….....…..16 Experimental Design………………………………………………………………..…...18 Procedure……………………………………………………………...……………...... 19 Interobserver Agreement………………………...………………………………………20 D. Results………………………………………………………………………….....……..23 E. Discussion……………………………………………………………………...………..46 F. References……………………………………………………….…………………...….48 G. Figure Captions……………………………………………………………………...…..57 H. Figures……………………………………………………………...……………...…….58 I. Appendices Appendix 1: Bank play set and materials……………………………..…………………65 Appendix 2: Mansion play set and materials……………………………...……...……..66 Appendix 3: Castle play set and materials……………………………….…………...…67 Appendix 4: Bank script……………………………..……………………………...…..68 Appendix 5: Mansion script………………………………………………………...…..69 Appendix 6: Castle script………………………………………………………...……..70 Appendix 7: 3D matrices with trained scripts………………………..………...……….71 Appendix 8: 3D matrices with alternative probes…………………………………...….72 Appendix 9: Scoring guidelines for bank script………………………………...………73 Appendix 10: Scoring guidelines for mansion script………………………….………..74 Appendix 11: Scoring guidelines for castle script………………….……...……………75

Video Modeling and Matrix Training 7

Using Video Modeling with Matrix Training to Teach Pretend Play.

Autism is a developmental disability which significantly affects a child’s education (U.S.

Department of Education, 1999). This disability often presents with severe deficits in displaying appropriate affect, which ultimately impedes social development (McGee, Feldman, & Chernin,

1991). Children with autism may engage in repetitive behaviors or resist changes in the environment (Delano, 2007). Another deficit observed within this population is that they may be less likely to accept or make social initiations (Koegal, Koegal, Frea, & Fredeen, 2001). As a result of these deficits, children with autism are less likely to develop play skills at the same rate as their typically developing peers.

In past studies researchers developed ways to teach different skills to children with autism. An effective way to teach children with autism is modeling. Modeling has been shown to be effective because it leads to skills being acquired and generalized quickly (Coleman &

Stedman, 1974). Learning through observation of peers has also been effective in teaching various skills to children with autism (Garfinkle & Schwartz, 2002). Charlop-Christy, Le, and

Freeman (2000) found that video modeling was more effective than in-vivo modeling. Video modeling interventions involve the learner watching a video of a typically developing peer or adult modeling appropriate skills. Ideally after repeated exposures the learner demonstrates imitation of those skills.

The use of a video modeling can be very effective because the same model can be used multiple times. If a video is developed to teach one skill it can be used with different students to teach the same skill. It has been found to be cost effective. Using video modeling can enhance learning by showing edited video which highlights particular behaviors and can be repeatedly viewed (Ayres & Langone, 2005).Videos that are developed to teach different skills may Video Modeling and Matrix Training 8 produce a learning tool that is student specific by incorporating stimuli that are salient to that student.

Reciprocating play actions and vocalizations are two marked deficits observed in children with autism. Taylor, Levin, and Jasper (1999) focused on teaching commenting during play to two children with autism via video modeling. A child with autism viewed a video with scripted play statements made between his sibling and an adult. Treatment sessions involved a child watching a video and then interacting with an adult. The children were taught to participate in an interaction with an adult. The adult also provided reinforcement for correct scripted comments made by the child. Results of this study showed that the children with autism were able to learn three different scripted scenarios. During these scenarios only scripted comments were made by the child. No novel unscripted comments were observed in this study.

Video modeling has been very successful in teaching both scripted play actions and scripted play statements. Children with autism have also been taught to initiate play with peers and adults through video modeling. MacDonald, Sacramone, Mansfield, Wiltz, and Ahearn

(2009) used video modeling to teach reciprocal play to children with autism. The purpose of this study was to teach the children with autism to engage in longer sequences of reciprocal play with their typically developing peers. The participants included two typically developing children and two children with autism. Three different play sets were used to teach the children reciprocal pretend play skills. During the study the researchers took data on scripted vocalizations and actions as well as those that were unscripted. Results of this study showed that video modeling was an effective tool to use to produce extended sequences of reciprocal play. Video Modeling and Matrix Training 9

Video modeling studies have also been shown to be effective for teaching reciprocal play.

Reagon, Higbee, and Endicott (2006) taught pretend play skills to a student with autism using video modeling with his sibling as a model and play partner. A 4-year-old boy diagnosed with autism participated in the study. The boy was provided instruction on play in four play scenarios, each was approximately 30 seconds long. Each episode of play consisted of 5 scripted statements and 6 scripted actions. The participant’s sibling and two typically developing peers were used as models for the scenarios. The video modeling phase of the study resulted in an increase in play skills for the child with autism.

Generative play is an area of focus in many video modeling studies. Generative play can be defined as play in which repertoires that have been taught in particular situations emerge in novel untrained situations (Dauphin, Kinney, & Stromer, 2004). A review of the current video modeling literature conducted by McCoy and Hermansen (2007) found that out of the three studies about play skills, only 1 of the studies showed a development of novel play.

Roberts, MacDonald, and Ahearn (2007) examined the use of video modeling and scripted substitutable loops to teach children to vary their play. Three children between the ages of 5 and 6 years old participated. Participants were shown a video of scripted toy play with a segment that was repeated using multiple characters. These scripted loops contained a series of actions and vocalizations that were modeled with several characters in the play set, additional characters were present but never trained. Following video modeling training, children used characters that were both trained and untrained in their play. The short substitutable loops resulted in greater variation in character use while longer loops resulted in play with trained characters instead of untrained characters. Video Modeling and Matrix Training 10

Another strategy used to develop generative play was included in a study conducted by

Paterson and Arco (2007). Four boys ranging in ages 6-9 were taught to interact with six sets of toys. The toys were broken up into two different groups; related toys and unrelated toys. The researchers found that appropriate play increased as a result of video modeling. Generalized appropriate play was observed in one of the participants when he was presented with toys that were related to each other. The authors used three sets of similar toys. The toys shared common components; the different sets of toys may have been less discriminable which lead to the generalized toy play that was observed.

Using a matrix has been an effective tool to produce untrained combinations. Goldstein and Mousetis (1989) used matrix training and observational learning to teach 6 children with developmental disabilities to combine known words into two- or three-word utterances. The researchers used 3 x 5 x 6 object-preposition-location language matrix, which included three known object words, two known and three unknown prepositions, and three known and three unknown locations. The participants demonstrated generalization with novel combinations or recombitative generalization of an unknown preposition or location with a known location or object. Using a matrix provided a way to organize the material that the participant was being taught and the recombinations that were possible.

Dauphin et al. (2004) examined the use of a video-enhanced activity schedule and matrix training to teach sociodramatic play to a child with autism. During the first phase of the study they used a computer based program was used to teach, various scripted actions and scripted vocalizations. The modeled actions and figurines formed a 3 x 3 matrix. Within the matrix there were three animals that performed three actions which formed nine combinations. The child was taught three of the nine combinations possible. The participant was also taught three 3-item Video Modeling and Matrix Training 11 schedules each contained one social activity, one closed activity, and one open activity.

Following the completion of training, the three schedules were intermixed which created three novel schedules which the participant completed with 96% accuracy or better. In the final phase of the study the researchers replicated the first phase of the study combined with an activity schedule. They did this by using three novel matrices with three exemplars of each toy set. The results showed that this type of teaching procedure was effective in teaching the participant to use an activity schedule and to engage in sociodramatic play. Using video modeling training with matrices resulted in an increase in novel play.

MacManus and MacDonald (2010) used video modeling combined with matrix training to teach three children with autism spectrum disorder to engage in novel play. In this study a

3x3x3 tiered matrix was used to teach play. Each play set was similar, but included different villains, heroes, vehicles, and objects. Each play scenario was made up of three scenes and all of the participants received instruction for three play scripts. Playing according to the scripts was trained to mastery criteria and subsequent alternative material probes were then conducted. In these probes the participants were presented with different materials from the training materials, these probes were conducted based on the matrix model used in the study. MacManus and

MacDonald found that once a participant had been trained on at least one of the play sets they performed both recombined actions and vocalizations with a play set which had not been previously trained. They demonstrated that using matrix training along with video modeling was a more systematic approach to increasing novel play.

The purpose of the current study is to replicate the MacManus and MacDonald study by examining the effects of combining matrix training with video modeling to increase pretend play, including scripted actions and vocalizations as well as recombined actions and vocalizations. Video Modeling and Matrix Training 12

Method

Participants

Three children diagnosed with an autism spectrum disorder were selected for

participation in this study Bonnie, Vincent, and Peter. The participants were enrolled in an

intensive instruction preschool program. Bonnie was five years old at the beginning of the study

(Peabody Picture Vocabulary Test- Fourth Edition Form A (PPVT- 4A ): age equivalent: 6 years

1 month, raw score: 99, 50 th percentile). Bonnie communicated vocally to peers and adults using phrases for the purposes of requesting desired items and activities. Vincent was a 4-year-old male at the beginning of the study (Peabody Picture Vocabulary Test- Fourth Edition Form A

(PPVT- 4A ): age equivalent: 5 years 4 months, raw score: 86, 63 th percentile). He

communicated primarily through 3-7 word phrases and sentences for the purposes of requesting

desired items and activities. Peter was a 4-year-old male at the beginning of the study (Peabody

Picture Vocabulary Test- Fourth Edition Form A (PPVT- 4A): age equivalent 5 years 2 months,

raw score: 92, 82 nd percentile). Peter communicated primarily through 1-4 word phrases and

sentences primarily for the purpose of requesting desired items. All participants were selected

based on their prior experience with video modeling and their scores on the video modeling

assessment conducted at the participants’ school. This assessment tested areas which included;

motor imitation, actions with objects, picture to object matching, computer picture to object,

delayed computer picture to object, and video modeling. Their scores ranged from .75- 1.0.

Video modeling was also included in each participant’s Individualized Education Plan (IEP).

Video Modeling and Matrix Training 13

Setting

All sessions were conducted in a research room (2.7 m x 4.3 m) at the participants’ school. The room was separate from the participants’ classroom and free from distractions. The room had a table and chair where the participant viewed the training videos. The play sets and materials were placed on the floor adjacent to the participant. A DVD with training videos for the scenario they were currently training on, DVD player and video camera which recorded all sessions for later scoring.

Materials

The materials used were toys primarily from the Fisher Price ® toy line for ages 3-8. The three play set locations included a bank play set (46 x 7 x 30 cm) from Hasbro

Marvel Squad range of preschool toys, a mansion play set (26 x 21 x 32 cm) from

Kenner Forever range of action figure toys, and a castle play set (53 x 30 x 42 cm) from

Fisher Price Magic Adventures range of preschool toys. Each play scenario had one of the play set locations (bank, mansion, or castle), 2 characters, 1 object, and 1 vehicle. A play set

(55x 15 x 39 cm), “Batcomputer” play piece (11 x 2.5 x 10.5 cm) and “Bat-a-rang” weapon piece

(1 x 5 cm) were available in each of the three scenarios. The Batcave play set had 2 tiers with a flat roof. It had 2 blue doors on the first tier that could be opened by turning red dials, a purple spiral door that could be opened with a yellow switch from the outside of the play set or a red dial on the inside, and elevator was located on the right side of the Batcave play set that could be raised by turning a red dial next to the top of the elevator. All the characters were (6 x 7 cm) and included hero characters and villain characters. Video Modeling and Matrix Training 14

The bank play set had 2 tiers on the left hand side of the structure with a flat roof, and one tier on the right hand side with a pointed roof. It had 2 doors that were able to be opened and a vault door that could be opened by using a TNT button. The characters for this play set included and Batman. Additional objects included a money piece (5 x 2 cm) and a

Batmobile vehicle (14.5 x 9 x 7 cm) which had a roof that could be opened (see Appendix 1).

The mansion play set had 2 tiers, with 2 windows on the bottom floor, a balcony and a flat roof.

The characters for this play set included Penguin and . Additional objects included a diamond piece (4 x 40 cm, and a vehicle (10.5 x 4 x 6cm) (see Appendix 2). The castle play set had 2 windows on the side wall, a hidden room on the left hand side, and a flat spot where characters could walk on the top of the castle. The characters for this play set included Mr.

Freeze and Superman. Additional objects included a ring piece (2 x 2 cm), and a vehicle (20 x 8 x 15 cm) which had a roof that was able to be opened and propellers that could be turned (see Appendix 3).

Videos

Scripts were recorded on a camcorder and shown on a Portable DVD player with a 13 x

18 cm screen. The training video included one adult holding and speaking for each character.

The scripts included the characters manipulating the play materials, for example Mr. Freeze putting the ring on his arm or Superman sitting in the Batcopter and flying it.

In the Bank scenario the Joker stole money from the bank. There was a call at the

Batcave that Batman answered and then he drove the to the bank to stop the Joker.

This script contained 41 actions and 30 vocalizations and had a running time of 3 minutes and 14 s (see Appendix 4). The characters in this script included: 1 Batman figurine and 1 Joker Video Modeling and Matrix Training 15 figurine. The scenario included the Fisher Price Batcave play set, Bank play set, Batmobile, money bag piece, and Bat-a-rang piece.

In the mansion scenario the Penguin stole the diamond from the mansion. There was a call at the Batcave that Robin answered and then he drove the Batcycle to the mansion to stop the

Penguin. This script contained 30 actions and 29 vocalizations and had a running time of 2 minutes and 52 seconds (see Appendix 5). The characters in this script included: 1 Robin figurine and 1 Penguin figurine. The scenario included the Fisher Price Batcave play set, mansion play set, Batcycle, diamond piece, and a Bat-a-rang piece.

In the castle scenario Mr. Freeze stole the ring from the castle. There was a call at the

Batcave that Superman answered and then he flew the Batcopter to the castle to stop Mr. Freeze.

This scenario contained 36 actions and 29 vocalizations and had a running time of 3 minutes and

21 seconds (see Appendix 6). The characters in this script included: 1 Superman figurine and 1

Mr. Freeze figurine. The scenario included the Fisher Price Batcave play set, Castle play set,

Batcopter, ring piece, and Bat-a-rang piece.

Independent Variable

The presentation of video models that showed the scripted play scenarios for each of the play sets, and the three dimensional matrix the scripts and probes were arranged in were the independent variables. Scripts were arranged in a three dimensional matrix based on Goldstein &

Mousetis (1989). The three 3 x 3 x 3 tiers in the matrix design describe each of the scenes in each scripted scenario (see Appendix 7). The 3D matrices on the left indicate the 3 scenes for the bank scenario and all of the combinations of the materials, the 3D matrices in the middle indicate the 3 scenes for the mansion scenario and all of the combinations of the materials, and the 3D Video Modeling and Matrix Training 16 matrices on the right indicate the 3 scenes for the castle scenario and the possible combinations.

The black shaded areas in the matrices indicate what was trained in each video model. For example, in the first script, the participants use the Joker figurine to steal the money from the bank on the first scene of the bank script. In the first script, second scene the matrix on the left in the middle (see Appendix 7) the participants use figurine to drive the Batmobile to the bank. In the third scene of the first script the participants use the Batman figurine to fight the

Joker figurine at the Bank play set, this is indicated by the black shaded in square (see Appendix

7). A matrix design identical to training was used during the alternative materials probe. During these probes materials other than what was trained with the play set were included (see Appendix

8). For example in the first scene of the alternative bank materials probe, the first scene in the top left corner, the participant is either given the Penguin figurine and the diamond or the Mr.

Freeze figurine and the ring. The participant uses the villain figurine to steal one of the objects.

Dependent Measures

All sessions were videotaped and data were collected on the occurrence of: (a) scripted play actions, (b) scripted vocalizations, (c) unscripted vocalizations, (d) recombined actions, and

(e) recombined vocalizations. Data were collected during each 5-minute session on the number of scripted vocalizations and scripted actions. Data were collected on unscripted vocalizations during the Baseline, Probe, and Alternative Probe sessions. Data were collected on script recombinations during the Alternative Probe sessions. Specific scoring guidelines were created for each script. (see Appendices 9, 10, and 11).

Scripted Actions. Scripted actions were defined as any action that was identical or similar to those modeled in the video, and resulted in the same change as seen in the model. For Video Modeling and Matrix Training 17 example, the participant puts Batman in the Batmobile to drive to the bank. A nonexample was the participant put the Joker in the Batmobile to drive to the bank.

Scripted Vocalizations. Scripted vocalizations were defined as vocal statements that matched the statement of the video model. In addition, statements that were similar to the modeled response but not identical were also scored. For example, when the participant said,

“Crime doesn’t pay Joker!” instead of “Remember Joker, crime doesn’t pay!” Statements that included the incorrect naming of characters, objects, vehicles, or locations were not scored as correct. For example, the participant referred to Batman as “Superman” when Superman was not present.

Unscripted Vocalizations. Unscripted vocalizations were scored as a vocal statement that did not meet the definition of the scripted statements in the video model, but that were contextually appropriate to the scenario or characters (e.g., talking about characters, objects, vehicles and/or locations, or talking for the characters). If the participant repeated the same statement data was only taken first time that the statement was made. Repetitions Unscripted vocalizations were scored by transcribing the unscripted vocalizations that each participant emitted and the time that the participant emitted such vocalizations. Unscripted vocalization data were taken during the Baseline sessions, Mastery Probe sessions, and Alternative Material Probe sessions. This was to assess the changes in unscripted vocalizations as a result of video modeling

(MacDonald et al., 2009).

Recombination Actions: Script recombination actions were defined as actions in which the participant substituted one character, object, or vehicle for another one. Script recombination data were taken during the Alternative Materials Probes. For example a recombined action was Video Modeling and Matrix Training 18 scored when the participant used the Joker to climb to the top of the Castle play-set instead of

Mr. Freeze which was unavailable during the Alternative Material Probe sessions or when the participant used the Batcopter to fly to the bank when the Batmobile was unavailable during the

Alternative Material Probe sessions.

Recombinations Vocalizations : Script recombinations vocalizations were defined as vocalizations in which the participant substituted one character’s scripted vocalization for another character. A recombined vocalization for example, the participant used Batman to say, “I arrived at the castle,” when the Bank play-set was unavailable during the Alternative Material

Probe sessions or when the participant uses the Joker to say, “I am going to steal the world’s biggest diamond from the mansion,” when the money and the Bank play-set were unavailable during the Alternative Material Probe sessions.

Experimental Design

A multiple probe design across play sets and a multiple probe design across participants were used to evaluate the effect of training on the dependent measures. In the multiple probe across participants design, the independent variable was applied to one of the participants while baseline conditions remain in effect for the other participants. Once the first participant performed at mastery criteria for each play scenario, the independent variable was applied to the next participant.

In the multiple probe design intermittent measures are taken at the beginning of the study for each participant and after each play set is mastered for each participant. The play sets were taught in the same order for all participants starting with the bank, the mansion, and then the castle. Participants were required to perform at or above mastery criteria for scripted Video Modeling and Matrix Training 19 vocalizations and scripted actions before training could begin on the next play set. Baseline sessions were conducted prior to the training for each play set and mastery probes were conducted following the acquisition of each script.

Procedure

Play Script Baseline: Prior to all Baseline, Training, and Probe sessions play materials were arranged in front of and to the side of the Batcave play set. During the baseline sessions the participant began the session sitting or standing beside the play materials. The session began when the experimenter gave the instruction, “It’s time to play,” and lasted approximately 5 minutes. All sessions were videotaped by the experimenter and later scored. During play script baseline sessions only the materials necessary for the targeted script were available.

All Materials Baseline: During all materials baselines all play-sets and figurines were present for the participant to engage with. The conditions and instruction given by the experimenter were identical to the play script baseline sessions.

Training: During training sessions the experimenter started the video which was set up in front of the participant on a table playing in the portable DVD player. The play script materials were positioned adjacent to the participant. After viewing the model twice the participant was instructed to play with the specific materials for the play scenario they were currently being trained on the participants were trained on the video scripts indicated by the black shaded areas in Appendix 7. During the first scene of the bank script the Joker stole the money from the bank, in the second scene Batman drove the Batmobile to the bank, and in the third scene of the bank script Batman stopped the Joker at the bank. The participant continued training until they scored

80% or better of the scripted actions and vocalizations for two consecutive sessions. Once Video Modeling and Matrix Training 20

mastery criteria were achieved a probe was conducted. They then began training on the mansion

scenario as indicated by the black shaded area in Appendix 7 of the middle matrices, finally after

that script was mastered the participant began training on the castle scenario as indicated by the

black shading in the matrices in the right of Appendix 7.

Mastery Probe: Mastery probes were identical to play script baseline sessions.

Participants were told, “It’s time to play,” and given 5 minutes to engage with the play materials.

If the participant scored 80% or better of the scripted actions and vocalizations for one session the play scenario was considered mastered. At this point an alternative materials probe was conducted.

Alternative Materials Probes: Alternative materials probes were identical to baseline sessions and mastery probes except the materials specific to the current script were removed and the materials from one of the other scripts were made available. In Appendix 11, the dark grey shaded areas indicate the recombinations that could emerge. After the participant met mastery criteria with the bank scenario they were presented with alternative figurines, play objects, and vehicles to the ones that had been previously trained with the bank scenario. For example during the alternative materials probe after the bank the participant was presented with the Penguin, diamond, Robin, and Batcycle instead of the Joker, money, Batman, and Batmobile.

Interobserver Agreement

Sessions were independently scored by a second observer for purposes of interobserver agreement (IOA). The observer scored scripted vocalizations and actions during baseline, training, and probe sessions. This observer also scored scripted vocalizations and actions as well as recombined vocalizations and actions in the alternative materials probes. Unscripted Video Modeling and Matrix Training 21 vocalizations were scored during baseline, mastery probe, and alternative material probe sessions. Interobserver agreement was calculated for by dividing the number of trials with agreement by the total number of sessions with agreement plus disagreement and multiplying by

100. IOA was scored for at least 33% of sessions for scripted and recombined vocalizations and actions across all participants. IOA was scored for 31% of scored sessions for Participant 1, 38% of scored sessions for Participant 2, and 36% of scored sessions for Participant 3.

Table 1 shows mean interobserver agreement for Participant 1. Participant 1’s mean IOA score for scripted vocalizations in the bank script was 97.5% (range, 88%-100%). The mean IOA for scripted actions for the bank script was 95.3% (range, 88%-100%). For the mansion script the mean IOA for scripted vocalizations was 99% (range, 97%-100%). The mean IOA for scripted actions was 95.3% (range, 93%-100%). For the castle script the mean IOA for scripted vocalizations was 98% (range, 97%-100%). Mean IOA for the scripted actions for the castle script was 96% (range, 91%-100%). For the alternative bank script IOA for recombined vocalizations was 96% and 95.5% for recombined actions. For the alternative mansion script mean IOA for recombined vocalizations was 98.2% (range, 96.3%-100%) and for recombined actions mean IOA was 91.5% (range, 86%-97%). For the alternative castle script IOA for recombined vocalizations was 98% and 80% for recombined actions. For unscripted vocalizations the mean IOA was 93% (range, 80%-100%).

Table 2 shows mean interobserver agreement for Participant 2. Participant 2’s mean IOA score for scripted vocalizations in the bank script was 97.4% (range, 93%-100%). The mean IOA for scripted actions for the bank script was 96.2% (range, 93%-100%). For the mansion script the mean IOA for scripted vocalizations was 99.3% (range, 97%-100%). The mean IOA for scripted actions was 97.4 (range, 87%-100%). For the castle script the mean IOA for scripted Video Modeling and Matrix Training 22 vocalizations was 99.3% (range, 97%-100%). Mean IOA for the scripted actions for the castle script was 96.5% (range, 89%-100%). For the alternative bank script IOA for recombined vocalizations was 100% and 95% for recombined actions. For the alternative mansion script mean IOA for recombined vocalizations was 94% (range, 88%-100%) and for recombined actions mean IOA was 86% (range, 83%-88%). For the alternative castle script IOA for recombined vocalizations was 98% (range, 96%-100%) and 83% (range, 80%-85%) for recombined actions. For unscripted vocalizations the mean IOA was 100%.

Table 3 shows mean interobserver agreement for Participant 3. Participant 3’s mean

IOA score for scripted vocalizations in the bank script was 95% (range, 90%-100%). The mean

IOA for scripted actions for the bank script was 93% (range, 80%-100%). For the mansion script the mean IOA for scripted vocalizations was 96% (range, 83%-100%). The mean IOA for scripted actions was 96% (range, 83%-100%). For the castle script the mean IOA for scripted vocalizations was 99.6% (range, 97%-100%). Mean IOA for the scripted actions for the castle script was 94% (range, 85%-100%). For the alternative bank script IOA for recombined vocalizations was 94% (range, 92%-96%) and 91.5% (range, 91%-92%) for recombined actions.

For the alternative mansion script mean IOA for recombined vocalizations was 98% and for recombined actions mean IOA was 93% (range, 91%-94%). For the alternative castle script IOA for recombined vocalizations was 100% and 100% for recombined actions. For unscripted vocalizations the mean IOA was 100%.

Video Modeling and Matrix Training 23

Results

Bonnie’s Baseline, Training, and Mastery Probes

Results for Bonnie are shown in Figure 1. Bonnie’s two initial baseline scores for the bank were 0% for scripted vocalizations, 0% and 3% for scripted actions. Bonnie completed 0% of the scripted vocalizations and 20% and 7% of the scripted actions of the mansion play script.

During the two initial castle baselines Bonnie completed 0% of the scripted vocalizations and 0% and 17% of the scripted actions.

Bonnie began training on the bank play set and took 6 sessions to meet mastery criteria.

In session 9, Bonnie completed 93% of the scripted vocalizations and 90% of the scripted actions. During the mansion and castle baselines following bank training, Bonnie completed 0% of the scripted vocalizations and 20% of the scripted actions of the mansion script and 17% of the scripted vocalizations and 33% of the scripted actions of the castle script.

Training began on the mansion play set in session 20. It took Bonnie 3 sessions to meet mastery criteria. Bonnie completed 83% of the scripted vocalizations and 100% of the scripted actions. Following mansion script training, Bonnie completed 24% of the scripted vocalizations and 39% of the scripted actions of the castle script.

Training for the castle play set began in session 28 and took Bonnie 3 sessions to meet mastery criteria. Bonnie completed 93% of the scripted vocalizations and 83% of the scripted actions. In bank probe 2 Bonnie completed 90% of scripted vocalizations and 93% of scripted actions. In mansion probe 2 Bonnie completed 76% of the scripted vocalizations and 73% of the scripted actions. Video Modeling and Matrix Training 24

Bonnie’s Alternative Material Probes

In the alternative bank material probes 1 and 2 (mansion and castle materials), sessions

10 and 11. Bonnie completed 13% of the recombined vocalizations and 24% of the recombined actions of bank alternative materials probe 1(mansion materials). Bonnie completed 6% of the recombined vocalizations and 29% of the recombined actions during the bank alternative materials probe 2 (castle materials). During the bank alternative probe 3 (mansion materials), session 36, Bonnie completed 67% of the recombined vocalizations and 92% of the recombined actions. Bonnie completed 67% of the recombined vocalizations and 91% of the scripted actions during the bank alternative probe 4 (castle materials), session 37.

During the alternative mansion material probes1 and 2 (bank materials and castle materials). During the alternative mansion material probe 1(bank materials) session 24, Bonnie completed 31% of the recombined vocalizations and 56% of the recombined actions. Bonnie completed 31% of the recombined vocalizations and 69% of the recombined actions during the mansion alternative materials probe 2(castle materials) session 25. During the mansion alternative probe 3 (bank materials), session 39, Bonnie completed 6% of the recombined vocalizations and 76% of the recombined actions. Bonnie completed 75% of the recombined vocalizations and 83% of the recombined actions of the mansion alternative probe 4 (castle materials), session 40.

During the castle alternative probe l (bank materials), session 32 Bonnie completed 88% of the recombined vocalizations and 82% of the recombined actions. Bonnie completed 79% of the recombined vocalizations and 90% of the recombined actions during the castle alternative probe 2 (mansion materials), session 33. Video Modeling and Matrix Training 25

Bonnie’s Recombined Vocalizations and Actions

Table 4 shows the percentage of actions and vocalizations recombined for each character, object, vehicle, and play set for Bonnie. In alternative bank probes 1 and 3 (mansion materials) sessions10 and 36, Bonnie recombined 55% of the actions completed by the villain after mastering the bank script and 88% after mastering the castle script. She recombined 0% of the actions using a hero character after mastering the bank script and 86% once the castle script was mastered. Recombined vocalizations said by the villain were scored at 0% after the bank script was mastered and 100% after the castle was mastered. Recombined vocalizations said by the hero increased 20% after bank script mastery to 100% after castle script mastery. Recombined vocalizations in which the object, villain, hero, vehicle, or bank were mentioned increased from an average of 24% after bank script mastery to 85% after castle script mastery.

In alternative bank probes 2 and 4 (castle materials) sessions 11 and 37, Bonnie recombined 33% of the actions completed by the villain after mastering the bank script and 67% of mastering the castle script. She recombined 27% of the actions using a hero character after mastering the bank script and 79% once the castle script was mastered. Recombined vocalizations said by the villain were scored at 0% after the bank script was mastered and 100% after the castle was mastered. Recombined vocalizations said by the hero increased 0% after bank script mastery to 60% after castle script mastery. Recombined vocalizations in which the object, villain, hero, vehicle, or bank were mentioned increased from an average of 23% after bank script mastery to 69% after castle script mastery.

In alternative mansion probes 1 and 3 (bank materials) sessions 24 and 39, Bonnie recombined 67% of the actions completed by the villain after mastering the mansion script and Video Modeling and Matrix Training 26

54% of mastering the castle script. She recombined 33% of the actions using a hero character after mastering the mansion script and 67% once the castle script was mastered. Recombined vocalizations said by the villain were scored at 25% after the mansion script was mastered and

25% after the castle was mastered. Recombined vocalizations said by the hero increased 27% after mansion script mastery to 100% after castle script mastery. Recombined vocalizations in which the object, villain, hero, vehicle, or mansion were mentioned decreased from 36% after mansion script mastery to 7% after castle script mastery. During the second alternative mansion probe A the participant ran out of time to complete the session.

In alternative mansion probes 2 and 4 (castle materials) sessions 25 and 40, Bonnie recombined 44% of the actions completed by the villain after mastering the mansion script and

56% of mastering the castle script. She recombined 67% of the actions using a hero character after mastering the mansion script and 87% once the castle script was mastered. Recombined vocalizations said by the villain were scored at 89% after the mansion script was mastered and

100% after the castle was mastered. Recombined vocalizations said by the hero increased 45% after mansion script mastery to 55% after castle script mastery. Recombined vocalizations in which the object, villain, hero, vehicle, or mansion were mentioned increased from an average of

0% after mansion script mastery to 81% after castle script mastery.

Two alternative castle probes were administered after mastery of the castle script. In alternative castle probe 1 (bank materials) session 32, Bonnie recombined 78% of the actions she recombined 85% of the actions using a hero. Recombined vocalizations said by the villain were scored at 100%. Recombined vocalizations said by the hero were scored at 100%. Recombined vocalizations in which the object, villain, hero, vehicle, or castle were mentioned were scored at

87%. Video Modeling and Matrix Training 27

In alternative castle probe 2 (mansion materials) session 33, Bonnie recombined 82% of the actions she recombined 85% of the actions using a hero. Recombined vocalizations said by the villain were scored at 67%. Recombined vocalizations said by the hero were scored at 100%.

Recombined vocalizations in which the object, villain, hero, vehicle, or castle were mentioned were scored at 76%.

Bonnie’s Unscripted Vocalizations

Bonnie’s unscripted vocalizations are shown in Figure 2. Unscripted vocalization data was taken in the initial baseline, baseline before training, mastery probe, and the alternative materials probes for each play set. In the first bank baseline, the session before training, Bonnie had 7 unscripted vocalizations scored. In the bank mastery probe, Bonnie had 0 vocalizations scored. In the bank alternative probe 1 (mansion materials), Bonnie had 9 vocalizations scored.

In the bank alternative probe 2 (castle materials), Bonnie had 8 vocalizations scored. In Bonnie’s second bank probe she had 5 unscripted vocalizations scored. In bank alternative probe 3

(mansion materials), Bonnie had 15 unscripted vocalizations scored and in bank alternative probe 4 (castle materials), Bonnie had 16 unscripted vocalizations scored.

In the first mansion baseline, Bonnie had 8 unscripted vocalizations scored. In the third mansion baseline, the session before training Bonnie had 18 unscripted vocalizations scored. In the mansion mastery probe, Bonnie had 0 unscripted vocalizations scored. In the mansion alternative probe 1 (bank materials); Bonnie had 5 unscripted vocalizations scored. In the mansion alternative probe 2 (castle materials), Bonnie had 0 unscripted vocalizations scored. In

Bonnie’s second mansion probe she had 10 unscripted vocalizations scored. In mansion Video Modeling and Matrix Training 28 alternative probe 3 (bank materials), Bonnie had 18 unscripted vocalizations scored and in mansion alternative probe 4 (castle materials), Bonnie had 13 unscripted vocalizations scored.

In the first castle baseline, Bonnie had 3 unscripted vocalizations scored. In fourth castle baseline, the baseline before training, Bonnie had 11 unscripted vocalizations scored. In the castle mastery probe, Bonnie had 8 unscripted vocalizations scored. In castle alternative probe

1(bank materials), Bonnie had 5 unscripted vocalizations scored and in the castle alternative probe 2 (mansion materials), Bonnie had 9 unscripted vocalizations scored.

Vincent’s Baseline, Training, and Mastery Probes

Results for Vincent are shown in Figure 3. Vincent had 5 baseline sessions on each play set prior to training. Vincent’s average baseline scores for the bank were 0% for scripted vocalizations, 9% for scripted actions. Average of the first five mansion baseline scores were 0% of the scripted vocalizations and 4% of the scripted actions. The average of the first five castle baseline session was 0% of the scripted vocalizations and 17% of the scripted actions.

Vincent began training on the bank play set at session 21 and took 5 sessions to meet mastery criteria. Vincent completed 96% of the scripted vocalizations and 95% of the scripted actions. Following bank training, Vincent completed 0% of the scripted vocalizations and 0% of the scripted actions of the mansion script and 0% of the scripted vocalizations and 27% of the scripted actions of the castle script.

Training began on the mansion play set in session 32. It took Vincent 6 sessions to meet mastery criteria. Vincent completed 100% of the scripted vocalizations and 87% of the scripted actions. Following mansion training, Vincent completed 5% of the scripted vocalizations and

17% of the scripted actions of the castle script. Video Modeling and Matrix Training 29

Training for the castle play set began in session 51 and took Vincent 3 sessions to meet mastery criteria. Vincent completed 97% of the scripted vocalizations and 94% of the scripted actions. In bank probe 2 Vincent completed 87% of scripted vocalizations and 95% of scripted actions. In mansion probe 2 Vincent completed 97% of the scripted vocalizations and 95% of the scripted actions.

Vincent’s Alternative Material Probes

During alternative bank material probes 1 and 2, sessions 27 and 28, Vincent completed

5% of the recombined vocalizations and 17% of the recombined actions of alternative bank probe 1(mansion materials). Vincent completed 0% of the recombined vocalizations and 37% of the recombined actions during the bank alternative materials probe 2 (castle materials). During the bank alternative probe 3 (mansion materials), session 50, Vincent completed 16% of the recombined vocalizations and 37% of the recombined actions. During bank alternative probe

4(castle materials), session 51, Vincent completed 6% of the recombined vocalizations and 16% of the recombined.

During mansion alternative materials probe 1 (bank materials) session 40, Vincent completed 6% of the recombined vocalizations and 44% of the recombined actions. During mansion alternative materials probe 2 (castle materials) session 41, Vincent completed 0% of the recombined vocalizations and 28% of the recombined scripted actions. During the mansion alternative probe 3 (bank materials) session 52 Vincent completed 2% of the recombined vocalizations and 42% of the recombined actions. During mansion alternative probe 4 (castle materials) session 53, Vincent completed 6% of the recombined vocalizations and 35% of the recombined actions. Video Modeling and Matrix Training 30

During the castle alternative probe l (bank materials) session 48 Vincent completed 2% of the recombined vocalizations and 36% of the recombined actions. Vincent completed 6% of the recombined vocalizations and 15% of the recombined actions during the castle alternative probe 2 (mansion materials) session 49.

Vincent’s Alternative Materials Probes after Instructional Session

Prior to session 54 an additional recombined instructional session was run with this participant to demonstrate how to recombine the actions and vocalizations with untrained characters and materials. All the alternative material probes were then conducted. During the bank alternative probe 5 (mansion materials) session 54, Vincent completed 88% of the recombined vocalizations and 92% of the recombined actions. Vincent completed 82% of the recombined vocalizations and 92% of the recombined actions during the bank alternative probe 6

(castle materials) session 55. During the mansion alternative probe 5 (bank materials) session 56

Vincent completed 79% of the recombined vocalizations and 92% of the recombined actions.

Vincent completed 6% of the recombined vocalizations and 35% of the recombined actions of the mansion alternative probe 6 (castle materials) session 57. During the castle alternative probe

3(bank materials) session 58, Vincent completed 80% of the recombined vocalizations and 88% of the recombined actions. Vincent completed 74% of the recombined vocalizations and 90% of the recombined actions during the castle alternative probe 4 (mansion materials) session 59.

Vincent’s Percentage of Recombined Vocalizations and Actions

Table 5 shows the percentage of actions and vocalizations recombined for each character, object, vehicle, and play set for Vincent. In alternative bank probes 1 and 3 (mansion materials) sessions 27 and 50, Vincent recombined 11% of the actions completed by the villain after Video Modeling and Matrix Training 31 mastering the bank script and 0% of mastering the castle script. He recombined 29% of the actions using a hero character after mastering the bank script and 43% once the castle script was mastered. Recombined vocalizations said by the villain were scored at 8% after the bank script was mastered and 20% after the castle was mastered. Recombined vocalizations said by the hero increased 0% after bank script mastery to 28% after castle script mastery. Recombined vocalizations in which the object, villain, hero, vehicle, or bank were mentioned increased from an average of 0% after bank script mastery to 8% after castle script mastery.

In alternative bank probes 2 and 4 (castle materials) sessions 28 and 51, Vincent recombined 0% of the actions completed by the villain after mastering the bank script and 0% of mastering the castle script. He recombined 43% of the actions using a hero character after mastering the bank script and 21% once the castle script was mastered. Recombined vocalizations said by the villain were scored at 0% after the bank script was mastered and 0% after the castle was mastered. Recombined vocalizations said by the hero increased 0% after bank script mastery to 6% after castle script mastery. Recombined vocalizations in which the object, villain, hero, vehicle, or bank were mentioned maintained an average of 0% after bank script mastery to 0% after castle script mastery.

In alternative mansion probes 1 and 3 (bank materials) sessions 40 and 52, Vincent recombined 0% of the actions completed by the villain after mastering the mansion script and

0% of mastering the castle script. He recombined 40% of the actions using a hero character after mastering the mansion script and 33% once the castle script was mastered. Recombined vocalizations said by the villain were scored at 0% after the mansion script was mastered and 0% after the castle was mastered. Recombined vocalizations said by the hero decreased from 13% after mansion script mastery to 0% after castle script mastery. Recombined vocalizations in Video Modeling and Matrix Training 32 which the object, villain, hero, vehicle, or mansion were mentioned decreased from 2% after mansion script mastery to 0% after castle script mastery.

In alternative mansion probes 2 and 4 (castle materials) session 41 and 53, Vincent recombined 0% of the actions completed by the villain after mastering the mansion script and

0% of mastering the castle script. He recombined 33% of the actions using a hero character after mastering the mansion script and 20% once the castle script was mastered. Recombined vocalizations said by the villain were scored at 0% after the mansion script was mastered and 0% after the castle was mastered. Recombined vocalizations said by the hero increased from 0% after mansion script mastery to 45% after castle script mastery. Recombined vocalizations in which the object, villain, hero, vehicle, or mansion were mentioned maintained at an average of

0% after mansion script mastery to 0% after castle script mastery.

In alternative castle probe 1(bank materials) session 48; Vincent recombined 11% of the actions using the villain. He recombined 31% of the actions using a hero. Recombined vocalizations said by the villain were scored at 0%. Recombined vocalizations said by the hero were scored at 0%. Recombined vocalizations in which the object, villain, hero, vehicle, or castle were mentioned were scored at 0%.

In alternative castle probe 2 (mansion materials) session 49, Vincent recombined 22% of the actions. He recombined 15% of the actions using a hero. Recombined vocalizations said by the villain were scored at 10%. Recombined vocalizations said by the hero were scored at 12%.

Recombined vocalizations in which the object, villain, hero, vehicle, or castle were mentioned were scored at 3%.

Video Modeling and Matrix Training 33

Vincent’s Recombined Vocalizations and Actions after Instructional Session

Prior to session 54 an additional recombined instructional session was run with this participant to demonstrate how to recombine the actions and vocalizations with untrained characters and materials. All the alternative material probes were administered. In alternative bank probe 5 (mansion materials) session 54; Vincent recombined an average of 5.5% of the actions completed by the villain prior to the instructional session and 100% after. He recombined an average of 36% of the actions using a hero character prior to the instructional session which increased to 70% after instruction. Recombined vocalizations said by the villain were scored at an average of 14% which increased to 83% after the instructional session.

Recombined vocalizations said by the hero increased averaged 14% prior to instruction and increased to 100%. Recombined vocalizations in which the object, villain, hero, vehicle, or bank were mentioned increased from an average of 4% prior to instructional recombination to 72% after training.

In alternative bank probe 6 (castle materials) session 55; Vincent recombined an average of 0% of the actions completed by the villain which increased to 100% after recombination instruction. He recombined an average of 32% of the actions using a hero character prior to instruction and 71% after. Recombined vocalizations said by the villain were scored at an average of 0% which increased to100% after the participant was taught to recombine.

Recombined vocalizations said by the hero increased from an average of 3% to 100%.

Recombined vocalizations in which the object, villain, hero, vehicle, or bank were mentioned increased from an average of 0% to 75% after recombination instruction. Video Modeling and Matrix Training 34

In alternative mansion probe 5 (bank materials) session 56; Vincent recombined an average of 0% of the actions completed by the villain and 100% after the instructional session.

He recombined an average of 37% of the actions using a hero character after prior to instruction and 87% after. Recombined vocalizations said by the villain was scored averaged 0% and increased to 73% after recombination instruction. Recombined vocalizations said by the hero increased from an average of 7% to 100% after the participant was taught to recombine.

Recombined vocalizations in which the object, villain, hero, vehicle, or mansion were mentioned increased from an average of 1% to an average of 85%.

In alternative mansion probe 6 (castle materials) session 57, Vincent recombined an average of 0% of the villain’s actions prior to the instructional session and increased to 89%. He recombined an average of 27% of the actions using a hero character which increased to 73% after instruction. Recombined vocalizations said by the villain was scored with an average of 0% and increased to 82%. Recombined vocalizations said by the hero averaged 23% and increased to

94% after recombination instruction. Recombined vocalizations in which the object, villain, hero, vehicle, or mansion were mentioned averaged 0% prior to instruction and increased to an average of 78% after training.

In alternative castle probe 3 (bank materials) session 58; prior to recombination instruction, Vincent recombined 11% of the actions using the villain which increased to 89%. He recombined 31% of the actions using a hero prior to instruction and 92% after. Recombined vocalizations said by the villain were scored at 0% prior to recombination instruction, which increased to 90%. Recombined vocalizations said by the hero were scored at 0% before training and 88% after. Prior to recombination instruction recombined vocalizations in which the object, villain, hero, vehicle, or castle were mentioned were scored at 0% this increased to76%. Video Modeling and Matrix Training 35

In alternative castle probe 4 (mansion materials) session 59, Vincent recombined 22% of the actions of the villain prior to recombination instruction which increased to 100%. He recombined 15% of the actions using a hero prior to instruction which increased to 85%. Prior to recombination instruction recombined vocalizations said by the villain was scored at 10% this increased to 80%. Recombined vocalizations said by the hero was scored at 12% before training and increased to 81%. Recombined vocalizations in which the object, villain, hero, vehicle, or castle were mentioned were scored at 3% prior to recombination instruction and increased to an average of 71%.

Vincent’s Unscripted Vocalizations

Vincent’s unscripted vocalizations are shown in Figure 4. Unscripted vocalization data was taken in the initial baseline, baseline before training, mastery probe, and the alternative materials probes for each play set. In the first bank baseline, Vincent had 0 unscripted vocalizations scored. In bank baseline 5, the baseline before training, Vincent had 0 unscripted vocalizations scored. In the bank mastery probe, Vincent had 0 vocalizations scored. In the bank alternative probe 1 (mansion materials), Vincent had 4 vocalizations scored. In the bank alternative probe 2 (castle materials), Vincent had 1 vocalizations scored. In Vincent’s second bank probe he had 0 unscripted vocalizations scored. In bank alternative probe 3 (mansion materials), Vincent had 1 unscripted vocalizations scored and in the bank alternative probe 4

(castle materials), Vincent had 4 unscripted vocalizations scored. In bank alternative probe 5

(mansion materials), Vincent had 0 unscripted vocalizations scored and in the bank alternative probe 6 (castle materials), Vincent had 0 unscripted vocalizations scored. Video Modeling and Matrix Training 36

In the first mansion baseline, Vincent had 0 unscripted vocalizations scored. In the sixth mansion baseline, the session before training, Vincent had 9 unscripted vocalizations scored. In the mansion mastery probe, Vincent had 0 unscripted vocalizations scored. In the mansion alternative probe 1 (bank materials), Vincent had 1 unscripted vocalizations scored. In the mansion alternative probe 2 (castle materials), Vincent had 1 unscripted vocalizations scored. In

Vincent’s second mansion probe he had 0 unscripted vocalizations scored. In mansion alternative probe 3 (bank materials), Vincent had 3 unscripted vocalizations scored and in mansion alternative probe 4 (castle materials), Vincent had 3 unscripted vocalizations scored. In mansion alternative probe 5 (bank materials), Vincent had 0 unscripted vocalizations scored and in mansion alternative probe 6 (castle materials),

Vincent had 0 unscripted vocalizations scored. In the first castle baseline, Vincent had 0 unscripted vocalizations scored. In seventh castle baseline, the baseline before training, Vincent had 6 unscripted vocalizations scored. In the castle mastery probe, Vincent had 0 unscripted vocalizations scored. In castle alternative probe 1(bank materials), Vincent had 2 unscripted vocalizations scored and in the castle alternative probe 2 (mansion materials), Vincent had 7 unscripted vocalizations scored. In castle alternative probe 3(bank materials), Vincent had 0 unscripted vocalizations scored and in the castle alternative probe 4 (mansion materials), Vincent had 0 unscripted vocalizations scored.

Peter’s Baseline, Training, and Mastery Probes

Results for Peter are shown in Figure 5. Peter had 8 baseline sessions before training began. Peter’s average baseline scores for the bank were 0% for scripted vocalizations, and 2 % for scripted actions. Peter completed 0% of the scripted vocalizations and 7% of the scripted Video Modeling and Matrix Training 37 actions for the mansion script. During the castle baselines Peter completed 1% of the scripted vocalizations and 13% of the scripted actions. Peter began training on the bank play set at session 31 and took 11 sessions to meet mastery criteria. Peter completed 93% of the scripted vocalizations and 90% of the scripted actions. During baseline sessions following bank training,

Peter completed 0% of the scripted vocalizations and 13% of the scripted actions of the mansion script and 0% of the scripted vocalizations and 14% of the scripted actions of the castle script.

Training began on the mansion play set in session 48. It took Peter 6 sessions to meet mastery criteria. Peter completed 100% of the scripted vocalizations and 83% of the scripted actions. In bank probe 2 Peter completed 93% of scripted vocalizations and 88% of scripted actions. During a castle baseline following mansion training Peter completed 0% of the scripted vocalizations and 14% of the scripted actions.

Training for the castle play set began in session 74 and took Peter 6 sessions to meet mastery criteria. Peter completed 97% of the scripted vocalizations and 81% of the scripted actions. In bank probe 3 Peter completed 100% of scripted vocalizations and 88% of scripted actions. In mansion probe 2 Peter completed 100% of the scripted vocalizations and 77% of the scripted actions.

Peter’s Alternative Materials Probes

During bank alternative materials probe 1 (mansion materials) session 43, Peter completed 2% of the recombined vocalizations and 13% of the recombined actions. During bank alternative probe 2 (castle materials), session 44, Peter completed 0% of the recombined vocalizations and 13% of the recombined actions. During the bank alternative materials probe 3

(mansion materials) session 58, Peter completed 0% of the recombined vocalizations and 3% of Video Modeling and Matrix Training 38 the recombined actions. Peter completed 0% of the recombined vocalizations and 14% of the recombined actions during the bank alternative materials probe 4 (castle materials) session 59.

During bank alternative probe 5 (bank materials) session 72, Peter completed 6% of the recombined vocalizations and 8% of the recombined actions. Peter completed 2% of the recombined vocalizations and 16% of the recombined actions during the bank alternative probe 6

(castle materials), session 73.

During mansion alternative materials probe 1 (bank materials) session 55, Peter completed 0% of the recombined vocalizations and 26% of the recombined actions. During mansion alternative materials probe 2 (castle materials) session 56, Peter completed 2% of the recombined vocalizations and 9% of the recombined actions. During the mansion alternative probe 3 (bank materials) session 75, Peter completed 0% of the recombined vocalizations and

6% of the recombined actions.

During the castle alternative probe l (bank materials) session 69, Peter completed 4% of the recombined vocalizations and 6% of the recombined actions. Peter completed 2% of the recombined vocalizations and 0% of the recombined actions during the castle alternative probe 2

(mansion materials) session 70.

Peter’s Alternative Materials Probes after Instructional Session

At this point an instruction session was run with this participant to demonstrate how to recombine actions and vocalizations with untrained characters and materials a preferred item was also given to the participant contingent upon completion of the session. Peter completed 65% of the recombined vocalizations and 43% of the recombined actions of the mansion alternative probe 4 (castle materials) session 76. Video Modeling and Matrix Training 39

After recombination training the previous alternative material probes were run.

During the castle alternative probe 3 (bank materials) session77, Peter completed 83% of the recombined vocalizations and 74% of the recombined actions. Peter completed 100% of the recombined vocalizations and 60% of the recombined actions during the castle alternative probe

4 (mansion materials) session 78. During the bank alternative probe 7 (mansion materials) session 79, Peter completed 96% of the recombined vocalizations and 81% of the recombined actions. Peter completed 89% of the recombined vocalizations and 80% of the recombined actions during the bank alternative probe 8 (castle materials) session 80. During the mansion alternative probe 5 (bank materials) session 81, Peter completed 98% of the recombined vocalizations and 81% of the recombined actions. Peter completed 94% of the scripted vocalizations and 82% of the scripted actions of the mansion alternative probe 6 (castle materials) session 82.

Peter’s Recombined Vocalizations and Actions

Table 6 shows the percentage of actions and vocalizations recombined for each character, object, vehicle, and play set for Peter. In alternative bank probes 1, 3, and 5 (mansion materials) sessions 43, 58, and 72, recombined 0% of the actions completed by the villain after mastering the bank script and 0% and recombined 11 % after mastery of both the mansion and the castle script. He recombined 11% of the actions using a hero character after mastering the bank script

11% after the mansion script and decreased to 0% once the castle script was mastered.

Recombined vocalizations said by the villain was scored at 0% after the bank script and remained at 0% after both the mansion and castle scripts were mastered. Recombined vocalizations remained at 0% after mastery of the bank script, mansion script, and castle script.

Recombined vocalizations in which the object, villain, hero, vehicle, or bank were mentioned Video Modeling and Matrix Training 40 decreased from an average of 3% after bank script mastery to 0% after both mansion script mastery and castle script mastery.

In alternative bank probes 2, 4, and 6 (castle materials) sessions 44, 59, and 73,Peter recombined 11% of the actions completed by the villain after mastering the bank script his percent of recombination’s remained at 11% after both the mansion and castle script were mastered. He recombined 7% of the actions using a hero character after mastering the bank script

8% once the mansion script was mastered and increased to 14% once the castle script was mastered. Recombined vocalizations said by the villain was scored at 0% after the bank script, and remained at 0% after mansion script, and castle script mastery. Recombined vocalizations said by the hero remained at 0% after the bank, mansion, and castle script were mastered.

Recombined vocalizations in which the object, villain, hero, vehicle, or bank were mentioned remained at 0% after bank mastery, mansion mastery, and castle mastery.

In alternative mansion probes 1 and 3 (bank materials) sessions 55 and 75, Peter recombined 0% of the actions completed by the villain after mastering the mansion script and

0% of mastering the castle script. He recombined 7% of the actions using a hero character after mastering the mansion script and 0% once the castle script was mastered. Recombined vocalizations said by the villain were scored at 0% after the mansion script was mastered and 0% after the castle was mastered. Recombined vocalizations said by the hero remained at 0% after mansion script mastery and 0% after castle script mastery. Recombined vocalizations in which the object, villain, hero, vehicle, or mansion were mentioned remained at 0% after mansion script mastery and 0% after castle script mastery. Video Modeling and Matrix Training 41

In alternative castle probe 1 (bank materials present) session 69, Peter recombined 0% of the actions using the villain. He recombined 0% of the actions using a hero. Recombined vocalizations said by the villain were scored at 0%. Recombined vocalizations said by the hero were scored at 6%. Recombined vocalizations in which the object, villain, hero, vehicle, or castle were mentioned averaged a score of 0%.

In alternative castle probe 2 (mansion materials) session 70, Peter recombined 0% of the actions. He recombined 0% of the actions using a hero. Recombined vocalizations said by the villain were scored at 0%. Recombined vocalizations said by the hero were scored at 6%.

Recombined vocalizations in which the object, villain, hero, vehicle, or castle were mentioned were scored at 0%.

Peter’s Recombined Vocalizations and Actions after Instructional Session

Prior to session 76, an additional recombination instructional session was administered a preferred item was also offered to the participant contingent on completion of the session. In alternative mansion probe 4 (castle materials) session 76, Peter recombined 0% of the actions completed by the villain after mastering the mansion script and 22% after the recombination instructional session. He recombined 7% of the actions using a hero character after mastering the mansion script and 20% once the instructional session was administered. Recombined vocalizations said by the villain was scored at 0% after the mansion and increased to 27% after the instructional session. Recombined vocalizations said by the hero increased from 0% after mansion script mastery to 94% after recombination instruction. Recombined vocalizations in which the object, villain, hero, vehicle, or mansion were mentioned increased from an average of

0% after mansion script mastery to 76% after recombination training. Video Modeling and Matrix Training 42

In alternative bank probe 7 (mansion materials) session 79; Peter recombined an average of 7% of the actions completed by the villain prior to the instructional session and 89% after. He recombined an average of 0% of the actions using a hero character prior to the instructional session which increased to 64% after instruction. Recombined vocalizations said by the villain were scored at an average of 0% which increased to 100% after the instructional session.

Recombined vocalizations said by the hero increased from an averaged 2% prior to instruction to100% after. Recombined vocalizations in which the object, villain, hero, vehicle, or bank were mentioned increased from an average of 0% prior to instructional recombination to 96% after training.

In alternative bank probe 8 (castle materials) session 80; Peter recombined an average of

7% of the actions completed by the villain which increased to an average of 78% after recombination instruction. He recombined an average of 10% of the actions using a hero character prior to instruction and 57% after. Recombined vocalizations said by the villain were scored at an average of 0% which increased to 92% after the participant was taught to recombine. Recombined vocalizations said by the hero increased from an average of 0% to

100%. Recombined vocalizations in which the object, villain, hero, vehicle, or bank were mentioned increased from an average of 0% to 96% after recombination instruction.

Two alternative mansion probes were administered after recombination instruction. In alternative mansion probe 5 (bank materials) session 81; Peter recombined an average of 0% of the actions completed by the villain and 78% after the instructional session. He recombined an average of 3.5% of the actions using a hero character after prior to instruction and 60% after.

Recombined vocalizations said by the villain was scored averaged 0% and increased to 92% after recombination instruction. Recombined vocalizations said by the hero increased from an average Video Modeling and Matrix Training 43 of 0% to 83% after the participant was taught to recombine. Recombined vocalizations in which the object, villain, hero, vehicle, or mansion were mentioned increased from an average of 0% to an average of 97%.

In alternative mansion probe 6 (castle materials present) session 82, Peter recombined an average of 0% of the villain’s actions prior to the instructional session and increased to an average 44.5%. He recombined an average of 7% of the actions using a hero character which increased to an average of 43.5% after instruction. Recombined vocalizations said by the villain was scored with an average of 0% and increased to an average of 59%. Recombined vocalizations said by the hero averaged 0% and increased to 97% after recombination instruction. Recombined vocalizations in which the object, villain, hero, vehicle, or mansion were mentioned averaged 0% prior to instruction and increased to an average of 88% after training.

In alternative castle probe 3 (bank materials) session 77, prior to recombination instruction, Peter recombined 0% of the actions using the villain which increased to 56%. He recombined 0% of the actions using a hero prior to instruction and 62% after. Recombined vocalizations said by the villain were scored at 0% prior to recombination instruction, which increased to 50%. Recombined vocalizations said by the hero were scored at 6% before training and 100% after. Prior to recombination instruction recombined vocalizations in which the object, villain, hero, vehicle, or castle were mentioned were scored at 0% this increased to 85%.

In alternative castle probe 4 (mansion materials) session 78, Peter recombined 0% of the actions of the villain prior to recombination instruction which increased to 67%. He recombined

0% of the actions using a hero prior to instruction which increased to 69%. Prior to Video Modeling and Matrix Training 44 recombination instruction recombined vocalizations said by the villain was scored at 0% this increased to 100%. Recombined vocalizations said by the hero was scored at 6% before the instructional session and increased to 100%. Recombined vocalizations in which the object, villain, hero, vehicle, or castle were mentioned were scored at 0% prior to recombination instruction and increased to an average of 100%.

Peter’s Unscripted Vocalizations

Results for Peter’s unscripted vocalizations are shown in Figure 6. Unscripted vocalization data was taken in the initial baseline, baseline before training, mastery probe, and the alternative materials probes for each play set. In the first bank baseline, Peter had 1 unscripted vocalization scored. In bank baseline 8, the baseline before training, Peter had 1 unscripted vocalization scored. In the bank mastery probe, Peter had 1 vocalization scored. In the bank alternative probe 1 (mansion materials), Peter had 0 vocalizations scored. In the bank alternative probe 2 (castle materials), Peter had 1 vocalization scored. In Peter’s second bank probe he had 1 unscripted vocalization scored. In bank alternative probe 3(mansion materials),

Peter had 1 unscripted vocalization scored and in the bank alternative probe 4 (castle materials),

Peter had 1 unscripted vocalization scored. In Peter’s third bank probe he had 1 unscripted vocalization scored. In bank alternative probe 5 (mansion materials), Peter had 2 unscripted vocalizations scored and in the bank alternative probe 6 (castle materials), Peter had 0 unscripted vocalizations scored. In bank alternative probe 7 (mansion materials), Peter had 0 unscripted vocalizations scored and in the bank alternative probe 8 (castle materials), Peter had 1 unscripted vocalization scored. Video Modeling and Matrix Training 45

In Peter’s first mansion baseline, 0 unscripted vocalizations were scored. In the first mansion baseline, Peter had 2 unscripted vocalizations scored. In the ninth mansion baseline, the session before training, Peter had 1 unscripted vocalization scored. In the mansion mastery probe, Peter had 0 unscripted vocalizations scored. In the mansion alternative probe 1 (bank materials); Peter had 1 unscripted vocalization scored. In the mansion alternative probe 2 (castle materials), Peter had 0 unscripted vocalizations scored. In Peter’s second mansion probe he had 1 unscripted vocalization scored. In mansion alternative probe 3 (bank materials), Peter had 0 unscripted vocalizations scored and in mansion alternative probe 4 (castle materials), Peter had 1 unscripted vocalization scored. In mansion alternative probe 5 (bank materials), Peter had 1 unscripted vocalization scored and in mansion alternative probe 6 (castle materials), Peter had 2 unscripted vocalizations scored.

In the first castle baseline, Peter had 1 unscripted vocalization scored. In tenth castle baseline, the baseline before training, Peter had 0 unscripted vocalizations scored. In the castle mastery probe, Peter had 0 unscripted vocalizations scored. In castle alternative probe 1(bank materials), Peter had 1 unscripted vocalization scored and in the castle alternative probe 2

(mansion materials), Peter had 0 unscripted vocalizations scored. In castle alternative probe 3

(bank materials), Peter had 0 unscripted vocalizations scored and in the castle alternative probe 4

(mansion materials), Peter had 1 unscripted vocalization scored.

Total Script Completion for All Participants

Figure 7 shows the total script completion for all 3 participants. Bonnie finished the study in session 40. She was in baseline until session 8. She had 6 bank training sessions before a Video Modeling and Matrix Training 46 mastery probe in session 14. Mansion training started in session 20 she mastered this script in 3 sessions, before a mastery probe in session 23. Castle training began in session 28.

Vincent completed the study in session 81. He was in baseline until session 39. He had 5 bank training sessions before a mastery probe in session 45. Mansion training started in session

53, he had 6 training session before a mastery probe in session 58. He had 3 castle training sessions starting in session 63. A mastery probe was run in session 66.

Peter completed the study in session 134. He was in baseline until session 83. He had 11 bank training sessions before a mastery probe in session 94. Mansion training started in session

100 and lasted until session 105. Castle training began in session 114 and finished in session 118.

Discussion

There have been several approaches to teaching children to engage in generative play.

The substitutable loops used in Roberts (2007), were successful in producing novel play behavior. Using toys that are share similar components may also produce generative play. The use of a matrix attempts to demonstrate the various recombinations are possible. The several studies that have employed matrix training to teach various skills have been successful producing novel or generative recombinations. Using matrices to teach participants to engage in novel play have proven to be successful.

In the current study, all three participants acquired the scripts at a rapid rate. Each participant mastered the scripts at a more rapid rate after being trained on at least one script. This may have been due to the similarities in each script. The training sessions only included the play sets and materials that were included in the specific training video. During training sessions only scripted actions and scripted vocalizations were scored. If all materials were present during training there may be a slightly different outcome. Video Modeling and Matrix Training 47

It would be interesting to examine the effects of combining substitutable loops in video modeling used in the Roberts (2007) study with the matrix. By adding these two techniques together it may be more likely to produce results in which the participant combines several of the scripts together. If all the materials were available in a probe after the participant learned the scripts with a substitutable loop the participant may perform more recombinations.

The results obtained from this study are similar to those found by MacManus and

MacDonald (2010). After one of the play sets was mastered the participants performed some of the recombinations with the other untrained characters, objects, and play sets. Two of the three participants did not recombine the vocalizations and actions until additional instructional sessions were administered. This may have been due to the length and difficulty of the scripts.

Although recombination did not occur during the first alternative probe sessions, both participants, performed the majority of the recombination’s immediately after the recombination instructional session without any further prompting. One of the participants was noncompliant during some of the baseline and probe sessions this could have been due to his familiarity with the experimenter. During all sessions participants were given approximately 5 minutes to manipulate the play materials. Given this short duration the participants only had a short time to perform the scripted play scenario. With longer sessions there is a possibility of more recombinations or more unscripted play.

There are many options for future research. Matrix training offers a systematic approach to developing probes to anticipate the recombinations that are possible with this type of procedure. Future research should aim to use matrix training when teaching various skills to children in this population. An important area of research may be teaching social skills to children with an autism spectrum disorder. Video Modeling and Matrix Training 48

References Ayres, K. M., & Langone, J., (2005). Intervention and instruction with video for students with autism: a review of the literature. Education and Training in Developmental Disabilities, 40, 183-196. Calabrese, N. (2003). Developing quality sociodramatic play for young children. Education,123 , 3. Charlop M., & Milstein, J. (1989). Teaching autistic children conversational speech using video modeling. Journal of Applied Behavior Analysis, 22, 275-285. Charlop-Christy, M. H., Le, L., & Freeman, K. A. (2000). A comparison of video modeling with in vivo modeling for teaching children with autism. Journal of Autism and Developmental Disorders, 30, 537-555. Coleman, S., & Stedman, J.M. (1974). Use of a peer model in language training in an echoic child. Journal of Behavior Therapy and Experimental Psychiatry , 5, 275 -279. Dauphin, M., Kinney, E.M., & Stromer, R. (2004). Using video-enhanced activity schedules and matrix training to teach sociodramatic play to a child with autism. Journal of Positive Behavior Interventions, 6, 238-251. Delano, M., E. (2007). Video modeling interventions for individuals with autism. Remedial and Special Education, 28, 33-42. Garfinkle, A. N., & Schwartz, I. S. (2002). Peer imitation: increasing social interactions in children with autism and other developmental disabilities in inclusive preschool classrooms. Topics in Early Childhood Special Education, 22, 26-38. Goldstein, H., & Mousetis, L., (1989). Generalized language learning by children with severe mental retardation: effects of peers’ expressive modeling. Journal of Applied Behavior Analysis, 22, 245-259. Koegel, L. K., Koegal, R. L., Frea, W. D., & Fredeen, R. M., (2001). Identifying early intervention target for children with autism in inclusive school settings. Behavior Modification, 25, 754-761. Video Modeling and Matrix Training 49

MacDonald, R.P.F., Sacramone, S.R., Mansfield, R., Wiltz, K., & Ahearn, W.H. (2009).

Using video modeling to teach reciprocal pretend play to children with autism. Journal

of Applied Behavior Analysis. 42, 43-55.

MacManus, C., & MacDonald, R.P.F. (May 2010). Video modeling and matrix training to teach pretend play in children with autism. Paper to be presented at the annual conference of the Association for Behavior Analysis, San Antonio, TX. McCoy, K., & Hermansen, E. (2007) Video modeling for individuals with autism: a review of model types and effects. Education & Treatment of Children, 30, 183-214. McGee, G.C., Feldman, R.S., & Chernin, L., (1991). A comparison of emotional facial display by children with autism and typical preschoolers. Journal of Applied Behavior Analysis, 18. 3-16. Nikopoulos, C. K., & Keenan, M. (2004). Effects of video modeling on social initiations by children with autism. Journal of Applied Behavior Analysis, 37, 93-96. Paterson, C. R., & Arco, L., (2007). Using Video modeling for generalizing toy play in children with autism, Behavior Modification 31 (5) 660-681. Reagon, K. A., Higbee, S., & Endicott, K. (2006). Teaching pretend play skills to a student with autism using video modeling with a sibling as model and play partner. Education and Treatment of Children, 29, 1-12. Roberts, S., MacDonald, R. P. F., & Ahearn, W. H. (2007). A method to teach varied play to

children with ASD using video modeling. Autism and Related Developmental

Disabilities: Special Interest Group Newsletter . 23, 1-2.

Taylor, B., Levin, L., & Jasper, S. (1999). Increasing play-related statements in children with autism toward their siblings: effects of video modeling, Journal of Developmental and Physical Disabilities, 11, 253-264. U.S. Department of Education. (1999). Individuals with Disabilities Education Act: Part 300. Federal Register, 64 (8), 12418-12480. Video Modeling and Matrix Training 50

Table 1 Mean interobserver Agreement for Bonnie across all play sets.

Bank Mansion Castle IOA for 97.5% 99% 98% Vocalizations (range, 88%-100%) (range, 97%-100%) (range, 97%-100%)

IOA for Actions 95.3% 95.3% 96% (range, 88%-100%) (range, 93%-100%) (range, 91%-100%) IOA for Recombined 98.2% Vocalizations 96% (range, 96.3%-100%) 98% (During Alterative Materials Probes)

IOA for 91.5% Recombined Actions 95.5% (range, 86%-97%) 80% (During Alterative Materials Probes)

IOA for Unscripted Vocalizations 93% (During Baseline, Probes, and (range, 80%-100%) Alternative Material Probes)

Video Modeling and Matrix Training 51

Table 2 Mean interobserver agreement for Vincent across all play sets.

Bank Mansion Castle IOA for 97.4% 99.3% 99.3% Vocalizations (range, 93%-100%) (range, 97%-100%) (range, 97%-100%)

IOA for Actions 96.2% 97.4% 96.5% (range, 93%-100%) (range, 87%-100%) (range, 89%-100%) IOA for Recombined 94% 98% Vocalizations 100% (range, 88%-100%) (range, 96%-100%) (During Alterative Materials Probes)

IOA for 86% 83% Recombined Actions 95% (range, 83%-88%) (range, 80%-85%) (During Alterative Materials Probes)

IOA for Unscripted Vocalizations 100% (During Baseline, Probes, and Alternative Material Probes)

Video Modeling and Matrix Training 52

Table 3 Mean interobserver agreement for Peter across all play sets.

Bank Mansion Castle IOA for 95% 96% 99.6% Vocalizations (range, 90%-100%) (range, 83%-100%) (range, 97%-100%)

IOA for Actions 93% 96% 94% (range, 80%-100%) (range, 83%-100%) (range, 85%-100%) IOA for Recombined 94% Vocalizations (range, 92%-96%) 98% 100% (During Alterative Materials Probes)

IOA for 91.5% 93% Recombined Actions (range, 91%-92%) (range, 91%-94%) 100% (During Alterative Materials Probes)

IOA for Unscripted 100% Vocalizations (During Baseline, Probes, and Alternative Material Probes)

Video Modeling and Matrix Training 53

Table 4 The percentage of actions and vocalizations recombined for each character, object, vehicle, and play set for Bonnie.

After Bank Training After Mansion Training After Castle Training

Alt. Bank Probe Alt. Bank Probe Alt. Bank Probe Alt. Bank Probe 1(Mansion 2 (Castle 3 (Mansion 4 (Castle materials) materials) materials) materials) 6 % of recombs. % of recombs. % of recombs. % of recombs. Actions using Villain 55 33 88 67 Actions using Hero 0 27 86 79 Actions using Object 0 33 100 75 Actions using Vehicle 0 75 100 100 Vocals said by Villain 20 0 60 60 Vocals said by Hero 6 11 94 100 Vocals in which Object is mentioned 20 0 40 60 Vocals in which Villain is mentioned 0 0 100 33 Vocals in which Hero is mentioned 0 0 100 100 Vocals in which Vehicle is mentioned 10 0 100 100 100 Vocals in which Bank is mentioned 0 17 83 50

Alt. Mansion Alt. Mansion Alt. Mansion Alt. Mansion Probe 1 (Bank Probe 2(Castle Probe 3 (Bank Probe 4(Castle materials) materials) materials) materials) % of recombs. % of recombs. % of recombs. % of recombs. Actions using Villain 67 44 44 56 Actions using Hero 33 67 67 87 Actions using Object 25 0 25 75 Actions using Vehicle 83 89 100 100 Vocals said by Villain 27 45 9 55 Vocals said by Hero 50 75 0 88 Vocals in which Object is mentioned 50 0 25 67 Vocals in which Villain is mentioned 17 0 0 83 Vocals in which Hero is mentioned 0 0 0 100 Vocals in which Vehicle is mentioned 100 0 0 100 Vocals in which Mansion is mentioned 0 0 14 57

Alt. Castle Probe Alt. Castle 1 (Bank Probe 2 materials) (Mansion materials) % of recombs. % of recombs. Actions using Villain 78 82 Actions using Hero 85 85 Actions using Object 100 67 Actions using Vehicle 100 100 Vocals said by Villain 82 80 Vocals said by Hero 100 100 Vocals in which Object is mentioned 100 0 Vocals in which Villain is mentioned 100 100 Vocals in which Hero is mentioned 100 100 Vocals in which Vehicle is mentioned 100 100 Vocals in which Castle is mentioned 33 50

Video Modeling and Matrix Training 54

Table 5 The percentage of actions and vocalizations recombined for each character, object,

vehicle, and play set for Vincent.

After Bank After Mansion After Castle RecAfter Recomb. Training Training Training Alt. B ank Alt. Bank Alt. B ank Alt. Bank Alt. B ank Alt. Bank Probe 1 Probe Probe 3 Probe 4 Probe 5 Probe 6 (Mansion 2(Castle (Mansion (Castle (Mansion (Castle materials) materials) materials) materials) materials) materials) % of % of % of % of % of % of recombs. recombs. recombs. recombs. recombs. recombs. Actions using Villain 11 43 0 0 100 100 Actions using Hero 29 0 43 21 70 71 Actions using Object 0 83 0 0 100 100 Actions using Vehicle 8 0 100 67 100 100 Vocals said by Villain 0 0 20 0 83 67 Vocals said by Hero 0 0 28 6 100 100 Vocals in which Object is mentioned 0 0 0 0 80 80 Vocals in which Villain is mentioned 0 0 0 0 100 83 Vocals in which Hero is mentioned 0 0 0 0 0 50 Vocals in which Vehicle is 0 0 0 0 100 100 mentioned Vocals in which Bank is mentioned 0 0 40 0 80 60

Alt. Alt. Alt. Alt. Alt. Alt. Mansion Mansion Mansion Mansion Mansion Mansion Probe 1 Probe 2 Probe Probe 4 Probe5(B Probe 6 (Bank (Castle 3Bank (Castle ank (Castle materials) materials) materials) materials) materials) materials) % of % of % of % of % of % of recombs. recombs. recombs. recombs. recombs. recombs. Actions using Villain 0 0 0 0 100 89 Actions using Hero 40 33 30 20 87 73 Actions using Object 0 0 0 0 100 75 Actions using Vehicle 83 44 100 67 100 100 Vocals said by Villain 0 0 0 0 73 82 Vocals said by Hero 13 0 0 13 100 94 Vocals in which Object is mentioned 0 0 0 0 83 50 Vocals in which Villain is mentioned 0 0 0 0 83 83 Vocals in which Hero is mentioned 0 0 0 0 100 100 Vocals in which Vehicle is 0 0 0 0 100 100 mentioned Vocals in which Mansion is 0 0 0 0 57 57 mentioned

Alt. Alt. Alt. Alt. Castle Castle Castle Castle Probe Probe 2 Probe 3 Probe 4 1(Bank (Mansion (Bank (Mansion materials) materials) materials) materials) % of % of % of % of recombs. recombs. recombs. recombs. Actions using Villain 11 22 89 100 Actions using Hero 31 15 92 85 Actions using Object 0 5 75 100 Actions using Vehicle 83 33 100 100 Vocals said by Villain 0 10 90 80 Vocals said by Hero 0 12 88 81 Vocals in which Object is mentioned 0 0 60 60 Vocals in which Villain is mentioned 0 0 50 80 Vocals in which Hero is mentioned 0 0 100 50 Vocals in which Vehicle is 0 0 100 100 mentioned Vocals in which Castle is mentioned 0 17 71 67

Video Modeling and Matrix Training 55

Table 6 The percentage of actions and vocalizations recombined for each character, object,

vehicle, and play set for Peter.

After Bank After Mansion After Castle Training Training Training Recomb. Alt. B ank Alt. Bank Alt. B ank Alt. Bank Alt. B ank Alt. Bank Alt. B ank Alt. Bank Probe 1 Probe2 Probe 3 Probe 4 Probe 5 Probe 6 Probe 7 Probe 8 (Mansion (Castle (Mansion (Castle (Mansion (Castle (Mansion (Castle materials) materials) materials) materials) materials) materials) materials) materials) % of % of % of % of % of % of % of % of recombs. recombs. recombs. recombs. recombs. recombs. recombs. recombs. Actions using Villain 0 7 11 0 11 0 89 78 Actions using Hero 0 0 0 8 0 14 64 57 Actions using Object 0 44 0 0 0 0 33 75 Actions using Vehicle 0 0 0 56 0 67 100 100 Vocals said by Villain 0 0 0 0 0 0 100 92 Vocals said by Hero 6 0 0 0 0 0 80 100 Vocals in which Object is mentioned 0 0 0 0 0 0 100 80 Vocals in which Villain is mentioned 17 0 0 0 0 0 100 100 Vocals in which Hero is mentioned 0 0 0 0 0 0 100 100 Vocals in which Vehicle is 0 0 0 0 0 0 100 100 mentioned Vocals in which Bank is mentioned 0 0 0 0 0 0 100 100

Alt. Alt. Alt. Alt. Alt. Alt. Mansion Mansion Mansion Mansion Mansion Mansion Probe Probe 2 Probe Probe 4 Probe Probe 6 1(Bank (Castle 3(Bank (Castle 5(Bank (Castle materials) materials) materials) materials) materials) materials) % of % of % of % of % of % of recombs. recombs. recombs. recombs. recombs. recombs. Actions using Villain 0 0 0 22 78 67 Actions using Hero 7 7 0 20 60 67 Actions using Object 0 0 0 25 75 75 Actions using Vehicle 0 44 0 78 100 100 Vocals said by Villain 0 0 0 27 91 91 Vocals said by Hero 0 0 0 94 83 100 Vocals in which Object is mentioned 0 0 0 33 83 100 Vocals in which Villain is mentioned 0 0 0 100 83 100 Vocals in which Hero is mentioned 0 0 0 100 100 100 Vocals in which Vehicle is 0 0 0 100 100 100 mentioned Vocals in which Mansion is 0 0 0 48 100 100 mentioned

Alt. Alt. Alt. Alt. Castle Castle Castle Castle Probe 1 Probe 2 Probe 3 Probe 4 (Bank (Mansion (Bank (Mansion materials) materials) materials) materials) % of % of % of % of recombs. recombs. recombs. recombs. Actions using Villain 0 0 56 67 Actions using Hero 0 0 62 69 Actions using Object 0 0 75 50 Actions using Vehicle 17 0 100 100 Vocals said by Villain 0 0 50 100 Vocals said by Hero 6 6 100 100 Vocals in which Object is mentioned 0 0 60 100 Vocals in which Villain is mentioned 0 0 100 100 Vocals in which Hero is mentioned 0 0 100 100 Vocals in which Vehicle is 0 0 100 100 mentioned Vocals in which Castle is mentioned 0 0 67 100

Video Modeling and Matrix Training 56

Figure Captions

Figure 1 Bonnie’s scripted vocals and actions and recombined vocals and actions for each play

set.

Figure 2 Bonnie’s unscripted vocals for baseline, mastery probes, and alternative material probes

for each play set.

Figure 3 Vincent’s scripted vocals and actions and recombined vocals and actions for each play

set.

Figure 4 Vincent’s unscripted vocals for baseline, mastery probes, and alternative material probes for each play set.

Figure 5 Peter’s scripted vocals and actions and recombined vocals and actions for each play set.

Figure 6 Peter’s unscripted vocals for baseline, mastery probes, and alternative material probes for each play set.

Figure 7 Combined percentage of vocals and actions for all participants.

Video Modeling and Matrix Training 57

Figure 1. Video Modeling and Matrix Training 58

Unscripted Vocalizations

20

18 Bank Mansion Castle 16

14

12

10

8

6

4

2

0

Figure 2.

Video Modeling and Matrix Training 59

Probes after Recombination Instruction

Probes after Recombination Instruction

Probes after Recombination Instruction

Figure 3. Video Modeling and Matrix Training 60

Unscripted Vocalizations

10 9

8 Castle Bank Mansion 7 6 5 4 3 2 1 0 Probe Probe Probe Probe Probe 2 Probe Baseline Baseline 5 Baseline 1 Baseline 6 Baseline 1 Baseline 7 Aternative Aternative Materials Probe 3 Alternative Alternative Materials Probe 1 Alternative Materials Probe 2 Alternative Materials Probe 3 Alternative Materials Probe 4 Alternative Materials Probe 5 Alternative Materials Probe 6 Alternative Materials Probe 1 Alternative Materials Probe 2 Alternative Materials Probe 3 Alternative Materials Probe 5 Alternative Materials Probe 6 Alternative Materials Probe 1 Alternative Materials Probe 2 Alternative Alternative Materials Probe 4 Alternative Materials Probe 4 Figure 4.

Video Modeling and Matrix Training 61

Probes after 1 Recombination 0.9 Instruction 0.8 0.7 Mastery Probes 0.6 0.5 Vocals 0.4 Baseline Recombined Actions 0.3 Probes 0.2 0.1 Training Bank 0 -0.1 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96

Probes after Recombination 1 Instruction 0.9 Mastery Probes 0.8 0.7 0.6 Baseline 0.5 Vocals 0.4 Recombined 0.3 Probes Actions 0.2 0.1 Training 0 Mansion -0.1 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96

1 0.9 0.8 Probes after Recombined Recombination 0.7 Probes Instruction 0.6 0.5 Mastery Probe 0.4 Baseline Vocals 0.3 Actions 0.2 0.1 Castle 0 -0.1 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96

Figure 5 Video Modeling and Matrix Training 62

Unscripted Vocalizations 2.5

2

Mansion Castle 1.5 Bank

1

0.5

0 Probe castle castle Probe Probe Probe 2 Probe 3 Probe 2 Probe Mansion Baseline Baseline 1 Baseline 8 Baseline 1 Baseline 9 Baseline 1 Baseline Baseline 10 Alternative Alternative Materials Probe 1 Alternative Materials Probe 2 Alternative Materials Probe 3 Alternative Materials Probe 4 Alternative Materials Probe 5 Alternative Materials Probe 6 Alternative Materials Probe 7 Alternative Materials Probe 8 Alternative Materials Probe 1 Alternative Materials Probe 2 Alternative Materials Probe 3 Alternative Materials Probe 4 Alternative Materials Probe 5 Alternative Materials Probe 6 Alternative Materials Probe 1 Alternative Materials Probe 2 Alternative Materials Probe 3 Alternative Materials Probe 4

Figure 6.

Video Modeling and Matrix Training 63

Figure 7.

Video Modeling and Matrix Training 64

Appendices

Appendix 1. Bank play set and materials

Video Modeling and Matrix Training 65

Appendix 2. Mansion play set and materials

Video Modeling and Matrix Training 66

Appendix 3. Castle play set and materials

Video Modeling and Matrix Training 67

Appendix 4 . Bank scenario script

OBJECTS ACTIONS (41) VOCALS (30)

Bank playset & Joker Joker walks in from the left Joker “Ha, ha, ha! I’m going to steal all the money from the bank!” Joker Walk Joker to the white door White door Open white door Joker “Rats! It’s not in this room!” White door Close white door Joker Walk Joker to the black door Black door Open black door Joker “Double rats! It’s not in this room!” Black door Close black door Joker “The money must be in the vault!” Joker Walk Joker to the vault door Joker “I’ll use the dynamite to open the vault TNT Button Push TNT button “BOOM!” Joker “Wow! That’s a lot of money! I’m going to be rich! Ha, ha, ha!” Joker & money Pick Joker up, place money on arm. Joker “I’ll climb to the top of the bank to wait for my getaway driver to arrive!” Joker & money Climb Joker to the roof of the bank Joker “Now I’ll be able to see any silly hero coming here to try and stop me before they see me!” Batcave playset “Ring, ring!” Batcave playset Slide yellow button to the right to open purple door revealing Batman Batman “An alarm at the Batcomputer, I’ll answer it!” Batman Pick up Batman, jump him down to the ground Batman Walk over to the Batcomputer “Someone has stolen the money from the bank!” Batman “It must be the Joker!” Batman “I’ll drive the Batmobile to the bank” Batcave Turn red circle beside middle blue door to the right Batmobile Reach through and pick up the Batmobile, drive through ramp Batman Walk Batman over to the Batmobile “Time to get in” Batmobile Open roof Batman & Batmobile Pick Batman up and sit him down into the Batmobile Batman “I’ll need the Bat-a-rang” Bat-a-rang Pick up Bat-a-rang from black alcove Bat-a-rang Put in Batmobile beside Batman Batmobile Close roof Batman in Batmobile “To the bank!” Batman in Batmobile “I must stop the Joker!” Batman in Batmobile Drive Batmobile to front of the bank “VROOM! VROOM!” Batman in Batmobile “I’ve arrived at the bank” Batmobile Open roof of Batmobile Bat-a-rang Pick Bat-a-rang out of the Batmobile Batman Pick Batman out of the Batmobile Batman & Bat-a-rang Pick up Bat-a-rang, put in Batman’s hand Batman “Now time to find the Joker!” Batman Walk Batman to the white door White door Open white door Batman “He’s not in here” White door Close white door Batman Walk Batman to the black door Black door Open black door Batman “He’s not in here either! Where is he?” Black door Close black door Joker “Oh no! That FATman has shown up to stop me! I’ll have to take care of him!” Joker Jump Joker down to the ground, facing towards Batman Batman Turn Batman to face Joker Batman “Joker! You’ll never get away with stealing the money!” Joker “You’re too late Batman! Ha, ha! Once I deal with you, I’ll be away scot-free! Ha, ha, ha!” Batman “I don’t think so Joker!” Batman & Bat-a-rang Pull Bat-a-rang off of Batman’s hand Batman “Take this!” Bat-a-rang Spin through air towards Joker Bat-a-rang & Joker Bat-a-rang hits Joker and knocks him over Joker “AHHH!” Joker & money Pull money off Joker Batman Walk Batman over to Joker Batman & money Put money on Batman’s arm Batman “Remember Joker, crime doesn’t pay!”

Video Modeling and Matrix Training 68

Appendix 5 . Mansion scenario script

OBJECTS ACTIONS (30) VOCALS (29)

Mansion playset & Penguin walks in from the right to the middle of the mansion playset Penguin Penguin “Ha, ha, ha! I’m going to steal the world’s biggest diamond from the mansion!” Penguin Walk Penguin to the window on the right Penguin “Rats! It’s not in this room!” Penguin Walk Penguin to the window on the left Penguin “Double rats! It’s not in this room!” Penguin Walk Penguin to the middle of the mansion Penguin “The diamond must be on the 2 nd floor” Penguin “I’ll climb up the side of the mansion to get it!” Penguin Climb Penguin up the side of the mansion to the 2 nd floor Penguin “Wow! That’s a really big diamond! I’m going to be rich! Ha, ha, ha!” Penguin & diamond Pick Penguin up, place diamond on arm. Penguin “I’ll climb to the top of the mansion to wait for my getaway driver to arrive!” Penguin & diamond Climb Penguin up to the roof of the mansion Penguin “Now I’ll be able to see any silly hero coming here to try and stop me before they see me!” Batcave playset “Ring, ring!” Batcave playset Turn red circle at the top right of the Batcave playset to raise the “An alarm at the Batcomputer, I’ll answer it!” elevator Robin Pick up Robin from the elevator Robin Climb Robin down the purple ladder Robin Walk over to the Batcomputer “Someone has stolen the diamond from the mansion!” Robin “It must be the Penguin!” Robin “I’ll drive the Batcycle to the mansion” Batcave & Batcycle Reach through left hand blue door and pick up the Batcycle, drive through ramp Robin Walk Robin over to the Batcycle “Time to get in” Robin & Batcycle Pick Robin up, and sit him down into the Batcycle Robin “I’ll need the Bat-a-rang” Bat-a-rang Pick up Bat-a-rang from black alcove Bat-a-rang Put in Batcycle beside Robin Robin in Batcycle “To the mansion!” Robin in the Batcycle “I must stop the Penguin!” Robin in Batcycle Drive Batcycle to front of the mansion “VROOM, VROOM!” Robin in Batcycle “I’ve arrived at the mansion” Bat-a-rang Pick Bat-a-rang out of the Batcycle Robin Pick Robin out of the Batcycle Robin & Bat-a-rang Pick up Bat-a-rang, put in Robin’s hand Robin “Now time to find the Penguin!” Robin Walk Robin to the window on the left Robin “He’s not in here” Robin Walk Robin to the window on the right Robin “He’s not in here either! Where is he?” Penguin “Oh no! That bird-brain Robin has shown up to stop me! I’ll have to take care of him!” Penguin Jump Penguin down to the bottom of mansion, facing towards Robin Robin Turn Robin to face Penguin Robin “Penguin! You’ll never get away with stealing the diamond!” Penguin “You’re too late Robin! Ha, ha! Once I deal with you, I’ll be away scot-free! Ha, ha, ha!” Robin “I don’t think so Penguin!” Robin & Bat-a-rang Pull Bat-a-rang off of Robin’s hand Robin “Take this!” Bat-a-rang Spin through air towards Penguin Bat-a-rang & Penguin Bat-a-rang hits Penguin and knocks him over Penguin “Oh no! The diamond!” Penguin & diamond Pull diamond off Penguin Robin Walk Robin over to Penguin

Robin & diamond Put diamond on Robin’s hand Robin “Remember Penguin, crime doesn’t pay!”

Video Modeling and Matrix Training 69

Appendix 6 . Castle scenario script

OBJECTS ACTIONS (36) VOCALS (29)

Castle playset & Mr. Mr. Freeze walks in from the right to the middle of the castle playset Freeze Mr. Freeze “Ha, ha, ha! I’m going to steal the world’s most expensive ring from the castle!” Mr. Freeze Walk Mr. Freeze to the window on the right, jump up to look in Mr. Freeze “Rats! It’s not in this room!” Mr. Freeze Walk Mr. Freeze to the window on the left, jump up to look in Mr. Freeze “Double rats! It’s not in this room!” Mr. Freeze “The ring must be hidden in the castle’s storage room”” Mr. Freeze Walk Mr. Freeze over to green alcove at the left side of the castle Mr. Freeze “Wow! That’s a really beautiful ring! I’m going to be rich! Ha, ha, ha!” Mr. Freeze & ring Pick Mr. Freeze up, place ring on arm. Mr. Freeze “I’ll climb to the top of the castle to wait for my cold cronies to help me escape!” Mr. Freeze Climb Mr. Freeze up steps on the left wall of the mansion Mr. Freeze Walk over to top right of castle Mr. Freeze “Now I’ll be able to see any silly hero coming here to try and stop me before they see me!” Batcave playset “Ring, ring!” Batcave playset Pull yellow rail at the right of the Batcave playset to reveal Superman “An alarm at the Batcomputer, I’ll answer it!” Superman Pick up Superman from the ledge Superman Fly Superman down to the centre of the playset Superman Walk over to the Batcomputer “Someone has stolen the ring from the castle!” Superman “It must be the Mr. Freeze!” Superman “I’ll fly the Batcopter to the castle” Batcopter Fly Batcopter down to the middle of the Batcave playset Batcopter Spin propellers “Choak, Choak, Choak” Superman Walk Superman over to the Batcopter “Time to get in” Batcopter Open Batcopter roof Superman & Batcopter Pick Superman up, bend legs and sit him down into the Batcopter Batman “I’ll need the Bat-a-rang” Bat-a-rang Pick up Bat-a-rang from black alcove Bat-a-rang Put in Batcopter beside Superman Batcopter Close Batcopter roof Superman in the “To the castle!” Batcopter Superman in the “I must stop the Mr. Freeze!” Batcopter Batcopter Lift Batcopter into the air “Choak, choak, choak” Batcopter Spin propeller Superman in Batcopter Fly Batcopter to front of the castle Superman in Batcopter “I’ve arrived at the castle” Batcopter Open roof Bat-a-rang Pick Bat-a-rang out of the Batcopter Superman Pick Superman out of the Batcopter Superman & Bat-a-rang Pick up Bat-a-rang, put in Superman’s hand Superman “Now time to find the Mr. Freeze!” Superman Walk Superman to the window on the right, jump up to look in Superman “He’s not in here” Superman Walk Superman to the window on the left, jump up to look in Superman “He’s not in here either! Where is he?” Mr. Freeze “Oh no! That stupid Superman has shown up to stop me! I’ll have to take care of him!” Mr. Freeze Jump Mr. Freeze down to the right hand side of mansion, facing towards Superman Superman Turn Superman to face Mr. Freeze Superman “Mr. Freeze! You’ll never get away with stealing the ring!” Mr. Freeze “You’re too late Superman! Ha, ha! Once I deal with you, I’ll be away scot-free! Ha, ha, ha!” Superman “I don’t think so Mr. Freeze!” Superman & Bat-a-rang Pull Bat-a-rang off of Superman’s hand Superman “Take this!” Bat-a-rang Spin through air towards Mr. Freeze Bat-a-rang & Mr. Freeze Bat-a-rang hits Mr. Freeze and knocks him over Mr. Freeze “AHHH!” Mr. Freeze & ring Pull ring off Mr. Freeze Superman Walk Superman over to Mr. Freeze Superman & ring Put ring on Superman’s arm Superman “Remember Mr. Freeze, crime doesn’t pay!”

Video Modeling and Matrix Training 70

Appendix 7. 3 dimensional matrices with black shaded boxes indicating trained scripts

Video Modeling and Matrix Training 71

Appendix 8. 3 dimensional matrices with dark grey shaded boxes indicating possible emerging scripts in participant 1 alternative probes

Video Modeling and Matrix Training 72

Appendix 9. Scoring guidelines for bank scenario script

Bank Playset: Action Guidelines General Scoring rules • Actions still count if a character/piece is placed somewhere and then falls off • Actions can occur out of order from the original script Walk Joker to the white door Joker doesn’t have to contact ground when walking Open White door Door can be opened from front or back of the playset Walk Joker to the black door Joker doesn’t have to contact ground when walking Open Black door Door can be opened from front or back of the playset Walk Joker to vault door Joker doesn’t have to contact ground when walking Joker & money may stand on the left or right of the bank. Money may or may not be Climb Joker to the roof of the bank on Joker’s arm but must be on bank roof Slide yellow button to the right to open Batman must be on the other side of the purple door purple door revealing Batman Walk to the Batcomputer Batman doesn’t have to contact ground when walking Turn red circle beside middle blue door to Do not score correct if Batmobile is not behind blue door the right Batman may not contact ground when walking but must go to the Batmobile, not Walk Batman over to the Batmobile Batmobile to where Batman is Put (Bat-a-rang) in Batmobile beside Still score correct even if participant does not retrieve from black alcove. May sit on Batman top of Batman in Batmobile. Batmobile can drive in straight line or side to side. Do not score correct if participant Drive Batmobile to front of the bank drives Batmobile but not to bank. Do not score if Batman is not in Batmobile. Pick up Bat-a-rang, put on Batman’s hand Bat-a-rang must clip into Batman’s hand Batman doesn’t have to contact ground when walking. Still score correct even if Bat- Walk Batman to the white door a-rang is not in Batman’s hand Open White door Door can be opened from front or back of the playset Batman doesn’t have to contact ground when walking. Still score correct even if Bat- Walk Batman to the black door a-rang is not in Batman’s hand Open Black door Door can be opened from front or back of the playset Turn Batman to face Joker Still correct if Batman is already facing Joker Pull Bat-a-rang off of Batman’s hand Bat-a-rang must already be in Batman’s hand Spin (Bat-a-rang) through air towards Joker Still score correct if Bat-a-rang doesn’t spin Still score correct if money falls off Joker’s hand when getting knocked over with Bat- Pull money off Joker a-rang Batman doesn’t have to contact ground when walking but must go to Joker, not Joker Walk Batman over to Joker to where Batman is Bank Playset: Vocal Guidelines General Scoring rules: • Scored as + if it is complete and matches the script. • Scored as + if it is a partial vocal (at least 50%) or a paraphrase of a vocal (example: script says “Remember Joker, crime doesn’t pay!” and student says “Crime doesn’t pay Joker!”) • Student does not have to be holding/manipulating character for + • Do not score as + if it is with the wrong character • Vocals are only counted once, but do not count against student if they repeat them • Scored as + if they occur out of order from the original script

Video Modeling and Matrix Training 73

Appendix 10 . Scoring guidelines for mansion scenario script

Mansion Playset: Action Guidelines General Scoring rules • Actions still count if a character/piece is placed somewhere and then falls off • Actions can occur out of order from the original script Walk Penguin to the window on the right Penguin doesn’t have to contact ground when walking Walk Penguin to the window on the left Penguin doesn’t have to contact ground when walking Penguin & diamond may stand on the left or right of the bank. Diamond may or may Climb Penguin to the roof of the mansion not be on Penguin’s arm but must be on mansion roof Turn red circle at the top right of the Robin must be in the elevator Batcave playset to raise the elevator Climb Robin down the purple ladder Robin touches purple ladder. Do not score correct if Robin Jumps down or flies down Robin doesn’t have to contact ground when walking. Score correct if Robin has Walk to the Batcomputer directly climbed to Batcomputer Robin may not contact ground when walking but must go to the Batcycle, not Batcycle Walk Robin over to the Batcycle to where Robin is Put (Bat-a-rang) in Batcycle beside Robin Still score correct even if participant does not retrieve from black alcove Batcycle can drive in straight line or side to side. Do not score correct if participant Drive Batcycle to front of the bank drives Batcycle but not to mansion. Do not score if Robin is not in Batcycle. Pick up Bat-a-rang, put on Robin’s hand Bat-a-rang must clip into Robin’s hand Robin doesn’t have to contact ground when walking. Still score correct even if Bat-a- Walk Robin to the window on the right rang is not in Robin’s hand Robin doesn’t have to contact ground when walking. Still score correct even if Bat-a- Walk Robin to the window on the left rang is not in Robin’s hand Turn Robin to face Penguin Still correct if Robin is already facing Penguin Pull Bat-a-rang off of Robin’s hand Bat-a-rang must already be in Robin’s hand Spin (Bat-a-rang) through air towards Still score correct if Bat-a-rang doesn’t spin Penguin Still score correct if diamond falls off Penguin’s hand when getting knocked over with Pull diamond off Penguin Bat-a-rang Robin doesn’t have to contact ground when walking but must go to Penguin, not Walk Robin over to Penguin Penguin to where Robin is Mansion Playset: Vocal Guidelines General Scoring rules: • Scored as + if it is complete and matches the script. • Scored as + if it is a partial vocal (at least 50%) or a paraphrase of a vocal (example: script says “Remember Penguin, crime doesn’t pay!” and student says “Crime doesn’t pay Penguin!”) • Student does not have to be holding/manipulating character for + • Do not score as + if it is with the wrong character • Vocals are only counted once, but do not count against student if they repeat them • Scored as + if they occur out of order from the original script

Video Modeling and Matrix Training 74

Appendix 11. Scoring guidelines for castle scenario script

Castle Playset: Action Guidelines General Scoring rules • Actions still count if a character/piece is placed somewhere and then falls off • Actions can occur out of order from the original script Walk Mr. Freeze to the window on the Mr. Freeze doesn’t have to contact ground when walking right, jump to look in Walk Mr. Freeze to the window on the left, Mr. Freeze doesn’t have to contact ground when walking jump to look in Walk Mr. Freeze to the green alcove at the Mr. Freeze doesn’t have to contact ground when walking left side of the castle Climb Mr. Freeze up steps on the left wall Mr. Freeze touches wall of the castle where steps are. Do not score correct if Mr. of the castle Freeze jumps up or flies up Walk Mr. Freeze to the top right of the Mr. Freeze doesn’t have to contact ground when walking. Ring may or may not be on castle Mr. Freeze’s arm but must be on castle roof Pull yellow rail at the right of the Batcave Superman must be on the platform playset to reveal Superman Fly Superman down to the centre of the Superman must fly down. Do not score correct if Superman uses the purple ladder to playset climb down. Score correct if Superman flies directly to Batcomputer. Superman doesn’t have to contact ground when walking. Score correct if Superman Walk to the Batcomputer has directly flown to Batcomputer Superman doesn’t have to contact ground when walking but must go to the Batcopter, Walk Superman over to the Batcopter not Batcopter to where Superman is Put (Bat-a-rang) in Batcopter beside Still score correct even if participant does not retrieve from black alcove Superman Batcopter can fly in straight line or side to side. Do not score correct if participant flies Fly Batcopter to front of the castle Batcycle but not to castle. Do not score if Superman is not in Batcopter. Pick up Bat-a-rang, put on Superman’s Bat-a-rang must clip into Superman’s hand hand Walk Superman to the window on the right, Superman doesn’t have to contact ground when walking. Still score correct even if jump to look in Bat-a-rang is not in Superman’s hand Walk Superman to the window on the left, Superman doesn’t have to contact ground when walking. Still score correct even if jump to look in Bat-a-rang is not in Superman’s hand Turn Superman to face Mr. Freeze Still correct if Superman is already facing Mr. Freeze Pull Bat-a-rang off of Superman’s hand Bat-a-rang must already be in Superman’s hand Spin (Bat-a-rang) through air towards Mr. Still score correct if Bat-a-rang doesn’t spin Freeze Still score correct if ring falls off Mr. Freeze’s hand when getting knocked over with Pull ring off Mr. Freeze Bat-a-rang Superman doesn’t have to contact ground when walking but must go to Mr. Freeze, Walk Superman over to Mr. Freeze not Mr. Freeze to where Superman is Castle Playset: Vocal Guidelines General Scoring rules: • Scored as + if it is complete and matches the script. • Scored as + if it is a partial vocal (at least 50%) or a paraphrase of a vocal (example: script says “Remember Mr. Freeze, crime doesn’t pay!” and student says “Crime doesn’t pay Mr. Freeze!”) • Student does not have to be holding/manipulating character for + • Do not score as + if it is with the wrong character • Vocals are only counted once, but do not count against student if they repeat them • Scored as + if they occur out of order from the original script

Video Modeling and Matrix Training 75