C- Hg: A COLLABORATIVE HAPTIGRIPPER VIRTUAL REALITY SYSTEM
6.5 Results and Discussions
6.5.3 Impact of Fine Motor Skill Training on Social Skills
The collaborative performance results are shown in Table 6-5. We found statistically significant improvements of both groups regarding the task score in the collaborative tasks (Collaborative Prize Claw tasks: RC = 9.5%, Z = -2.55, p = .011, r = -.57; Collaborative Green Path tasks: RC = 16.75%, Z = -2.8, p
= .005, r = -.63). The Spearman’s rank correlation analysis also found strong/medium correlations between the performance results of individual tasks and collaborative tasks (ASD in Prize Claw tasks: ρ = 0.498, p
= .025; TD in Prize Claw tasks: ρ = 0.208, p = .379; ASD in Green Path tasks: ρ = 0.427, p = .061; TD in Green Path tasks: ρ = 0.574, p = .008), which indicated that the improvements in collaborative tasks were related to improvements in individual tasks. These results might suggest that through practicing fine motor
skills in the individual training tasks, participants were more likely to perform better in collaborative fine motor tasks, which required communication and collaborative operations.
Table 6-5. Collaborative Performance Results of TD-ASD Pairs (N = 20, Pairs = 10)
Metrics Pre Post RC
(%) Z p r
Mdn Mdn
Col PC 9.5 12.75 34.2 -2.55 .011* -.57† Col GP 16.75 31 85.0 -2.80 .005* -.63† ASD PC Col Ratio 0.49 0.63 26.6 -2.40 .017* -.54† TD PC Col Ratio 0.57 0.65 13.2 -1.58 .114 -.35 ASD GP Col Ratio 0.75 0.81 7.85 -1.99 .047* -.44 TD GP Col Ratio 0.77 0.81 4.94 -1.27 .203 -.29 ASD Col PC INIT 12 15.25 27.1 -0.77 .44 -.17 TD Col PC INIT 20.75 15.25 -26.5 1.27 .20 .28 ASD Col PC RESP 10 6 -.4 0.49 .62 .11 TD Col PC RESP 8.47 7 -.2 0.35 .72 .08 PC Col INIT Switch 15.25 16.25 6.56 0.49 .62 .11 ASD Col GP INIT 13.25 18.75 41.5 -2.14 .033* -.48
TD Col GP INIT 17.75 18.5 4.23 -1.48 .139 -.33 ASD Col GP RESP 5.75 5.5 -4.35 0.41 .682 .09
TD Col GP RESP 5.75 4.75 -17.4 1.07 .282 .24 GP Col INIT Switch 14.5 17 17.2 -1.89 .059 -.42
Col PC, collaborative Prize Claw task score Col GP, collaborative Green Path task score Col Ratio, the ratio of collaborative operations to total operations
INIT, the frequency of initiating a conversation RESP, the response frequency
INIT Switch, the switch frequency of conversation initiator RC, relative change computed by (post - pre)/pre *100%
*p<.05, †|r|>0.5
We also found that all participants, whether with ASD or TD performed much more collaborative manipulations in the collaborative tasks of the post-test. Specifically, participants with ASD significantly increased their collaborative manipulations in the collaborative tasks of post-test (Collaborative Ratio of ASD in collaborative Prize Claw tasks: RC = 26.6%, Z = -2.4, p = .017, r = -.54; Collaborative Ratio of ASD in collaborative Green Path tasks: RC = 7.85%, Z = -1.99, p = .047, r = -.44). These results indicated increased interaction and collaboration between partners in the collaborative tasks of post-test, and might suggest that improved fine motor skills would promote collaborative activities of participants.
To analyze the conversation pattern of participants, two human coders (trained graduate students with
experience collecting and analyzing qualitative and quantitative data) were recruited to manually transcribed and coded the conversation audios from 80 recorded videos of collaborative tasks (four hours of video data in total). We provided a framework that described the concrete definitions of types of utterances, and a few examples about how to code the conversation data for the human coders. Each human coder independently coded the same data. After each rater completed coding, a percent agreement of 92.5%
was found. For the codes that were not in agreement, the human coders reconciled differences via a consensus in which any discrepancies were discussed and resolved.
As described in Section 6.3, the conversations between two players in all the collaborative tasks involved strategy discussion and information sharing. Timely communication was important for players to obtain higher scores. In order to evaluate how often one player communicated with his/her partner, we defined three types of utterances, Initiation (INIT), Response (RESP), and Initiation Switch (INIT Switch). Initiation represented one player’s statement that started a conversation. Response represented one player’s feedback to the partner’s statement. Initiation Switch represented the switch of the conversation initiator from one player to the other one. Based on the final coding data, we calculated the frequency of INIT, RESP, and INIT Switch in each collaborative task. Table 6-6 shows a sample of three conversations recorded in a collaborative Green Path task. First, Player A started a conversation asking where to go, and Player B responded with the direction information (recall that path position was only visible to one player). Second, Player B started a new conversation to direct the movement. Then, Player A started another conversation to provide the reward information (recall that reward position was only visible to one player), and Player B responded with an acknowledgement. Therefore, there were two INITs of Player A, one INITs of Player B, and two RESPs of Player B. We also see that the conversation initiator switched from Player A to Player B, and then switched back to Player A. Thus, we counted the frequency of INIT Switch as 2.
Table 6-6. A sample of conversations recorded in a collaborative Green Path task.
No. Player A Player B
1 Now where will we go? <Initiation>
Left. <Response>
2 Right. Down. Right, right. <Initiation>
3 Wait, right here. There is a reward. <Initiation>
Okay. <Response>
As shown in Table 6-7, the results indicated that in all the collaborative tasks of the pre-test and post- test, both the ASD and TD participants initiated more conversations than providing responses, which is reasonable since participants could respond to his/her partner with an action instead of an utterance.
However, we found that the TD participants maintained a significant difference between initiations and
responses, no matter in the pre-test or in the post-rest (TD in Collaborative Prize Claw tasks of Pre-test: Z
= 2.81, p = .005, r = -.63; TD in Collaborative Prize Claw tasks of Post-test: Z = 2.66, p = .008, r = -.59;
TD in Collaborative Green Path tasks of Pre-test: Z = 2.70, p = .007, r = -.60; TD in Collaborative Green Path tasks of Post-test: Z = 2.80, p = .005, r = -.63), while the difference among ASD participants is only significant in the post-test (ASD in Collaborative Green Path tasks of Post-test: Z = 2.30, p = .021, r = - .51; ASD in Collaborative Green Path tasks of Post-test: Z = 2.80, p = .005, r = -.63). The results showed in Table 6-5 also indicated that ASD participants made more initiations in the post-test, and the significant improvements were found in the Collaborative Green Path tasks (Z = -2.14, p = .033, r = -.48). These results might suggest increased active communication among the ASD participants, and they tended to provide information or commands in the post-test. In addition, in the post-test, the switch frequency of who initiated a conversation increased (INIT Switch in collaborative Prize Claw tasks: RC = 6.56%, Z = 0.49, p = .62, r = -.11; INIT Switch in collaborative Green Path tasks: RC = 17.2%, Z = -1.89, p = .059, r = - .42), though no significant difference was found, which also suggested that participants more actively and effectively communicated with their partners to share information or provide commands in the post-test collaborative tasks. Considering the improvements in task performance, these results might suggest that having the opportunity to practice fine motor skills could foster more active communication.
Table 6-7. The comparison between initiation frequency and response frequency among the ASD group (N = 10) and TD group (N = 10)
Metrics INIT RESP
Z p r
Mdn Mdn
ASD Col PC PRE 12 10 1.74 .082 .39 TD Col PC PRE 20.75 8.47 2.81 .005* .63† ASD Col PC POST 15.25 6 2.30 .021* .51† TD Col PC POST 15.25 7 2.66 .008* .59† ASD Col GP PRE 13.25 5.75 1.38 .169 .31 TD Col GP PRE 17.75 5.75 2.70 .007* .60† ASD Col GP POST 18.75 5.5 2.80 .005* .63† TD Col GP POST 18.5 4.75 2.80 .005* .63†
Col PC, collaborative Prize Claw task score Col GP, collaborative Green Path task score INIT, the frequency of initiating a conversation
RESP, the response frequency
*p<.05, †|r|>0.5