SPSS 8.01 for Windows was used to analyse the data. The mean pre and post VRDE T-scores on the Hyperactivity Index and the 7 CPRS-93 subscales for the AD/HD and control groups are displayed in Figures 3 and Figure 4.

Figure 3
Mean Pre and Post VRDE T-Scores for CPRS-93 Hyperactivity Index and Subscales

A 2 (group identification) × 2 (time) repeated measures Analysis of Variance (ANOVA) was conducted with time (of questionnaire administration) as the within subjects variable for the Hyperactivity Index and each of the 7 subscales of the CPRS-93.

For the Hyperactivity Index, a significant difference was found between mean T-scores obtained on the first and second administration of the CPRS-93 (F(1,45) = 36.10, p < .001, n² = .445), as well as a significant difference between mean T-scores obtained by each group (F(1,45) = 83.30, p < .001, n² = .649).

Figure 4
Mean Pre and Post VRDE T-Scores for CPRS-93 Subscales

However since a significant interaction between group identification and time was found (F(1,45) = 10.04, p < .01, n² = .182), these significant main effects must be viewed with caution. Separate analyses, comparing each group's mean CPRS-93 Hyperactivity Index T-scores before and after the VRDE, using paired samples T-tests, were conducted for each group. A significant difference in mean Hyperactivity Index T-scores obtained on questionnaires administered before and after the VRDE was found for the AD/HD group (t(25) = 4.44, p < .001), and the control group (t(20) = 2.65, p < .05).

For the Conduct Disorder subscale there was a significant difference between mean T-scores obtained on the first and second questionnaire (F(1,45) = 22.81, p < .001, n² = .336), as well a significant difference between the two groups' mean T-scores (F(1,45) = 58.11, p < .001, n² = .564). However the significant interaction between time of questionnaire administration and group identification (F(1,45) = 7.14, p =.01, n² = .137) necessitated separate paired samples T-tests of pre and post VRDE T-scores for the each of the two groups. There was a significant difference in mean T-scores obtained prior to the VRDE, and following it, for the control group (t(20) = 2.65, p < .05) and the ADD group (t(25) = 4.44, p < .001).

Mean Anxious-Shy subscale T-scores were significantly different after the VRDE session (F(1,45) = 37.17, p < .001, n² = .452) and also differed significantly between groups (F(1,45) = 8.33, p < .01, n² = .156). The interaction between group and time of questionnaire administration was not significant (F(1,45) = 2.76, p > .05). Separate analyses for the AD/HD and control groups revealed a significant difference between T-scores obtained before and after the VRDE for the AD/HD group (t(25) = 5.39, p <. 001) and the control group (t( 20) = 3.33, p < .01).

Mean Restless-Disorganized subscale T-scores differed significantly between the two administrations of the CPRS-93 (F(1,45) = 27.34, p < .001, n² = .378), as well as between the two groups (F(1,45) = 39.91, p < .001, n² = .47), however there was a significant interaction between the group identification and time of administration (F(1,45) = 5.77, p < .05, n² = .114), indicating that caution is required in the interpretation of the significant main effects. Separate analyses were conducted for the AD/HD and control groups. For both these groups a significant difference was found in mean T-scores obtained on the CPRS-93 before and after the VRDE (AD/HD: (t(25) = 4.68, p < .001), control: (t(20) = 3.04, p < .01).

Mean T-scores obtained on the Psychosomatic subscale differed significantly before and after the VRDE (F( 1,45) = 16.31, p < .001, n² = 2.66 ), as well as between the groups (F(1,45) = 36.89, p < .001, n² = .450). However there was a significant interaction between administration time and group identification (F(1,45) = 13.38 , p = .001, n² = .229), requiring cautious interpretation of these significant main effects. Separate analyses indicated there was a significant difference in pre and post VRDE CPRS-93 scores on the Psychosomatic subscale for both AD/HD (t(25) = 4.74, p < .001) and control groups (t(20) = 4. 90, p < .001).

Mean Obsessive Compulsive subscale T-scores were significantly different after the VRDE session (F(1,45) = 38.84, p < .001, n² = .174), and also differed significantly between groups (F(1,45)=11.03, p < .01, n² = .197). The interaction between group and time of questionnaire administration was not significant (F(1,45) = 1.96, p > .05). Separate analyses comparing each group's mean CPRS-93 Obsessive Compulsive subscale T-scores before and after the VRDE, using paired samples T-tests, found significant differences for the ADD group (t(25) = 2.56, p < .05), but not the control group (t(20) = 1.86, p > .05).

Mean Antisocial subscale T-scores were significantly different after the VRDE session (F(1,45) = 8.14, p < .01, n² = 1.53) and also differed significantly between groups(F(1,45) = 14.91, p < .001 n² = .249). The interaction between group and time of questionnaire administration was not significant (F(1,45) = 3.30, p > .05). Separate analyses comparing each group's mean CPRS-93 Antisocial subscale T-scores before and after the VRDE, using paired samples T-tests, found significant differences for the ADD group (t(25) = 2.65 p < .05) and for the control group (t(20) = 2.39, p < .05).

Mean Hyperactive-Immature subscale scores were significantly different before and after the VRDE (F(1,45) = 51.56, p < .001, n² = .534), and between the groups (F(1,45) = 45.49, p < .001, n² = .503). However there was a significant interaction between the two (F(1,45) = 8.00, p < .01, n² = .151), necessitating cautious interpretation of the significant main effects Separate analyses comparing each group's Hyperactive - Immature CPRS-93 subscale T-scores before and after the VRDE, using paired samples T-tests, found significant differences for the AD/HD group (t(25) = 18.70 p < .001) and for the control group (t(20) = 15.46, p < .001).

There was a significant difference in the scores obtained before and after the VRDE on the Hyperactivity Index and all 7 subscales for the AD/HD group, and on the Hyperactivity Index and 6 subscales for the control group. Further analysis was warranted to assess if there was a difference in the mean magnitude of score change pre and post VRDE, between the groups. Difference scores were calculated as the score obtained after the VRDE subtracted from the score obtained prior to the VRDE, for each subject. Means of these scores are displayed in Figure 5.

Figure 5
Mean CPRS-93 Hyperactivity Index and Subscales Difference Scores

An Independent Groups T-test was conducted on the AD/HD and control groups' mean CPRS-93 subscale T-score differences. Significant differences in mean magnitude of score change between the AD/HD and control groups was found for the Hyperactivity Index (t(42.43) = 3.33, p < .01), Conduct Disorder subscale (t(35.30) = 2.88, p < .01). Restless Disorganised subscale (t (38.77) = 2.56, p < .05) and the Hyperactive - Immature subscale (t(45) = 2.83, p < .01.) The remaining 4 subscales' mean change scores did not differ significantly between the two groups, however the Antisocial subscale's significance was borderline, (t(28) = 2.00, p = .055.).

To evaluate whether subjects' age, gender, medication status and their VRDE session duration affected the magnitude of the change between the T-scores obtained on the first and second questionnaire, a multiple regression was performed for each CPRS-93 subscale. The regression statistics appear in table 1.

Table 1
Regression Statistics for CPRS-93 Hyperactivity Index and Subscales Difference Scores and Demographic Variables

As displayed in table 1, there was only a significant relationship between group identification and the change score obtained on all subscales and the Hyperactivity Index. There was no significant relationship between age, gender or session duration and change scores on any of the subscales or the hyperactivity index. The Antisocial subscale was the only one where change scores and medication status had a significant relationship.

Although there was sufficient data for the regressions, the small sample size suggested that confirmatory correlations using Pearson's R would be useful. These correlations and their significance levels are displayed in table 2.

Table 2
Correlation between Group(AD/HD or Control) and CPRS-93 Hyperactivity Index and Subscales

"...the dolphins can see me - I know they can, I know they can!" (4 years)

"...I can see the bottom of the water..." (5 years)

" ...I was there..." (15 years)

"...sharks, they are sharks!' (5 years)

"...I can reach out and touch them!" (9 years)

"...it's good Mum ...I wish I had one of these at home ...I could watch for hours..." (8 years)

"...I want to watch more ...I want to watch till the end..." (12 years)

"...cool..." (10 years)

"...now where am I? ...who are the dolphins talking to? ...did you hear that Mum..?" (7 years)

please click to return to the table of contents