A significant difference between before and after behaviour in both the AD/HD and control groups was found. There was also a significant difference in magnitude of improvement between the AD/Hd and control groups on five of the eight measures of behaviour. Age, session duration, medication status and gender were not found to significantly influence the magnitude of improvement. However whether subjects belonged to the AD/HD or control group was the significant determinant of magnitude of improvement. The results of significant differences in behaviour before and after the VRDE, and larger differences in before and after behaviour for the AD/HD group, fulfilled the expectations of this study. The large amount of analyses conducted on results obtained with a relatively small sample size, must be considered when assessing the importance of the present findings. Although it was beyond the scope of the present study, future investigations could undertake a more detailed analysis of potential changes in behaviour for AD/HD and controls, based on CPRS-93 subscales.

In an effort to understand factors that may have contributed to these present findings, the possibility that components of the experience not directly related to dolphins or VR affected changes in behaviour, must also be examined. Parents, as well as most of the children participating, were aware of the current study's aims, as well as the possibility that they may be invited to participate in future investigations involving interactions with live dolphins. Most participants appeared highly enthused by this prospect, and enquired about the possibilities. Parental responses to the questionnaires, or even the children's behaviour may have been inadvertently influenced through this motivating factor. Many parents invested a great deal of time and effort to attend the sessions. This is attested to by the large number of participants who travelled from regional Victoria to participate. The present findings may therefore need to be interpreted with the same amount of caution required in evaluating Nathanson's (1980, 1987) findings, in which most data was collected in a clinical setting, from individuals who volitionally presented at his facilities for treatment. The great financial and emotional expenditure invested by these families may of itself have provided an impetus to improve, or for parents to gauge their children's behaviour as better than before the intervention.

Obtaining baseline measures of behaviour based on a period during the school term and conducting the study during the holidays may account for some of the discrepancy in behaviour displayed before and after the VRDE in the present study. However it has been noted that children with AD/HD require structure and routine (Long, 1995) and may exhibit more disruptive behaviour when their usual schedule is disturbed such as during school holidays. This factor may therefore have been unimportant in influencing the differences in behaviours observed. A factor that possibly influenced parental assessment may have been the two administrations of the CPRS-93 in close succession. Although parents were instructed to complete the first copy a week prior to the scheduled VRDE, which would have allowed a 9 day interval before the second copy was administered, several parents indicated that they had not been able to complete the first copy in the specified time. Practise effects and recollection of their previous responses may have influenced their evaluations on the second copy of the scale.

A criticism of the CPRS-93 which was used in the present study is that the scale gives the appearance of objective data by assigning numerical scores to judgements which actually reflect subjective impressions (Conners, 1989). It has been suggested that reliability can be increased if a teacher rated version of the CPRS-93 is also used. This was impractible in the current investigation which took place during school term holidays. However it is generally accepted that parents who tend to spend more time with the child on a daily basis, and may have greater knowledge of the situational context of the child's behaviour, are reliable assessors of their children's behaviour (Conners, 1989).

Using measures that may be more objective than a parent rated questionnaire, such as neuropsychological testing may be a useful adjunct in future investigations. Neuropsychological tests, assessing functions and abilities thought to be involved in processes of attention such as the Neurospychological Assessment of Children (NEPSY) (Korkman, 1988a in Korkman &Pesonen, 1994) may be applicable. However, with AD/HD participants, laboratory test and psychological assessments that are currently available must be cautiously interpreted. They are not regarded as accurate reflections of the state of the pathological condition (Catellanos, 1997). Diagnosis of these conditions is usually made exclusively by history (Catellanos, 1997). Laboratory tests of attention and impulsiveness may not be reflective because symptoms often disappear in the presence of unfamiliar people, in novel situations, when under scrutiny or performing an interesting task (Long,1995).

Although controversial in their use for identifying correlates of AD/HD symptomology, brain imaging and electroencephalograph (EEG) techniques could be considered, to circumvent the aforementioned difficulties, if the appropriate evaluative niches became apparent (Crawford & Barabasz, 1996) in future studies.

Less subjective physiological measures may also be a more accurate measure of effects on control subjects, since a questionnaire based on pathology may have been less sensitive to changes experienced by healthy individuals, something that several parents of control subjects reported to the author.

In the present study, the novelty of the situation and the attention received by participants may also have affected behaviour. Controlling for novelty by conducting several sessions with the same subjects may indicate whether this was a factor. Investigating whether stimuli other than dolphins, including other animals, presented in a VR medium have comparative effects, would also be useful.

The inclusion of a control group is an initial effort at redressing the all- clinical populations usually assessed in dolphin encounter research. These prior studies did not address whether the reported improvements in learning are restricted to children with a psychopathology, or could potentially be experienced by healthy children encountering dolphins. Although co-morbid conditions, and families with several children affected by the disorder are regarded as common in the AD/HD population (Catellanos, 1997), future studies may wish to restrict participants to either ADHD or ADD, and to one participant per family. This may assist in isolating confounding variables on any possible changes in behaviour. Further studies could also investigate the effects of VRDE on adults.

The more dramatic change for the AD/HD group could be attributed to differences in how the two groups actually experienced the virtual environment, or to the impact that viewing dolphins or nature footage has on healthy individuals, and those with an attentional deficit.

please click to return to the table of contents