4. THE PILOT TEST
4.4 The Software assessment
4.4.2 An Objective Evaluation
Table 4.5 displays some of the information about the ABA strategies that the ABCD SW provided as data storage and analytics that enable accurate tracking and precise monitoring of the learning path. The results are presented as medium values of the participants’ learning path.
Table 4.5: Medium value of learning path indicators in matching programs for the
entire user group
Image Image Receptive Image
Average SD Average SD
Sessions 97 60.43 7 3.1
Introduced articles 316 82.6 43 14
Mastered articles 133 89.7 8 2
As shown in Table 4.5, in order to evaluate the entire child sample we focused on two of the six programs implemented by ABCD SW: matching program (Image/Image) and receptive program. Although all the children did the same training in terms of hours spent with the SW, each of them worked differently contributing to this analysis with a non-homogeneous number of observations. For example, some children need to work in many sessions of short duration, with an intermediate pause, other children work better without interruptions. During the pilot test only two programs were completed at 80% by all the children; the other programs were started in some cases or not explored at all in other cases. The matching program and specifically Image/Image is introduced to all the children at the beginning of the intervention because it is the simplest and requires only basic expertise. If the child is unable to work with words (due to age or ability) all matching programs involving text are temporally bypassed and the student moves to the Receptive program based on images. Thus the matching image/image and the receptive with images programs are used the most. With these first programs, the tutor was able to verify that:
Each child had become familiar with the tablet and with the new work environment; familiarization is more successful if done starting from the basic program, more immediate and intuitive.
The child has learned the concepts proposed to him with the traditional ABA mode and he/she is able to generalize them via the abstraction introduced by the software. With respect to this use of the software as a means of generalization, we verify a greater number of trials made on the upper levels of a single program; it means that for many items we started directly with the mass trial with one non-neutral distracter. Only if an item was introduced for the first time (or not actually learned), did the tutor propose the full work flow starting from the easier levels of difficulty: Mass Trial, Mass Trial with one neutral distracter, etc.
In about 6 months (January-June 2012), each child did an average of 97 sessions. The high SD (Standard Deviation) indicate that children did not work uniformly, as previously mentioned. However, the number of sessions per se does not help us if we do not connect them with the results in terms of concepts introduced and mastered. Each child had 316 items introduced, about 3 items per session, considering that the child works with one article at a time and moves to the next one only if he performed all the levels for that article. About one-third of the articles proposed were mastered. Both the SD values of the proposed articles and the articles mastered are high for the usual reason that children worked in different ways (number of articles introduced) and achieved non-homogeneous results (number of articles mastered) due to the different children’s profiles with respect to the syndrome.
The average number of categories introduced (13 for image/image and 2 for receptive image with SD of 0.9 and 0.5 respectively) is homogeneous between the children; it means that all children worked more or less with the same number of categories. Comparing data regarding sessions done and articles introduced, it would seem that those who worked more (greater number of sessions and articles
introduced) did so within the same category. This is possible because of some large categories such as Labels (including a large number of items not otherwise classifiable).
Considering the other matching programs involving words, carried out by only two children, as an example, Table 4.6 shows the results obtained by one of them.
Table 4.6: Number of sessions, introduced and mastered articles and introduced
categories in the other Matching programs of one child (not included in Table 4.5).
Word Word Word Image Image Word
(n) (n) (n)
Sessions 66 18 9
Introduced articles 151 115 48
Mastered articles 95 28 11
Introduced categories 8 5 3
Overall VABS-II data were aggregated and analyzed. Statistics for changes from T0–T1 on measures showed significant differences (P < 0.05) on Communication dimension (from a scoring of 101.0 SD = 50.4 to a score of 118.5 SD = 55.6) and in the Expressive subscale (from a value of 57.7 SD = 33.6 to 69.3 SD = 37.7) (Table 4.7).
Table 4.7: Pre-Post intervention Vineland Adaptive Behavior Scale/Subscale mean
scores
4.4.3 Discussion
For each child the mastery time was shortened using the technology-enhanced intervention compared to the classic one, also since the setting and evaluation times are notably reduced with automatic procedures [4]. For the same reasons, children had more learning opportunities in terms of categories and articles introduced during computer-enhanced sessions, and this greater effort required was well-accepted.
Concerning statistic indexes, our study highlights that participants achieved improved communication skills in general, and specifically in expressive communication (for the two children who have reached this program in the test period). This is in line with previous findings. A meta-analysis of scientific literature (until 2007) conducted by Ospina et al. provides some evidence of improvement in adaptive behavior, communication, interaction, comprehensive language, daily living skills, expressive language, overall intellectual functioning and socialization when delivering behavioral treatments [70]. This benefit increases with increased intensity of the intervention [80]. However, authors question the validity of these findings since there are few studies, with few participants over a short time period.