Abstract
This paper presents the development and evaluation of PICTOAPRENDE, which is an interactive software designed to improve oral communication. Additionally, it contributes to the development of children and youth who are diagnosed with autism spectrum disorder (ASD) in Ecuador. To fulfill this purpose initially analyzes the intervention area where the general characteristics of people with ASD and their status in Ecuador is described. Statistical techniques used for this evaluation constitutes the basis of this study. A section that presents the development of research-based cognitive and social parameters of the area of intervention is also shown. Finally, the algorithms to obtain the measurements and experimental results along with the analysis of them are presented.
Abstract in Spanish:
En este trabajo se presenta el desarrollo y evaluación de Pictoaprende un software interactivo diseñado para mejorar la comunicación oral y contribuir al desarrollo personal de niños y jóvenes con diagnóstico de trastorno del espectro autista (TEA) en Ecuador. Para cumplir con este propósito inicialmente se analiza el área de intervención que describe las características generales de las personas con TEA y su situación en el Ecuador. Las técnicas estadísticas utilizadas para esta evaluación constituye la base de este estudio. También se muestra una sección que presenta el desarrollo de parámetros cognitivos y sociales basadas en la investigación de la zona de intervención. Finalmente, se presentan los algoritmos para obtener las medidas y los resultados experimentales, junto con el análisis de los mismos.
The Autism Spectrum Disorder(ASD), present in childhood, is identified due to a person’s difficulties, changes or delays to develop social relations and communication. It also affects his/her behavior, which ends up attempting against his/her independence and ability to get into social relationships. [1]
Verbal communication is very low or almost null on these people. They present lack of attention and interest in daily activities, thus this situation becomes a great challenge for the common learning process.[2]
The most common communication methods for people diagnosed with ASD have been based on the use of pictograms - graphic representations of words or actions to learn interactively, formulate sentences or perform various routines. Daily activities are described by sequential graphics to be accomplished by these people. For example, activities such as brushing their teeth or taking a shower were proposed to be performed in their cleaning places so that the activities can be imitated.
Nowadays, mobile devices like smartphones and tablets are widely used for teaching purposes because children and adolescents diagnosed with ASD present a great deal of interest on them. These technologies allow to develop therapeutic games to stimulate and help the educational cognitive development in an enjoyable manner.[3] They also permit to assess the learning process by using statistical methods such as Bootstrap, hence realizing the Android application PICTOAPRENDE. This is an interactive software for children and young people who were diagnosed with ASD. The cognitive process is improved by using this method as well as it helps to ensure the independence of the users.
Research for creating PICTOAPRENDE and the analysis of results was performed during a period of time of one year and six months. A sample of 20 children between 10 and 17 years old was identified for this study and they belong to the Foundation that was taken for the analysis.
This application has incorporated pictograms by taking advantage of the wide spread use of electronic communication media through a series of messages and sounds that ease the daily interaction with this segment of population, which has been historically invisible.
The development of this application is very important due to the lack of such tools that contain requirements and parameters within the Ecuadorian region such as lexicon, the use of images known by them, valid emergency numbers, data base containing telephone numbers. [4]
The use of the Ecuadorian dialect within the application is very important to consider since the utilization of countless words and phrases typical of the region are present. Foreign dialects bewilders and fool the learning process for people that suffer this kind of disorder. [5]
The results obtained by the use of this tool were analyzed through the application of statistical techniques such as bootstrap, hypothesis validation, confidence interval and descriptive statistics, obtaining satisfactory results.
Within the characteristics of autistic children, it is not odd that it comes accompanied with a certain mental retard degree, which consists of the existence of a great difficulty towards oral communication when expressing their feelings and needs.
Children and young people with autism have an inflexible behavior. In other words they do not take the changes in routine, sudden facts or last time modifications adequately. [6]
Another feature of these children is the fact that they are obsessed with a certain behavior or one of their objects (toys). This added up to their lack of capacity of communication could result in a really difficult behavior and even somewhat aggressive when integrating the toddler with the rest of society. [7]
Eye contact and attention are deficient. They are not often able to use gestures as a manner of communication. They speak aloud or like automatons or robots. Commonly, they do not pay attention when other people speak, they are not receptive and cannot even respond to their own names.
It is possible that they spend too much time sorting things before they can pay attention or they have to say the same phrase over and over again to calm down. It seems somehow that they are in their “own world”. [8]
Particularly in Ecuador, there are no formal statistics of people diagnosed with autism and neither there are devices or applications to treat this problem.
The status of ASD in the country regarding the attention to autism can be summarized as follows:
It is the most versatile and well-known sampling technique. The basic idea is to treat samples as they were population and extract with replacement a large number of resamples of size n. Thus, although each resample has the same number of elements as the original sample, each one could include some of the original data more than once. As a result each resample will probably be different from the original sample; whereby a statistical , calculated from one of those resamples will take a different value from the one that produces another resample. The fundamental affirmation of bootstrap is that a frequency distribution of statistics calculated from the resamples is an estimate of the sampling distribution of population.[11]
Three practical applications are derived as a result of this process:
PICTOAPRENDE is an interactive software oriented to children and young people with autism spectrum disorder. It aims to improve the verbal process and interaction with the environment. [1]
This study was performed in Quito - Ecuador thanks to the proposed requirements by a number of foundations dedicated to the treatment of this disorder as follows:
Due to the delay or partial lack of oral language that children with autism possess, the need to compensate this deficiency is generated through alternative modes of communication such as pictograms - systems that allow represent a real object or figure schematically, thus developing their social and communication skills.
PICTOAPRENDE has seven options available, as shown in Figure 1. These help to learn daily sequences, common and basic phrases that express basic needs. It also teaches digits, sounds that are associated with different pets, and the correct pronunciation of vowels.
Through logic and memory game, it teaches emergency numbers and allows users to send e-mails and text messages expressing needs and possibly emergency.
This application incorporates lexicon and Ecuadorian dialect deployed into an amicable voice for users as well as the integration of images and clear symbols, thus making it attractive for children and young people with Autism Spectrum Disorder.[1]
PICTOAPRENDE was tested on the target population for six months through monthly evaluation on each option of the application.
A scale of skills assessment was utilized; this consisted of the tabulation of 10 tries in each option. The assessment was applied before and after using the tool so that it can be seen how the aptitudes of the population sample changed.
Finally, the results were analyzed by using the following algorithms:
The number of samples in the research is limited due to the size of population, thus data could be compromised. For this reason bootstrap technique was implemented to assess the difference between the initial vs final learning data; this according to the following algorithm. [14]
The obtained distribution constitutes a Bootstrap estimation of the sample distribution for the media difference statistics.
If we count the number of times that the difference in Medias exceeds or equals the observed value in the original samples, we can obtain the relative frequency; this will be considered as an approximation of the probability to find, accepting the null hypothesis as true, a difference of Medias more or equal to the one observed. See equation 1:
| (1) |
If the result of the probability is less than 0,05 significance level, we reject the null hypothesis and it is asserted that the obtained data with the sample population is 95%.
]]>Confidence intervals were implemented by the method of percentiles, which determines the more likely range of learning with the use of PICTOAPRENDE. Considering that the values can be represented visually as a normal distribution or commonly called Gaussian distribution, it is determined that according to the media value, the learning percentage is estimated.
To develop the Bootstrap confidence intervals, it was used the percentiles method where approximates to . The idea is that an interval with a confidence level of includes all the values of between percentiles and of the distribution of . [15]
To develop the confidence intervals we will work separately with the initial and final data, obtaining thus independent results of CI for the learning media of each option of the application. The following algorithm was set as pattern:
The following section describes the findings along with the technical differences. All data went through hypothesis testing for validation purposes; a value of 0,05 was yielded for all options of PICTOAPRENDE, thus claiming 95% of certainty on the data. Furthermore, this allows the realization of confidence intervals to determine the most probable learning range.
Next, the results are displayed according to the classical statistics by using pie charts, which represent the final percentage of learning. Then this is compared with the results obtained by using bootstrap technique, which is based on confidence intervals. This was done for every single option of the application.
During the period of evaluation, visual and aural memory were improved. It has to be consider that several of this actions were realized by users with almost no difficulty, thus getting better interaction in his/her daily life.
The percentages achieved by users at the end of the learning process on each activity is presented by measuring the aptitudes on the skills performance scale set before.
Figure 13 shows the confidence interval of the population sampling. It ranges from 23.5 to 30.5 percent; this shows a deficient knowledge before the use of PICTOAPRENDE because the starting point corresponds to 27% of the knowledge, which was acquired by using traditional methods.
Figure 14 presents the learning evolution within the population, since the media has a value of 69%. Additionally, this percentage determines an optimal learning average, which is taken with this option and considering that the learning percentage for hygiene sequences was raised by 42%. The confidence interval was obtained as 62-75 percent. This range guarantees 95% of certainty for the learning process of the sequences. A slow improvement was reflected, consequently the target was not achieved with some of the users, and this could be visualized through non-verbal communication, eye contact, facial expression, and social interaction regulatory gestures.[16]
]]>
It is an option that provides users with phrases that are used daily such as I want to go to the bathroom, I am hungry, I am thirsty and others. This options helps to encourage people’s attention and concentration as well as their bond with the surrounding environment. Vocalization of words through repletion is also stimulated.
Figure 15 shows that 70% of beneficiaries got to pronounce phrases formed by two or three words. This represents a big step ahead due to the users of this application were monosyllables at the beginning and in some cases they were non-verbal.
The use of bootstrap initially yields a confidence interval of 3-19.25 percent; this shows a low concatenation of words to form phrases as it can be seen in Figure 16, where the corresponding media is 10.5 %.
Figure 17 shows an average of 53% and a confidence interval of 40.5 - 64 percent; range in which we can state a significant advance in oral communication of children and young people with ASD. The obtained results were very close compared to the ones gotten with descriptive statistics.
Memorization and repetition exercises were realized by the use of this option. Thus, getting the right pronunciation and identification of the natural numbers, which shows the numbers from 0 to 9 along with their pronunciation in Spanish.
A high percentage was gotten by using pictonumbers, as shown in Figure 18. This was achieved due to the majority of children and young people already had a basic knowledge about digits, thus getting reinforcement and sound knowledge on this topic.
Utilizing bootstrap allowed to determine that people already had a prior knowledge about this actions, hence obtaining a media of 45% and a confidence interval of 36-54 percent. This shows an acceptable knowledge of the numbers from 1 to 9 before running the PICTOAPRENDE study. See Figure 19.
Using the final data, a media of 75% was obtained. This shows that the population under study possesses a good knowledge about this topic. The confidence interval ranges from 64 to 84 percent; this range guarantee the learning process with a certainty of 95%. See Figure 20.
By the deployment of didactic methods for learning embedded in the PICTOAPRENDE application, the five Spanish vowels were recognized in the 90% of users as shown in Figure 21.
]]> Because of the previous job realized by the foundation when learning vowels, the obtained results regarding reinforcement and learning are considered optimum. Once the learning process was finished, this option was mastered and it is confirmed by the utilization of bootstrap from where the confidence interval, 65 - 86.5 percent, is expected to receive the 95% of the population under study. The obtained media is 77%. See Figure 23
PictoAprende was used to perform memorization and repetition exercises so that children can get trained with the available emergency numbers in Ecuador, this in the case a real event may happen.
The application includes Police Station, Firefighters Department, Red Cross and emergency numbers. It also teaches in an interactive and didactic way, thus getting a permanent attention during the learning process. As result 8 out of 20 users recognize and are able to use the application in the case an emergency arises. See Figure 24.
From a sample of 8 children, who learned to manage correctly the emergency numbers application, a gradual progress is showed. It has to be considered that a familiarization stage was done at the beginning as well.
During the second stage, one of the emergency numbers which corresponds to 911 was memorized.
During the last two stages, 100% of the proposed numbers was learned. See Figure 25.
This process was carried out over a long period of time and a low overall rate in user performance.
The memory is not affected by these disorders although this accompanied by a mental retard, it edges the use of the semantic memory, which is in charge of coding and storing general and specific information in a very structured way. [17]
In order to establish a 95% certainty for the results, confidence intervals were obtained at the beginning as well as the end of the use of this option, thus determining a 9% media before the use of PICTOAPRENDE. See Figure 26.
By using PictoMessages, users were able to interact with their surrounding environment and showing their needs and emergencies by sending text messages and emails.
As it can be seen in Figure 28, 50% of the sample was able to understand the functioning of this option and the remaining 50% did not achieve the desired results because they did not realized that at selecting a different option, a message is sent to family’s pre-established phone number.
The study showed null knowledge before the use of this application. This due to this option is uniquely presented in PICTOAPRENDE. An initial media of 1.5% was determined by using bootstrap technique. See figure 29.
Once completed the learning period, a average of 43% was obtained, thus determining a confidence interval of 30.5% to 54.5%. This range guarantees the learning with 95% certainty. See Figure 30.
Because of the percentages obtained by the two techniques, a progressive advancement and a good reaction towards this option were demonstrated. Therefore, it is recommended to perform studies and test for longer periods of time.
As it was determined in the former paragraph, the learning results, obtained with the PICTOAPRENDE tool, were satisfactory. This considering that the data was validated through statistical techniques such as descriptive statistics and bootstrap. The first one tries to summarize the data in a quantitative manner, thus getting learning percentages after the study of the tabulated data once the tool was utilized. The second one is applied to support the learning affirmations as well as make decisions and come up with conclusions from the collected data of the population. It has to be considered that the earning process of the population with ASD is closer to reality, since a great number of random samples is generated, hence emulating a wider population sample.
Table 1 shows the obtained results after processing the information. For instance, considering the PictoActions option and descriptive statistics, it can be observed that 14 out of 20 children learnt this option. Additionally, through the use of confidence intervals, it can be assured that the learning process of the users would be placed in a range between 62 and 72% with a certainty of 95%. Based on this, it can be reflected that between 2 and 3 out of 4 actions were mastered by using this option and this after the continuous use of the tool for at least six months.
Grounded on the yielded results with PICTOAPRENDE, it can be said that the application is versatile and usefulness for people diagnosed with ASD; this considering that in modern times people are opting for “mobile learning”. Furthermore, it has to be mentioned that this tendency is modifying the educational habits. Technological breakthroughs during the last decades are making possible the development of studying methods that are more dynamic, complementary, and interactive; they are usually wireless, therefore catching the attention of children and people with ASD. [18]
A drawback of this tools is the fact that it cannot be assured that the user is able to alert when an emergency situation may arise. This due to the low learning indexes shown with the PictoMessages and Emergency numbers options. Also, users tend to develop a tendency to depend on the device, consequently they end up in a lack of concentration and attention of their surrounding environment. This fool the objective of the tool, which is the insertion of the user into society. As a contrast this would foster the lack of capacity to communicate and develop social interaction skills. This application shall be used under the supervision and responsibility of adults.
PICTOAPRENDE is an interactive application which is oriented to children and young people diagnosed with moderate ASD in Ecuador. The application counts with a series of options that aims to help people during the learning process. This helps to people’s reinsertion into society by improving their communication skills. PICTOAPRENDE can be run in electronic devices such as tablets and cell phones, since technology arises interest on children and young people with ASD, this has allowed to obtain satisfactory results. Therefore, it can be also concluded that specialized education since a very early stage in life helps to achieve significant indexes of independence on people diagnosed with ASD. After the study, it can be asserted with a 95% certainty that users are able to achieve high memorizing rates. However, although good results are obtained after the evaluation, it is recommended to realize more trials and studies for longer periods of time.
[2] M. Bernardos and A. R. Gómez, Autismo, 2nd ed. España: Alianza Editorial, 2004.
[3] L. M. Josefina, “Software para enseñar emociones al alumnado con trastorno del espectro autista,” Revista Comunicar, vol. 18, no. 36, pp. 139–148, 2011. [Online]. Available: http://www.redalyc.org/articulo.oa?id=15817007017
[4] D. Avilés, “Diseño de un centro recreacional para niños con autismo en el valle de cumbaya, quito,” Universidad Internacional SEK, 2014. [Online]. Available: http://repositorio.uisek.edu.ec/jspui/handle/123456789/910
[5] K. Tutivén, Propuesta operativa para la inclusión educativa de niños y niñas con necesidades educativas especiales derivadas de la discapacidad intelectual de 3 a 5 años de edad en los centros de desarrollo infantil municipales de la ciudad de Cuenca, Cuenca, 2012. [Online]. Available: http://dspace.uazuay.edu.ec/handle/datos/3040
[6] G. Herrera, X. Casas, J. Sevilla, L. Rosa, C. Pardo, J. Plaza, and R. S. G. Jordan, “Pictogram room : Aplicación de tecnologías de interacción natural para el desarrollo del niño con autismo,” Anuario de Psicología Clínica y de la Salud, vol. 08, pp. 41–46, 2011. [Online]. Available: https://dialnet.unirioja.es/servlet/articulo?codigo=4620745
[7] A. Riviere, “Modificación de conducta en el autismo infantil,” Revista Española de Pedagogía, vol. 42, no. 164, pp. 283–316, 1984. [Online]. Available: http://www.jstor.org/stable/23764136
[8] A. Díez-Cuervo, J. Muñoz-Yunta, J. Fuentes-Biggi, R. Canal-Bedia, M. Idiazábal-Aletxa, M. Ferrari-Arroyo, F. Mulas, J. Tamarit, J. Valdizán, A. Hervás-Zúñiga et al., “Guía de buena práctica para el diagnóstico de los trastornos del espectro autistal,” Rev Neurol, vol. 41, no. 5, pp. 299–310, 2005.
[9] S. F. Fernández, J. M. C. Sánchez, A. Córdoba, J. M. Cordero, and A. C. Largo, Estadística descriptiva. Madrid: ESIC Editorial, 2002.
[10] L. Margulis, “Funcionamiento de los sistemas de memoria en niños con trastorno autista y trastorno de asperger,” Revista Argentina de Neuropsicología, vol. 13, pp. 29–48, 2009. [Online]. Available: http://revneuropsi.com.ar/pdf/numero13/Margulis.pdf
[11] C. Z. Mooney, R. D. Duval, and R. Duval, Bootstrapping: A nonparametric approach to statistical inference. Sage, 1993, no. 94-95.
[12] A. Rodríguez-Barrionuevo and M. Rodríguez-Vives, “Diagnóstico clínico del autismo,” Revista de Neurología, vol. 34, no. 1, pp. 72–77, 2002. [Online]. Available: http://www.neurologia.com/pdf/web/34s1/ms10072.pdf
[13] J. Ortega, Hasta 30 días de cárcel por llamadas falsas al ECU 911, Ecuador, 2014. [Online]. Available: http://www.elcomercio.com/actualidad/carcel-llamadas-ecu911-codigo-penal.html
[14] A. M. Moles and T. D. L. C. S. Ayçaguer, El Método de Remuestreo y su Aplicación en la Investigación Biomédica, Habana, 2003. [Online]. Available: http://www.sld.cu/galerias/pdf/sitios/revsalud/tesis_de_resampling.pdf
[15] J. Gil-Flores, “Aplicación del método bootstrap al contraste de hipótesis en la investigación educativa,” Revista de educación, no. 336, pp. 251–265, 2005. [Online]. Available: https://dialnet.unirioja.es/servlet/articulo?codigo=1195518
[16] A. NAVARRO, “Nuevas tecnologías aplicadas a la discapacidad,” IMSERSO, Madrid, pp. 391–397, 1995.
[17] L. Pérez, “Aplicaciones informáticas para personas con trastornos del espectro autista,” Congreso Regional Las Necesidades Educativas Especiales: Situación actual y retos de futuro, pp. 391–397, 2013. [Online]. Available: https://www.researchgate.net/publication/39208688_Aplicaciones_informaticas_para_personas_con_trastornos_del_espectro_autista
]]> [18] J. Vila Rosas, “El teléfono móvil como herramienta educativa: el m-learning.. repaso a iniciativas y proyectos de investigación para conocer las posibilidades que ofrece el uso en los sistemas educativos de pequeños dispositivos móviles.” Primeras Noticias: Comunicación y Pedagogía, no. 226, pp. 52–55, 2008.