CONTACTO: A Social Networking Site for Supported-Employment Interventions for People With Intellectual and Developmental Disabilities

The main objective of this work is to determine whether social networks can be used to improve self-management skills in people with Intellectual or Developmental Disability (IDD) given the high rate of acceptance of such tools among them. The proposed methodology applies an employment-based intervention to evaluate whether social network services help improve self-management skills. A fit-for-purpose social networking service is built, and different evaluation mechanisms are applied to determine the degree to which users with IDD could use the tool without assistance. The working hypothesis is that in order to ensure the usability of the implemented tool, users (with and without IDD) should be involved throughout the whole software development cycle, following a co-design methodological approach. Results demonstrate that this methodological approach leads to high satisfaction and motivation rates. In terms of effectiveness and efficiency, results demonstrate that the tool is useful and usable for people with IDD. It can therefore be concluded that social network services built having users with IDD at its center will help improve their self-management skills. The main implication of this research is that when using social network services as a tool to support interventions, special attention should be paid to previously evaluating the usability and motivation rate as a measure of how autonomously the person with IDD can use the tool. Nevertheless, further research is needed to determine the impact that such tools can have on other types of intervention.

form. The new communication possibilities brought about by 22 The associate editor coordinating the review of this manuscript and approving it for publication was Orazio Gambino . the development of the Internet and online social networks 23 have therefore radically changed the way social interactions 24 and markets were conceived [1]. According to Eurostat [2], 25 57% of people in the EU in the period of time 26 2013-2020 participate in social networks. In people aged 27 16-24 the percentage reaches an 88%. It is uncertain, how-28 ever, how extended it is the use of internet and social media 29 among people, of any age, with Intellectual and Development 30 tacts, number of recommendations, or the profile photograph, 87 are used to assess the career benefits derived from the use 88 of such tools [25]. Nonetheless, despite the benefits that 89 social networking sites like LinkedIn has on employment 90 and career promotion and the importance that integrated-91 employment interventions has on people with IDD [26], lit-92 tle is known about how to extrapolate that to the field of 93 employment-based interventions for people with IDD. 94 One of the skills pursued by the different interventions 95 addressed to people with IDD is self-management. Accord-96 ing to Browder and Shapiro [27] self-management refers to 97 the different practices that an individual employs to man-98 age his/her own behaviour (self-delivered prompting, self-99 monitoring, self-recording, and self-delivered consequences). 100 Self-management skills therefore lead to minimizing the need 101 for external support. In this sense, there is an increasing inter-102 est about how online social networks could be employed to 103 support interventions intended to promote self-management. 104 Nevertheless, little is known about this fact and researchers 105 like [28] call for further research on the use of social media 106 technologies as a powerful tool to assist individuals with 107 disabilities seeking employment. The main obstacle to carry 108 out such study is the lack of online social networks that 109 are made fit-for-purpose for people with IDD. In fact, the 110 workaround implemented by researchers like [28] is to resort 111 to employment specialists assisting individuals with IDD in 112 using existing networks (Facebook, Twitter, and LinkedIn). 113 Ideally, this type of studies should be carried out with tools 114 that were usable and useful for people with IDD (and people 115 without it) so that the tool does not interfere in the ultimate 116 goal of the study, as it is to determine whether online social 117 networks can be used to improve self-management skills in 118 people with IDD. 119 The objective of this study is therefore to provide a tool for 120 supporting employment-based interventions that it is not only 121 usable but also useful for both people with or without IDD, 122 to determine whether social networks can be used to improve 123 self-management skills. 124 This study is motivated by the interest that online 125 social networks stir among specialists in the context of 126 employment-based interventions in people with IDD. Due to 127 the lack of fit-for-purpose online social networks for people 128 with IDD and the identified limitations of existing tools, the 129 authors of this papers understand it is necessary to implement 130 a tool that minimizes the need for external assistance. In this 131 sense, to ensure the usability of the implemented tool, users 132 (with and without IDD) should be involved throughout the 133 whole software development cycle, following a co-design 134 methodological approach. The significance of this study lies 135 in the knowledge it provides regarding the role online social 136 networks can play in promoting self-management skills in 137 people with IDD. This knowledge will enable more and 138 better-informed interventions which eventually will improve 139 the quality of life of people with IDD. 178 The use of online social media or social networks is con-179 sidered to be very beneficial for individuals with IDD as 180 it contributes to their development of social identity and 181 self-esteem as well as for enjoyment [4] [31]. This lead to the digital exclusion of people 194 with IDD what limit the benefits they could potentially 195 receive from the use of digital technologies [32]. For this 196 reason, efforts like the one presented in this work, are still 197 needed to achieve the online social inclusion stated by [33] 198 mainly ensuring secure spaces and usable technologies.

199
One of the reasons that have been pointed out behind this 200 reduced use of social network services is the fear to the 201 inherent risk related to the cyberspace [34], particularly for 202 vulnerable groups. The study in [31] analyses the safety per-203 ception of young people with learning disabilities when using 204 online services. Although this study concludes that young 205 individuals with IDD feel confident about their knowledge of 206 online safety their parents did have serious concerns about 207 online risks. In fact, emotional distress has been reported 208 as the most common risks derived from the use of online 209 services in people with IDD, according to [35]. 210 It seems obvious that if online social inclusion is sought, 211 stringent mechanisms should be in place to guarantee safety 212 and ethics. In this sense, there are already efforts on this line, 213 such as the work in [33] that proposes a framework to design 214 online social media platforms for individuals with IDD. How-215 ever, it is currently unknown what mechanisms should be 216 implemented in existing platforms so that safety for individ-217 uals with IDD could be ensured although current research, 218 such as the works of [32] and [36], works on destigmatising 219 the use of internet and online social network in people with 220 IDD. Another approach, as the one presented in [28] advo-221 cates for the use of existing social media platforms such as 222 LinkedIn, Facebook, and Twitter, with the employment spe-223 cialist providing assistance and supervision. This evidences 224 that either because of the risks or due to the complexity of 225 the interaction mechanisms of such platforms, most users 226 with IDD find extremely difficult to use these platforms on 227 their own.

228
Nonetheless, there is a lack of consensus, from the state 229 of the art, on whether the online social inclusion should 230 be carried out by building isolated online social networks 231 that ensure security and safety, as stated in [33], or whether 232 to resort to well-known platforms, along with employment 233 specialists, as argued by [28]. The work presented here under-234 stands that inclusion must be adopted from the early stages 235 of the design process, implementing an inclusion-by-design 236 approach. For this reason, this work has opted for building 237 a new social media platform fit-for-purpose. This tool is 238 intended to leverage the increase in self-esteem and auton-239 omy in the process of an employment-based intervention, 240 seeking to minimize the need of support from the employ-241 ment specialist. By doing so, CONTACTO contributes to the 242 online social inclusion of people with IDD by providing an 243 online social networking site that it is usable and secure. 244 In order to achieve this goal, the tool is co-designed with the 245 main stakeholders (people with IDD, employment specialist, 246 or associations). Furthermore, previous works have already 247 identified the main risks and difficulties found by people with 248 IDD when using technologies [10], [31], [32], [36]. 249 Involving users with IDD in a software co-design process 250 is something relatively novel although some previous works 251 have already explored its use, as in [29], [37], and [38]. 252 Co-design ensures, as stated in [29], that requirements and 253 VOLUME 10, 2022 is devoted to the system and user requirement specification.

290
The second stage undertakes the design and implementation 291 of the system. During the third the system is evaluated. The 292 methodology proposed here differs from the work in [45] 293 and [42] in that the principles of agile methodologies are  recently been gaining attention, in particular for people with 308 severe mental illness [47]. Despite the need for tools support-309 ing employment-based interventions for people with disabil-310 ities, as stated in [48], the scarce literature found to the date 311 suggest that no major contributions have been made in this 312 field. The use of social media has proven to be a powerful tool 313 for such purpose [28], but the challenges found in existing 314 platforms (LinkedIn, Twitter, or Facebook) by people with 315 IDD limit their autonomy to use such platforms.

316
CONTACTO is proposed as a tool, which is specifically 317 devoted to SE for people with IDD, taking into consideration 318 the participation of the SE professionals and the organisations 319 that provide placements for employment and also training 320 courses. The system therefore supports three main roles, 321 namely: 322 1) Candidate: This is the role of an individual with IDD 323 who goes through a SE intervention. This intervention 324 might also involve training courses.

325
2) Reference role: This role is played by the SE profes-326 sional that assist the candidate during the intervention. 327 This person might also provide support on the place-328 ment where the SE is taking place.

329
3) Supervisor: This is a system administration role. This 330 role is supposed to be held by an Information Technol-331 ogy staff memeber.  Figure 1 outlines the different purposes for candidates to 336 use the platform. One of the main uses of CONTACTO is to 337 provide a candidate profile, as a Curriculum Vitae, in which 338 they can state their training and learning experiences. This 339 will help human resource services when identifying those 340 candidates that better fit the available positions. Furthermore, 341 there is an implicit benefit in encouraging candidates to keep 342 their profile updated. Autonomy is worked in this way, as the 343 tool has been designed following a simple interface and lan-344 guage, minimizing the need for support while including new 345 merits. Figure 2 shows the profile of a candidate.

346
CONTACTO has been designed as a meeting point for 347 candidates, seeking training or job opportunities, and organ-348 isations, offering both jobs and training. Organisations can 349 adopt an active role in the platform looking for candidates 350 whose profile fits their needs. In this sense, they can look for 351 specific profiles and contact the candidate or the reference 352 role that supervises that candidate. Organisations can also 353 adopt a more passive role, just posting an offer, and waiting 354 for candidates to show their interests, and then selecting 355 among those who have shown interest. The platform offers 356 candidates a functionality to subscribe through alerts sent 357 via email to offers that fit their profile, based on interests, 358 location, whether they imply a job offer or just training, etc. 359 Candidates, when interested in a specific offer, can notify the 360 organization about their interest. Furthermore, they can select 361 the offers they like, so that the system profiles the user based 362 on the offers that have been liked. Figure 3 shows the interface 363  offer that can be of relevance for any of the supervised 372 candidates. The offer can be directly notified to a candidate. 373 This functionality ensures that candidates have not missed 374 and opportunity while their autonomy is being encouraged. 375 Candidates are encouraged to notify their interest in an offer 376 in an autonomous way, requesting or receiving help only 377 when needed. 378 VOLUME 10, 2022  It has to be highlighted that in the proposed architec-423 ture, NGINX features (high performance and load balancing) 424 have been used to handle API traffic, thus functioning as an 425 API gateway.

426
The proposed platform can be accessed via a web browser 427 but also using the App available in the Android Play Store. with which users achieve specified goals'' or the number 441 of tasks that the users have been able to complete, taking 442 into account the errors they have made in order to reach 443 the specified goal.

444
• Efficiency defined as ''the resources expended in rela-445 tion to the accuracy and completeness with which users 446 achieve goals'' or the time or effort spent by the user to 447 complete the proposed task or objective.

448
• Satisfaction defined as ''freedom from discomfort, and 449 positive attitude to the use of product'' or the level of 450 personal comfort with the task performed.
The evaluation of effectiveness ( ) will be carried out through 453 observation and will be calculated by means of the completion 454 rate. It is also interesting to count the number of errors in the 455 execution of each task.
where N ts is the number of task completed successfully and 458 N tu is the total number of task undertaken. On the other hand, the evaluation of efficiency requires 461 measuring the time it took the user to complete the task. 462 In addition, it has to be determined whether the task has been 463 completed successfully or not. Efficiency can be calculated 464 either as efficiency based on time or as the overall relative 465 efficiency.

466
Time-Based Efficiency (η t ) is obtained as follow: where N is the total number of tasks (goals), R is the number 469 of users, n ij is the result of task i undertaken by user j (if the 470 user successfully completes task then n ij = 1, else n ij = 0), 471 t ij is the time spent by user j to complete task i (if the task is 472 not successfully complete, then the time is measured till the 473 moment the user quits the task).

474
The following formula is used to calculate the overall 475 relative efficiency (η r ):

478
Finally, the evaluation of satisfaction is done through two 479 questionnaires. One of the questionnaires will be used to perceived while carrying out the task. Therefore, the user 491 responded with an item on a likert scale from 1 (very easy) 492 to 7 (very difficult) to the only question in the questionnaire 493 which is ''In general, the task was. . . ''.

494
The SUS questionnaire was carried out when the user 495 finished the test, indicating the general impression they had 496 regarding the use of the tool. The user rates with an item on 497 a likert scale from 1 (strongly disagree) to 5 (strongly agree) 498 each of the 10 questions that are included in the questionnaire. 499

500
To meet the definition of the ISO 9241-11 standard, in addi-501 tion to evaluating usability through these three dimensions 502 VOLUME 10, 2022    performance with the aim of counting the tasks that were suc-524 cessfully completed and detecting and counting errors while 525   time. For some particular participants, the supervision of a 545 professional was required to reinforce the explanation of the 546 test.

547
All assessments were carried out in an office belonging 548 to the ASPRONA association, in order to avoid participants 549 having to travel. The room was equipped with a laptop for 550 the participants to perform the test and another laptop for the 551 evaluator to manage the whole process. The observer also had 552 a support laptop to take notes. After studying the usability of the tool, we evaluated the 556 motivation of users to use CONTACTO. This measure also 557 determines the system acceptance and willingness to use the 558 propose technology. To perform this evaluation, the Intrinsic 559 Motivation Inventory (IMI) questionnaire [49] is proposed, 560 with a selection of 26 questions adapted to this study.

565
According to Ryan [50] and Deci [51], the subscale that 566 best defines the measurement of intrinsic motivation is inter-567 est/enjoyment although perceived competence is considered 568 to be a closely related subscale. In addition, perceived choice 569 can provide us with a positive predictor of both self-reported 570 and behavioral measures of intrinsic motivation. On the 571 other hand, we have the pressure/tension subscale that is 572 handled as a negative predictor of intrinsic motivation. The 573 effort/importance subscale is a separate variable and will 574 therefore be used in some items if relevant. In some cases 575 it is also important to take into account whether the experi-576 ence is considered useful for the subjects, and for this, the 577 value/utility subscale is employed. Finally, the relatedness 578 subscale is used in studies that take into account interactions 579 between people.

580
The inclusion or exclusion of specific subscales does not 581 seem to have any impact on the others, and even the order 582 of the items seems to be insignificant according to the afore-583 mentioned studies [50], [51]. Another fundamental aspect in 584 the development of this questionnaire is redundancy, as the 585 items of the subscales overlap significantly, but this fact is 586 irrelevant to the participant due to the random way in which 587 the items are presented.

588
For our study we selected 26 items from the original 589 IMI questionnaire whose classification and categorisation of 590 items is based on the self-determination theory used in numer-591 ous studies. With the help of ASPRONA professionals, the 592 subscales listed in the following table were selected. For the 593 selection of the subscales and the respective items, the infor-594 mation to be obtained in the study was taken into account, 595 as well as the users to whom the questionnaire was addressed. 596 Thus, the aim was to obtain feedback on the interest generated 597 by the use of the application, the value or usefulness they 598 find in it or the importance of having such an application. 599 VOLUME 10, 2022 it and choose to use this application among others that exist 601 without feeling any kind of pressure.    According to the systematic review presented in [52], the 652 sample sizes of similar interventions greatly vary from one 653 study to another, from a minimum of 1 subject to a maximum 654 of 57. In fact, more than 50% of the surveyed studies have 655 less than five participants. Despite being limited in terms of 656 the potential generalization of the findings, the sample size 657 is similar those found in the state of the art for similar inter-658 ventions. Moreover, according to [53], in qualitative research, 659 data needs to be collected until saturation occurs. Saturation 660 occurs when there are no new issues raising after a certain 661 amount of ideas have already been stated. In this sense, when 662 there are no new data or ideas, there is no need to continue 663 broadening the sample size [54]. 664 Initially, the sample consisted of 32 volunteers. Within this 665 sample we differentiated 22 volunteers for the role of candi-666 date and 10 volunteers for the other roles and who participated 667 with a more expert opinion regarding the field in which 668 the CONTACT application will provide its service. Of those 669 22 participants in the candidate role, 3 were excluded because 670 they did not want to participate in the study, so our sample was 671 reduced to 19 in that role. Subsequently, 2 volunteers did not 672 meet the inclusion criteria, 4 were excluded from the test by 673 determination of their reference roles and 1 finally declined 674 to participate in the test, as Figure 10 depicts. Finally, 12 vol-675 unteers with the role of Candidate with different levels of 676 disability were used to carry out this study, of which 8 were 677 male and 4 were female with an average age of 25 years. As this application involves other users associated with this 679 process, it was necessary to collect information on the tasks 680 proposed for the other roles in order to obtain feedback from 681 the point of view of the professional in charge of candidates 682 (referent roles), from the main organisation that proposes 683 job offers and internships to candidates from the ASPRONA 684 association, and from the association's supervisors.

685
The tests were carried out by alternating roles in order to 686 avoid procedures or responses conditioned by the previous 687 user's performance. On the other hand, different tasks were 688 selected within the platform to find out, in a qualitative way, 689 what they thought of the proposed environment for the search 690 for internships and employments. 691 96468 VOLUME 10, 2022 The sample comprising the roles that form part of the 692 supervision of the job and internship search process was 693 initially composed of 10 volunteers, of which 3 were supervi-694 sors, 5 were reference roles and 2 were organisers. All of them 695 fulfilled the inclusion criteria, but finally, the sample was 696 reduced to 7 volunteers because they could not attend the day 697 of the tests, leaving the sample composed of 2 supervisors, 698 4 reference roles and 1 organiser, of which 5 were women 699 and 2 were men with an average age of 36 years old.

700
Each volunteer is identified with a test participant ID.

701
In Table 2 the IDs related to their role are detailed.  Table 2. Out 715 of these 4 candidates, only 1 had problems with more than one 716 task.

717
Furthermore, the analysis of the number of errors commit-718 ted by candidates in the tasks they were assigned, shows that, 719 in general, the average number of errors per task was low, 720 as Figure 11 illustrates. It was found that the tasks in which  research is also needed to determine the causes of failing to 726 complete the three aforementioned tasks. If the reason is due 727 to a poor design, appropriate actions will be carried out to 728 amend them. If the reason is due to the lack of familiarity 729 with this type of tasks, training actions can be organised to 730 reinforce those tasks they are less familiar with.

731
Regarding efficiency, refer to Table 6, it is observed how 732 there are three tasks from Table 1, namely: ''Reset Password'', 733 ''Create Training'' and ''Create Experience'', which show a 734 low time-based efficiency. This result suggests that candi-735 dates have to dedicate more time to find their way through 736 this task, as shown in the Figure 14. On the other hand, it is 737 also observed that the tasks with the lowest percentage of 738 overall relative efficiency are ''View My Favourites'', ''Reset 739 Password'' and ''Search for an Offer''. This implies that these 740 tasks were not even completed by some users.

741
In order to summarize the obtained results: Tasks A1, A4, 742 and A8 are those more prone to errors. It is necessary to 743 evaluate whether the reason of these failures is due to a poor 744 system interaction design or just because the complexity of 745 the task cannot be successfully addressed by most of the 746 people with IDD. Tasks A2, A3, and A1 are those that require 747 more time from the users. In this sense, further analysis 748 is required to determine whether these tasks could be split 749 or simplified in order to reduce the time spent completing 750 them. Finally, tasks A1, A5, and A8 are the ones with the 751 worse overall relative efficiency. Task A1 (reset password) 752 obtains bad results in the three metrics. It suggests the need 753 to reorganize that task in such a way that it is more feasible 754 by people with IDD to complete it. 755 Finally, as shown in figure 12 and in Table 4, users evaluate 756 all tasks performed above 4.83 in average. The tasks that 757 obtained the lowest scores were ''Reset Password'', ''Consult 758 your Reference role'' and ''View My Favorites''. It seems that 759 the same tasks that obtained low results in the previously 760 analyzed metrics are the same with which users are less 761     think that they are quite capable of using the tool, as ''Per-824 ceived Competence'' is the next best-rated subscale, and 825 even the average obtained in ''Value/Usefulness'' indicates 826 that they consider this application useful or see an added 827 value in using this digital solution in the development of 828 their activity. In addition, the next highest scoring subscale, 829 ''Interest/Enjoyment'', indicates that these volunteers have an 830 interest in or enjoy using the tool.

831
The obtained score in the Perceived choice indicates that 832 they would choose this application among others to carry out 833 the different tasks they need to do for the development of their 834 activity.

835
Finally, the Pressure/Tension subscale is the one that 836 obtained the lowest score, which means that the volunteers 837 who underwent the test did not experience moments of pres-838 sure or tension when performing the different tasks.

840
This study addresses the question of whether online social 841 networks can be used to improve self-management skills, 842 in the context of an employment-based intervention, for 843 people with IDD. The results show that, despite having con-844 sidered employment specialists providing assistance during 845 the intervention, as suggested by [28], most users with IDD 846 managed to complete all the proposed tasks. In fact, only 847 one user with IDD shows difficulties in performing more 848 than one task. In terms of motivation, the results obtained 849 in perceived competence suggest that users find themselves 850 capable of using the tool. In the overall, results demonstrate 851 that a fit-for-purpose online social network can be used to 852 support interventions addressed to improve self-management. 853 The results supporting this conclusion are discussed in this 854 section.

855
Similarly to the works in [42], [43], and [44] the suc-856 cess of the proposed technology is measured based on its 857 user acceptance and perceived usefulness. More specifically, 858 usability is measured in terms of effectiveness and efficiency. 859 VOLUME 10, 2022  use of this online social networks is greatly influence by the 887 experience users have on interacting with existing networks 888 like Facebook. Hitting the ''Add to my favourites'' button 889 follows the same approach as that of Facebook. In this sense, 890 eight out of the twelve individuals with IDD are males. This 891 could explain that they are less familiar with the hit ''Like'' 892 button and therefore they have more problems retrieving 893 his favourite offers. Special training on this aspect could be 894 planned to improve interactions with favourites offers. The 895 case of the low performance achieved by the ''Consult your 896 Reference role'' task is explained, after the qualitative analy-897 sis carried out during the interviews, due to a poor design that 898 requires improvement.

899
The results obtained from the motivation analysis are 900 in line with the previous studies that have assessed 901 the impact that online social networks have on career 902 promotion [23], [24]. In this sense, similarly to the results 903 obtained for other well-known online social networks, CON-904 TACTO users perceive it as a useful support for career pro-905 motion. More relevant are the obtained results with regards 906 to the perceived competence, as this is directly related to 907 self-management skills. In line with the conclusions of 908 the systematic review performed in [52], self-management 909 improves with targeted interventions. In this case, the per-910 ceived competence to use CONTACTO can be directly trans-911 lated into a positive impact on self-management skills. This 912 opens new and future lines of research to explore how these 913 skills could be improved, over time, while using the tool by 914 including new functionalities.

916
There is a lack of consensus on whether online social inclu-917 sion should be carried out by adapting existing tools (such as 918 LinkedIn, Facebook, or Twitter) or, on the contrary, by build-919 ing new platforms specifically dedicated to people with IDD. 920 Section II have identified works that support the use of 921 existing tools, with external support, such as the work in [28].

922
Other works, such as those of [32] or [36] promote the online 923 social inclusion through the use of existing tools.

924
The practical implications of using existing tools, such as 925 LinkedIn, rather than a fit-for-purpose one, such as CON- ventions. This is a very specific questions that yet had not 946 been addressed by previous works.

976
The objective of this study was to determine whether 977 online social networks could be employed for improving 978 self-management skills through supported-employment inter-979 ventions for people with IDD. Due to the barriers found by 980 people with IDD when using existing tools like LinkedIn, 981 and for the sake of ensuring their online social inclusion 982 from the beginning, a fit-for-purpose social networking site 983 was built and evaluated in terms of its usability and moti-984 vation to use it. Results demonstrate that self-management 985 is improved by means of the CONTACTO tool, based 986 on the obtained results regarding the usability, motivation 987 and, more specifically, the perceived competence to use the 988 tool.

989
This work brings about important practical, theoretical, and 990 societal implications. Regarding the practical implications, 991 the data collected from the interaction between users and 992 job offers generates new opportunities for the adoption of 993 data analytics, that can positively impact the decision-making 994 processes and the development of new policies regarding 995 the assistance provided to individuals with IDD. In terms 996 of the theoretical implications, this work has contributed to 997 fill the knowledge gap about how online social networks 998 could support interventions for people with IDD. Finally, 999 societal implications reveal the improvement in the quality 1000 of life of people with IDD as result of the improvement of 1001 their self-management skills.

1002
Further research is required to determine whether this 1003 type of platform, based on online social network approaches, 1004 can be useful for other types of interventions addressed to 1005 people with IDD. Among the limitation of this work, it has 1006 to be noted that the impact on self-management has not 1007 been quantitatively measured in relation with the use of the 1008 CONTACTO tool. Further research is required to determine if 1009 this impact can be quantitatively measured and to determine 1010 how these skills could be improved over time by means of 1011 new functionalities. The sample size can also be considered 1012 an additional limitation of the study in order to generalize the 1013 conclusions. Nevertheless, the results and findings provide 1014 new and valuable knowledge to be considered in conjunction 1015 with other similar studies. 1016 analysis and review of randomized controlled trials of individual placement 1182 and support,'' PLoS ONE, vol. 14, no. 2, Feb. 2019, Art. no. e0212208.