Using Accessibility Awareness Interventions to Improve Computing Education

Research demonstrates that we continue to be de-ficient in creating inclusive and equitable software, with a lack of empathy among software developers attributed to this issue, even if unintentional. Although increasing empathy has been shown to be part of the solution, unfortunately there is a lack of data that can be analyzed that shows how empathy-building and accessibility awareness interventions can be utilized in computing education, along with a shortage of free, public, and easily adoptable educational material regarding empathy and the creation of accessible software. Through large in-person studies that involve real-world par-ticipants, we are able to contribute to the knowledge regarding the potential benefits that can be used to improve computing education by implementing empathy-building strategies in the curriculum. Two forms of accessibility awareness exercises are utilized to display the effectiveness of empathy-based learning and teach the importance of accessibility in computing, including experiential and expression-based interventions. In these inter-ventions, we expect to find an increase in awareness and empathy in participants, as well as an increase in participant interest in the topic of software inclusivity and accessibility by placing the participant in a situation they may encounter in the real world. We additionally present a web-hosted, easily adoptable ex-periential and expression-based educational lab to support the inclusion of inclusive development and/or empathy-focused topics into a variety of curricula formats. Complete materials are available on the project website: https://all.rit.edu


I. INTRODUCTION
Creating inclusive software has fortunately become an area of increasing recognized need [13].Even with this recent focus, research demonstrates that we continue to be deficient in creating inclusive and equitable software [1,6].A lack of empathy among software developers has been attributed to the creation of biased, inequitable software [1,10].Research has shown that increasing empathy can lead to software that is developed more accessible, inclusive, and equitable [3,16,17,18].Experiential empathy-building interventions have been explored in various non-computing domains, such as in medicine [9,19], and for creating tolerance in social situations [4].Unfortunately, despite preliminary efforts in this area [17,18] the use of experiential empathy-building interventions have not been sufficiently examined in computing education.
Previous work has demonstrated the potential benefits of experiential interventions to create empathy in computing education [17,18].This work also hopes to provide additional information on the best practices and possible benefits of empathy-building interventions.By holding a large in-person activity involving real-world participants that attend a local institute-wide community event that includes a large percentage of participants that are in the field of software engineering and computing, as well as some who are outside of the field, we materialize our hypothesis and utilize two separate study groups to allow for the documentation of a more diverse range of participant demographics in support of the examined research questions.Both evaluations are to be conducted using a systematically designed pre/post survey summation assessment.This will be achieved using both experiential and expression-based empathy building interventions.
In summary, this work hopes to make the following contributions: • Research Findings: Suggest that both experiential and expression-based interventions are helpful in building empathy and there is no significant difference in creating empathy between both interventions.This statement will be verified by comparing both interventions in three aspects: interest, awareness, and empathy.• Educational Material: We present a hosted, easily adoptable experiential and expression-based educational lab to support the inclusion of inclusive development and / or empathy-focused topics in a variety of curricula formats.Complete materials are publicly available on the project website: https://all.rit.edu

II. CREATED PUBLICLY ACCESSIBLE EDUCATIONAL ACTIVITY
The developed empathy-building educational activities serve to both help address the examined research objectives (Section III), but also to serve as an empathy-building educational  intervention that has demonstrated its educational effectiveness and capabilities [17,18].These intervention formats and labs are publicly available for educational adoption, while also serving as the basis for our research evaluation and the preliminary results described in Section III.

A. Experience-Based Intervention Activity
Participants are provided with an opportunity to gain firsthand insight into the challenges faced by individuals with color-blindness when using poorly designed and inaccessible apps.For a demonstration activity, participants engage in a game that involves three colored circles (Figure 1a).Their task is to click the circle when it turns red, while refraining from clicking when it appears in the other two colors.For participants who can see the colors, this is a simple and easy task.For those who cannot discern colors (i.e., those who are colorblind), this game is challenging to impossible.Successful clicks are rewarded with points and the objective is to achieve the highest score possible within the 15-second time limit.By implementing this points-based challenge, the participant is incentivized to watch closely for the color changing circle and to create a baseline score to refer to when confronted with the second part of the experiential lab module.
Following their first iteration of the game, participants have another opportunity to play.However, this time, the user interface (UI) simulates the experience of colorblindness (Figure 1b).The once distinct circle colors are now transformed into nearly indistinguishable shades of gray, emulating what a colorblind user would encounter, thus presenting a significant challenge to achieving a high score.This exercise serves as a compelling demonstration of the crucial role played by accessible web design in ensuring inclusivity for those with colorblindness.

B. Expression-Based Intervention Activity
The initial stage of the expression-based empathy-building activity is the same as the experiential activity, where participants are presented with the same objective in playing the game that involves clicking on colored circles to achieve a high score, the user will be presented with a new element to deepen their emphatic understanding.The participants next watch a video clip featuring a color-blind person attempting to play the same game.The participant observes the frustration of the user and their attempt to play the game even though it is inaccessible to those who are colorblind (experience is demonstrated in Figure 1).
Figure 2 demonstrates the video of a frustrated colorblind user using the inaccessible application.By incorporating this expressive empathy-building component, the activity aims to foster a more empathetic appreciation for the experiences of individuals with color blindness.Participants gain valuable insights into the frustrations and difficulties faced by users with this condition when interacting with digital interfaces that lack appropriate accessibility features.This exercise encourages participants to reflect on the importance of designing inclusive interfaces that cater to the diverse needs of all users.

C. Educational Usage
Complete materials are hosted on our server and only require a browser for adoption.This enables institutions and individual learners to easily utilize the material for brief

D. Objectives of Utilized Lab Format
The provided educational lab has been systematically developed to achieve the following key goals: 1) The lab does not require any special hardware or software: Only a web browser is required to operate the lab, enabling both institutions and independent learners who lack the ability to install specific software or have older computer systems to conveniently use the labs.This will also promote classroom inclusivity by removing the need for instructors to spend time pre-configuring classroom computers 2) Instructors and students will only require very basic programming/computing skills to utilize the lab: It is crucial for individuals with varying degrees of software development expertise to understand the significance of developing software that is accessible.3) Labs should fit into already crowded foundational computing courses: The intervention is designed to take approximately 10-20.The succinctness of the lab will enable them to fit into courses that are already heavily time-constrained.
4) The lab should demonstrate the need to create accessible software: The main objective of the labs are to emphasize the significance of developing software that is accessible to students.The lab allows students to directly encounter the accessibility concern addressed (as realistically as feasible).

III. STUDY DESIGN
Our study included data collected from 121 real-world participants at a local community event.An approved IRB was obtained prior to beginning our study.Participant Recruitment: Our human study was conducted at the Rochester Institute of Technology during Imagine RIT 2023, a single-day event where thousands of people from the local community visit the RIT campus and view a variety of scientific and educational venues, including robotics, software projects, and engineering activities.Visitors to the institutewide event ranged in age 0-51+ and were generally representative of the general population of the non-technology-oriented local population.
With the help of several student workers, tables will be set up with laptops running the application variations described in Section II.Participants will be recruited by asking visitors passing by the tables if they would like to participate in the experiential or expressive activities of the intervention.Other than basic technical and process-related questions, participants will not receive guidance in experiencing the activity as an unbiased user.
Since our IRB only covered users of at least 18 years of age, we will not retain the results of anyone younger than this age that participates in our study.We will also exclude participants who did not complete the prepost survey completely (Section III).This resulted in a total of 121 people participating in our study.
Data Collected: Results were collected using a pre-post survey process outlined in Figure 3 and described in the following sections.

Experiential Application Usage
Post-Survey Fig. 3: Utilized data collection process for in-person user study

Pre-post Survey Data Collection Instruments:
Our user study comprised of three primary phases that are further described later in our work: 1) Pre-Survey: Collect participant demographic data and elicit response to empathy-building intervention related questions.2) Experiential/Expression-Based Exercise: Participants interact with the experiential or expression-based activity.3) Post-Survey: Participants answer the same empathybuilding intervention related questions.
Participants were asked to complete surveys both before and after the exercise, comprising the same questions as are shown below in both the pre-and post-surveys.Participants were not allowed to use the application until they had completed the initial pre-survey, and they completed the post-survey only after concluding using the application.Participants who did not complete both parts of the survey were not included in the study.This survey component measured the participant's feeling of regrading bias.Due to the brevity of the largescale in-person study, we kept the number of pre-post survey questions brief.The pre-post survey and high-level results are shown below: Pre-Survey Participant Background Questions

IV. PRELIMINARY RESULTS
In our analysis, we leverage two non-parametric tests: the Wilcoxon signed-rank test (which evaluates the null hypothesis that two samples originate from identical distributions) and the Kruskal-Wallis test (which examines the null hypothesis that the population median is the same for all the groups).We choose these tests because they don't rely on assumptions about the data distribution as parametric tests do, thus providing more robust testing results.In our analysis, we follow the convention by choosing 5% as our significant level.
Our evaluation focused primarily on addressing the following research question:

Are empathy-building interventions effective in terms of increasing interest?
To answer this research question, we performed a thorough analysis to verify the significance of both interventions in changing the interests of the participants.For this, we consider survey question: How interested are you in the topic of software inclusiveness? to analyze the statistical significance.We choose to use non-parametric tests, such as Wilcoxon test, in our analysis, since these tests don't make overly strong assumptions regarding our data.The overall Wilcoxon test results for both groups of interventions are provided in Table I The p-value of experiential-based intervention in Table I shows significance in the border range i.e., close to 5%.Due to the small sample size of the experiential-based group, we attempt to test whether adding or removing one or two participants matters or not.We found a notable difference in the p-value while removing only one participant.This suggests that we need strong evidence to maintain the significance of this group.Hence, we mentioned it as not significant with p-value of 0.042.For expression-based intervention, it has a large p-value of 0.305 which is larger than our chosen significant level, meaning there is no increasing in the interest of the participants.Further, we demonstrate the transition Likert score plot for experiential-based intervention in Figure 4 and for expression-based intervention in Figure 5.
In both graphical representations, the Likert scores predominantly cluster around values of 4 and 5, exhibiting minimal fluctuations from pre-to post-survey.This concentration implies that participants, particularly those from the experiential-based group, display motivation induced by the respective interventions.However, the limited variability in scores suggests that the impact may not be substantial enough to yield statistical significance in terms of augmenting interest through both interventions.
V. RELATED WORK Empathy-building interventions have been explored in various non-computing domains such as in medicine [9,19], and for creating tolerance in social situations [4].These works have found that empathy-building interventions can be beneficial for things such as understanding the perspective of the patient, or even in more optimal communication practices with a patient.These existing efforts differ from our work in that these do not focus on the educational computing domain.
Several recent studies have focused on the importance of empathy in educational activities and explored effective implementation practices.Motahar et al. [12] conducted research to understand the reasons behind the shortcomings of existing methods in fostering "design empathy".They emphasized the need for the HCI education sector to address the unresolved challenge of teaching empathetic design to students.Olesonet al. [14] introduced the Critique, Imagine, Design, Ex-pand, Repeat (CIDER) assumption elicitation technique, which aims to teach inclusive design abilities.This method helps designers recognize and address biases by critically examining assumptions about users.The researchers conducted an 11week mixed-method case study involving 40 undergraduates enrolled in an interaction design course.They supplemented this study with subsequent interviews.The results indicated that activities based on the CIDER technique had the potential to facilitate the identification of various design biases over time and encouraged students to reflect on unconscious biases related to users.
In previous research [2], empathy-building activities aimed at understanding people with disabilities have been analyzed.However, this work reveals that unfortunately, these acts of empathy can sometimes lead to a disconnect between people with disabilities and the intended support from designers.Crabb et al. [5] conducted a study with 197 participants highlighting the existence of knowledge gaps among developers in creating accessible software, despite efforts to promote accessible design.In contrast, our developed educational material serves to address and close this knowledge gap among developers, empowering them to effectively create inclusive software.
Similar to our work, several existing efforts have utilized Likert measurements to evaluate empathy [11,15].These works served as an inspiration for our Likert evaluations.
Experiential activities have demonstrated its potential benefits for creating empathy in users [7,8,16,17,18].Our work differs from this previous effort in that our analysis includes a much more diverse population set, and conducts a deeper analysis of empathy-focused research questions (e.g., a comparative analysis between expression and experientialbased interventions).

VI. CONCLUSION
The empathy-building intervention activities described in Section II has demonstrated their pedagogical effectiveness (Section III), and have been made publicly available on the project website: all.rit.edu.These brief interventions may be utilized in various educational settings, ranging from individual learning sessions, to classrooms and outreach events.The activity can benefit a diverse set of proficiency levels and focus areas, ranging from introductory to advanced computing courses, to non-computing focused curriculum such as business and project management courses.The hosted nature of the intervention will support easy adoption, since only a browser and internet connection are required for usage of the self-encapsulated activity.
This work additionally provides a hosted, easy to adopt educational resource that may be utilized in various computing and non-computing classrooms.Complete educational material is available: https://all.rit.edu (a) Experiential Activity Pre-Colorblindness Filter, where colors are easily discernible to users, emulating the experience of non colorblind users.(b) Experiential Activity Post-Colorblindness Filter where color variations and text are difficult to discern due to 'colorblindness' emulator, emulating the experience of colorblind users.

Fig. 1 :
Fig. 1: Example Intervention Screenshots where Figure 1b contains indiscernible color shapes and texts due to emulated accessibility-related issues compared to the more distinguishable accessible text shown in Figure 1a.

Fig. 2 :
Fig. 2: Expressive intervention activity where the user is tasked with recognizing discomfort in a user experiencing the colorblindness activity.

1 ) 3 ) 1 )
What Is Your Age? a) 0-17 years b) 18-29 years c) 30-50 years d) 51+ years 2) What Is Your Gender?What demographic do you most closely identify with?a) White b) Asian or P.I. c) Hispanic d) OtherPre-/Post-Survey The following questions are used in both pre-and postsurvey.The participants are asked to respond on a 1-5 Likert scale.How interested are you in the topic of software inclusiveness? 2) I can identify with the challenges posed by inaccessible software.3) I get upset when I see inequitable software.4) I have empathy for people that are challenged by issues of inclusion.5) I find that it is difficult to understand the impacts of non-inclusive software.(Circle One)

TABLE I :
. Testing results of effectiveness in terms of increasing interest for both interventions: both interventions are not significant in increasing interest.