App Choice and AAC Outcomes: Insights from a Large-Scale Family-Reported Outcomes Dataset

Executive Summary

Suggested Citation

DeYoung, C. M., Fisher, K., & Kemerling, L. (2026). App choice and AAC outcomes: Insights from a large-scale family-reported outcomes dataset (AbleNet Research & Insights White Paper No. 1). AbleNet, Inc. https://www.ablenetinc.com/

Conflict of Interest Disclosure

The author is employed by AbleNet, Inc. and holds equity in the company as a participant in its Employee Stock Ownership Plan (ESOP). This white paper was produced as part of the company’s internal research program. Readers should consider this context when evaluating the findings.

Background

The Growing Diversity of AAC App Options

The AAC app marketplace has expanded considerably in recent years, giving families and practitioners a wider range of options than at any point in the field’s history. Among the most widely adopted apps in current use are TouchChat, TD Snap, Proloquo2Go, and LAMP Words for Life, all of which fall into the category of “robust” AAC systems: apps designed to support a large, generative vocabulary rather than a small, fixed set of preprogrammed phrases. A recent survey of school-based AAC practitioners found that these four apps, along with low-tech communication boards, accounted for the substantial majority of systems currently in use, with TouchChat with WordPower being the most commonly reported single preferred system, followed by LAMP Words for Life, Proloquo2Go, and TD Snap with Core First (Senner et al., 2025).

Each of these apps differs in its default layout and vocabulary architecture. TouchChat’s default configuration uses a 60-icon grid with a vocabulary exceeding 2,000 words. TD Snap similarly defaults to a 60-icon layout but supports a larger vocabulary set of more than 2,700 words. Proloquo2Go uses a denser 77-icon grid with access to over 4,700 words, and LAMP Words for Life uses an 84-icon grid paired with a vocabulary exceeding 4,000 words. Despite these differences in grid density and vocabulary size, all four apps share a common design philosophy: each is built around a robust, generative language system intended to grow with the user, rather than a limited set of static phrases. This shared foundation is worth noting because it means that, at least in principle, differences in communication outcomes across these apps would need to be explained by differences in layout, navigation, and vocabulary organization rather than by fundamentally different approaches to language representation.

What’s notably absent from the literature is research that directly evaluates whether these structural differences translate into differences in communication outcomes for the people who use them. This gap sets up the central question this white paper is positioned to address.

Gaps in the Outcomes Literature

The existing body of AAC research offers a great deal of insight into whether AAC works, but comparatively little insight into whether the specific app a person uses shapes how well it works. Much of the comparative literature that does exist compares broad categories of AAC (manual sign, picture exchange, and speech-generating devices) rather than comparing named commercial apps against one another. A systematic review and meta-analysis of single-subject design studies found that picture exchange and speech-generating device applications were comparably effective for teaching social communication skills to individuals with autism spectrum disorder, while manual sign was generally less effective, though the review also noted that a large number of comparative studies failed to meet rigorous design standards because they compared more than two systems within the same experimental phase (Aydin & Diken, 2020). This body of work is informative, but it speaks to modality-level differences (aided versus unaided AAC) rather than to differences among today’s commercially available, feature-matched apps.

Research that does focus on specific apps has tended to concentrate on design and usability rather than on outcomes. A design-based study conducted with special education teachers and university faculty identified a set of design features associated with more usable and engaging AAC apps, including personalization options, simple and uncluttered interfaces, hierarchical presentation of vocabulary, responsive feedback, built-in help documentation, and parental control features (Polat & Delialioğlu, 2023). This work is valuable for understanding what makes an app usable, but it does not address whether these design differences produce measurable differences in communication outcomes once an app is deployed in real-world use.

Separately, research on how practitioners select AAC systems suggests that the selection process itself may be a more significant variable than has been widely recognized. A recent survey of AAC professionals in US schools found that practitioners using a single, preferred system across their caseload were significantly less likely to involve other professionals or parents in the decision-making process, and that clearly defined criteria for evaluating a student’s success with a given system were reported by fewer than one in four respondents (Senner et al., 2025). This finding does not test outcomes directly, but it raises an important possibility worth carrying into this analysis: that variability in how confidently and consistently an app is selected and implemented may influence outcomes as much as, or more than, which specific app is chosen.

Taken together, the literature leaves a clear and specific gap. We know a fair amount about AAC modalities in general, and we know something about what makes an app usable in principle, but we know very little about whether communication outcomes actually differ across the robust AAC apps most commonly used today.

Study Rationale and Research Questions

AbleNet’s Family-Reported Outcomes (FRO) dataset, comprising responses from families of QuickTalker Freestyle users, provided a distinctive opportunity to begin addressing this gap. Because the FRO dataset already captures which AAC app each respondent’s family uses, alongside a range of communication outcome measures, it allowed us to explore, for the first time at this scale, whether app choice is associated with differences in communication outcomes as reported by families. Given how limited the existing literature is on this specific question, this analysis was designed as an exploratory first step rather than a confirmatory test of a well-established hypothesis. The guiding research question was straightforward: do families of QuickTalker Freestyle users report different communication outcomes depending on which AAC app their child uses?

Scope and Limitations of This Analysis

Several important boundaries define what this analysis can and cannot tell us. First, and most importantly, this analysis is observational rather than experimental. App group was not randomly assigned; families arrived at their current app through their own clinical and personal decision-making process. To compare outcomes across app groups, we used Kruskal-Wallis H tests, which assess whether the distribution of an outcome variable differs across independent groups. These tests can tell us whether reported outcomes differed by app group, but because app assignment was not randomized, they cannot tell us whether the app itself caused any observed differences (or lack thereof), as opposed to unmeasured factors such as practitioner confidence, implementation fidelity, family engagement, or the severity of the user’s communication needs at the time of app selection.

Second, this analysis is limited to apps loaded onto the QuickTalker Freestyle device, and does not extend to AAC apps used on other hardware platforms. Third, the analysis focuses specifically on the four most commonly used apps in our customer database (TouchChat, TD Snap, Proloquo2Go, and LAMP Words for Life) because sample sizes for less commonly used apps were not sufficient to support reliable statistical comparison. As our second year of Family Reported Outcomes data collection continues, we anticipate that growing sample sizes may make additional app-level comparisons possible in future analyses.

Dataset Overview

The analyses in this white paper draw on a subset (n = 5,449) of AbleNet’s complete Family-Reported Outcomes dataset (n = 5,785), one of the largest caregiver-reported AAC outcomes collections in existence. The subset includes only the four AAC apps with sufficient sample sizes for meaningful comparison: TouchChat HD with WordPower, Proloquo2Go, TD Snap, and LAMP Words for Life.

Data Collection

• Surveys were distributed to families and caregivers of QuickTalker Freestyle users via email outreach, administered using Survey Monkey.
• Respondents reported on their QuickTalker Freestyle user’s communication behaviors, AAC use patterns, and related outcomes.

Sample Characteristics

Table 1 summarizes key demographic characteristics of the 5,449 QuickTalker Freestyle users included in this analysis, spanning age, diagnosis, and AAC app. The sample is predominantly young (91% under age 18) and diagnosed with a language or autism spectrum disorder. TouchChat HD is the most commonly used app, accounting for 42% of the sample.

Analytical Approach

The goal of this analysis was to examine whether app choice was associated with communication-related outcomes, using a subsample restricted to the four AAC apps with sufficient sample sizes for comparison.

Outcome Variables

Twelve outcome variables were examined in this analysis, covering caregiver-reported satisfaction, communication-related change, and a composite summary score.

Satisfaction

Caregiver-reported satisfaction with the QuickTalker Freestyle device, rated on a 5-point scale from Very Unsatisfied to Very Satisfied.

Frustration/Behavior

Change in the user’s frustration and challenging behavior since starting to use the device (Answer choices: decreased, stayed the same, or increased). Higher scores reflect a decrease in frustration/behavior, consistent with a better outcome.

Language Understanding

Change in the user’s understanding of language since starting to use the device (Answer choices: decreased, stayed the same, or increased).

Word/Phrase Use

Change in the user’s use of words, phrases, or sentences since starting to use the device (Answer choices: using less, using about the same amount, or using more). 

Verbal Speech

Change in the user’s verbal speech since starting to use the device (Answer choices: decreased, stayed the same, or increased). Users not yet using verbal speech were excluded from this item rather than scored as a negative outcome.

Communication Function Items (six domains)

Change in the user’s use of each of the following communication functions since starting to use the device (Answer choices: decreased, stayed the same, or increased):

• Requesting (asking for things, actions, or attention)
• Protesting (expressing disagreement or objection)
• Describing (sharing information about objects)
• Asking and Answering Questions
• Commenting (sharing thoughts or opinions)
• Expressing Feelings (e.g., “I’m happy,” “I’m tired”)

Users not yet demonstrating a given communication function were excluded from that item rather than scored as a negative outcome.

Composite Outcome Score

A single summary score calculated as the row-wise mean across all eleven individual outcome items (Answer choices: satisfaction, frustration/behavior, language understanding, word/phrase use, verbal speech, and the six communication function items), using all available data for each respondent.

Predictor Variable

The primary predictor variable was app choice: the AAC application in use on the QuickTalker Freestyle at the time of data collection. Rather than relying on caregiver report, app choice was operationalized using AbleNet’s Salesforce system, reflecting the app currently loaded on the user’s device. Raw app entries included multiple licensing, subscription, and bundle variants for some platforms; these were consolidated into a single label per platform as follows:

• TouchChat HD with WordPower: TouchChat HD – AAC with WordPower, and the TouchChat HD – AAC with WordPower + PCS bundle
• Proloquo2Go: Proloquo2Go, Proloquo (Subscription), and the Proloquo2Go + Gateway bundle
• TD Snap: TD Snap AAC (Perpetual License), TD Snap AAC (Subscription), and the TD Snap AAC (Subscription) + Gateway and + PODD bundles
• LAMP Words for Life: LAMP Words for Life (no variants)

Users of apps outside these four groups were excluded from the analytic sample rather than grouped into an “Other” category, since the remaining apps did not have sufficient sample sizes to support reliable comparison.

Statistical Methods

Given the ordinal and non-normal distribution of outcome variables, non-parametric methods were used throughout.

• Kruskal-Wallis H tests were used to compare outcome distributions across app groups, run separately for each of 12 outcome variables (11 individual items plus a composite score).
• Effect sizes were calculated using epsilon-squared to contextualize the practical significance of statistically significant findings (small < .04, medium .04–.16, large > .16).
• Because all 12 outcomes were tested against the same app grouping within a single analytic family, a Bonferroni correction was applied across the 12 tests, setting the adjusted significance threshold at α = .05 / 12 = .0042.
• Sample sizes vary slightly across outcomes due to pairwise handling of missing data; each test includes all respondents with non-missing data for that specific outcome.
• All analyses were conducted in R (version 4.6.0) using the dplyr, tidyr, and rstatix packages.

Covariates and Limitations

Covariates and Limitations

This analysis did not statistically control for covariates. App group was compared against each outcome variable using bivariate Kruskal-Wallis tests, without adjusting for age, diagnosis, length of AAC device use, SLP engagement frequency, or other factors that may plausibly relate to both app choice and outcomes. As a result, observed differences between app groups cannot be attributed to app choice alone, and may partly reflect differences in the populations using each app rather than the app itself.

Several additional limitations should be considered when interpreting these findings:

• Non-random app assignment. Users were not randomly assigned to an app; app choice may be influenced by SLP recommendation, clinical setting, insurance or funding pathways, diagnosis, or family preference. Any of these factors could independently affect outcomes, making it difficult to isolate the effect of the app itself.
• Self-report and recall bias. Outcome measures reflect caregiver-reported perceptions of change since starting device use, collected at a single point in time. These ratings rely on caregiver recall and interpretation rather than standardized, independently observed measures of communication behavior.
• Cross-sectional design. Outcomes were assessed once per respondent rather than tracked longitudinally, so the analysis cannot speak to the trajectory, timing, or durability of change over time.
• Survey non-response. Findings are based only on caregivers who chose to respond to the survey; families who did not respond may differ in ways that affect both app use and reported outcomes, limiting generalizability to the full user population.
• Restricted app sample. Respondents using apps outside the four analyzed groups were excluded rather than retained in an “Other” category, since remaining apps lacked sufficient sample size for reliable comparison. Findings apply only to users of these four apps and should not be generalized beyond them.
• Multiple comparisons. Twelve outcome variables were tested against the same app grouping. A Bonferroni correction was applied to reduce the risk of false positive findings across this set of tests, which is a conservative approach that may also increase the risk of overlooking true effects that don’t survive correction.

Findings

Primary Finding: App Choice and Outcomes

Because twelve outcome variables were tested, a Bonferroni correction was applied, setting the corrected significance threshold at α = .0042 (.05/12).

App choice was not associated with meaningful differences in outcomes. Of the twelve outcome variables examined, eleven showed no statistically significant differences across the four app groups, and the one outcome that did reach significance (Commenting) reflected a negligible effect size.

In plain terms: families using TouchChat HD, Proloquo2Go, TD Snap, and LAMP Words for Life reported broadly similar outcomes for their QuickTalker Freestyle users, regardless of which app their loved one used. This includes the Composite Outcome score, which summarizes all individual outcome items into a single measure and also showed no significant difference across app groups (H(3) = 4.098, p = .2511).

Notable Exception: Commenting

Commenting was the only outcome to show a statistically significant difference across app groups, and it remained significant after Bonferroni correction (adjusted p = .0228). The effect size (ε² = .004) falls well below conventional thresholds for even a small effect (Tomczak & Tomczak, 2014), suggesting that group differences, while statistically detectable, account for a negligible proportion of variance in the reported Commenting outcomes. In practical terms, this difference is unlikely to be meaningful to families or clinicians choosing between apps, even though it is unlikely to be due to chance alone given the large sample size.

This pattern, a statistically significant result with a negligible effect size, is common in large-sample research: with N in the thousands, even very small group differences can reach statistical significance. Statistical significance and practical significance are not the same thing, and this finding should be interpreted with that distinction in mind.

Discussion

Interpreting the Findings

The absence of meaningful outcome differences across app groups, including on the Composite Outcome score, is a notable finding, but it is not an unexpected one when viewed through the lens of how the AAC field understands app selection. Clinical guidance on AAC device and app selection has long centered on feature matching rather than brand selection. This means that the process of matching a person’s strengths, needs, and communication goals to available tools and strategies, not the choice of one commercially available app over another, is what clinical frameworks treat as the active ingredient in AAC success(Glennen & DeCoste, 1997). Gosnell et al. (2011) extended this feature-matching approach directly to communication apps, cautioning the field against selecting communication apps based on media coverage, testimonials, or informal recommendations rather than a clinically grounded matching process.

If feature matching, rather than app identity, is the mechanism driving outcomes, then a finding of no meaningful difference across four major commercially available apps is consistent with theory rather than contrary to it. TouchChat HD, Proloquo2Go, TD Snap, and LAMP Words for Life all offer core AAC functionality (symbol-based communication, customizable vocabulary, and speech output), and when any of these tools is matched well to a user’s needs, similar outcomes would be expected regardless of brand. This reframes the discussion in a clinically useful way: the meaningful variable is not which app a family selects, but how well that app was matched to the individual’s access needs, language level, and communication goals, and how consistently it was implemented across environments (Light & McNaughton, 2015).

It’s also worth grounding the Commenting finding in appropriate statistical context. With a sample this large, even trivial group differences can reach statistical significance, a well-documented phenomenon in large-sample research (Lin et al., 2013). Sullivan and Feinn (2012) make the case plainly: statistical significance indicates that an observed difference is unlikely to be due to chance, while effect size indicates whether that difference is large enough to matter in practice. An effect size of ε² = .004 falls far below even a small effect by conventional standards, so despite reaching significance, the Commenting finding does not represent a difference that should influence app selection decisions in practice.

Taken together, these findings shift the conversation away from “which app is best” and toward “how is the chosen app being matched and implemented for this specific user,” which is a more actionable and evidence-aligned question for the field.

Implications for Clinicians

• App selection: This analysis reinforces that app choice, in isolation, should not be treated as a primary clinical lever for improving outcomes. SLPs conducting AAC assessments are better served by centering the evaluation on feature matching (vocabulary organization, access method, rate enhancement, and language representation matched to the individual) rather than steering families toward one specific commercial product over another based on popularity or anecdote (Gosnell et al., 2011). Given that none of the four apps in this sample showed a meaningfully different relationship with outcomes, SLPs can reasonably reassure families that, within this set, the decision is less about finding the single “right” app and more about how well any of these tools fits the individual and how consistently it’s implemented.
• SLP engagement: The null result on app choice does not mean intervention doesn’t matter. It suggests the opposite: if the brand of app isn’t driving outcomes, the quality and intensity of clinical support around whichever app is chosen likely matters more. Reviews of AAC modeling interventions have found that structured, partner-implemented strategies produce meaningful linguistic gains across pragmatics, vocabulary, and syntax (Sennott et al., 2016), and Light and McNaughton (2015) argue that improving outcomes for AAC users requires rethinking intervention design itself, not simply the device involved. This points toward SLP-led coaching, modeling, and follow-through as a more promising lever for outcomes research and clinical practice than app brand comparisons.
• Family and caregiver role: Caregivers are consistently identified in the literature as central implementers of AAC strategies in daily life, not passive recipients of a device recommendation (Elmquist et al., 2023). Kim and Soto (2024) found that AAC implementation and potential abandonment are shaped heavily by the strength of collaboration between families and school-based professionals, and by whether families feel equipped and supported in using the system at home. This suggests that the quality of caregiver coaching and ongoing collaboration between families and SLPs may be a more productive area for future outcomes research than app-level comparisons.

Implications for Families

If you’re a family choosing between AAC apps, this research offers some reassurance: the specific app your loved one uses is unlikely to be the deciding factor in their communication progress. What matters more is whether the app fits your loved one’s needs and abilities, and whether you and your speech-language pathologist can consistently support its use across home, school, and community settings. Rather than searching for the “best” app, it may be more helpful to focus on working closely with your SLP to match features to your loved one’s specific strengths and goals, and to build consistent routines for using the app throughout daily life.

Limitations

As with any observational dataset, these findings should be interpreted in light of several important limitations.

• The dataset is cross-sectional in nature for the primary outcomes measure, which limits causal inference.
• Respondents are self-selected and may not be representative of all QuickTalker Freestyle users.
• App usage was family-reported and may not reflect the full complexity of a user’s AAC experience (e.g., device sharing, app switching).
• This analysis examined app identity as a categorical grouping variable but did not account for how well each app’s features were matched to individual users at the point of selection, which the feature-matching literature suggests may be the more clinically meaningful variable (Gosnell et al., 2011).
• The analysis does not capture intervention intensity, SLP involvement, or caregiver implementation fidelity, all of which the literature identifies as potentially stronger drivers of outcomes than app identity alone (Sennott et al., 2016; Elmquist et al., 2023).
• With a Bonferroni-corrected sample of this size, the analysis is well powered to detect even very small effects, meaning the null results for 11 of 12 outcomes reflect a genuine absence of meaningful group differences rather than insufficient statistical power.

Conclusion

This analysis examined one of the most common questions facing AAC users and their support teams: does the specific communication app a user relies on shape their outcomes?

[ Write 2-3 paragraphs summarizing the core conclusion, its significance for the field, and what it means for AbleNet’s ongoing research mission. End with a forward-looking statement about future research directions or next steps. ]

About This Research

About AbleNet, Inc.

AbleNet, Inc. is a 100% employee-owned company dedicated to supporting individuals with complex communication needs and their caregivers. Through its QuickTalker Freestyle product line and expanding research program, AbleNet is committed to generating and sharing evidence that improves outcomes for AAC users worldwide.

Data Availability

The de-identified dataset underlying this analysis is available for qualified researchers at [ OSF project URL or ‘Open Science Framework (OSF): link forthcoming’ ]. Researchers interested in collaboration or data access may contact [ contact email or department ].

Acknowledgments

[ Acknowledge any collaborators, IRB support, university partners, or individuals who contributed to data collection or analysis. ]

Conflict of Interest Disclosure

The author is employed by AbleNet, Inc. and holds equity in the company as a participant in its Employee Stock Ownership Plan (ESOP). This white paper was produced as part of the company’s internal research program. Readers should consider this context when evaluating the findings.

Suggested Citation

DeYoung, C. M., Fisher, K., & Kemerling, L. (2026). App choice and AAC outcomes: Insights from a large-scale family-reported outcomes dataset (AbleNet Research & Insights White Paper No. 1). AbleNet, Inc. https://www.ablenetinc.com/

References

Aydin, O., & Diken, I. H. (2020). Studies comparing augmentative and alternative communication systems (AAC) applications for individuals with autism spectrum disorder: A systematic review and meta-analysis. Education and Training in Autism and Developmental Disabilities, 55(2), 119–141.

Elmquist, M., Crowe, B., Wattanawongwan, S., Reichle, J., Pierson, L., Simacek, J., Hong, E. R., Liao, C.-Y., & Ganz, J. B. (2023). Caregiver-implemented AAC interventions for children with intellectual or developmental disabilities: A systematic review. Review Journal of Autism and Developmental Disorders. Advance online publication. https://doi.org/10.1007/s40489-023-00394-2

Glennen, S. L., & DeCoste, D. C. (1997). The handbook of augmentative and alternative communication. Singular Publishing Group.

Gosnell, J., Costello, J., & Shane, H. (2011). Using a clinical approach to answer “What communication apps should we use?” Perspectives on Augmentative and Alternative Communication, 20(3), 87–96. https://doi.org/10.1044/aac20.3.87

Kim, J., & Soto, G. (2024). A comprehensive scoping review of caregivers’ experiences with augmentative and alternative communication and their collaboration with school professionals. Language, Speech, and Hearing Services in Schools, 55(2), 607–627. https://doi.org/10.1044/2024_LSHSS-23-00117

Light, J., & McNaughton, D. (2015). Designing AAC research and intervention to improve outcomes for individuals with complex communication needs. Augmentative and Alternative Communication, 31(2), 85–96. https://doi.org/10.3109/07434618.2015.1036458

Lin, M., Lucas, H. C., & Shmueli, G. (2013). Too big to fail: Large samples and the p-value problem. Information Systems Research, 24(4), 906–917. https://doi.org/10.1287/isre.2013.0480

Polat, H., & Delialioğlu, Ö. (2023). Design specifications of augmentative and alternative communication apps for individuals with autism spectrum disorder. International Journal of Technology in Education (IJTE), 6(1), 19–36. https://doi.org/10.46328/ijte.396

Senner, J. E., Quach, W., Chung, Y., Patterson, B., Goldman, A., Blackstone, S., & Post, K. A. (2025). Utilization of a single preferred AAC system in US schools. Augmentative and Alternative Communication. https://doi.org/10.1080/07434618.2025.2516763

Sennott, S. C., Light, J. C., & McNaughton, D. (2016). AAC modeling intervention research review. Research and Practice for Persons with Severe Disabilities, 41(2), 101–115. https://doi.org/10.1177/1540796916638822

Sullivan, G. M., & Feinn, R. (2012). Using effect size, or why the P value is not enough. Journal of Graduate Medical Education, 4(3), 279–282. https://doi.org/10.4300/JGME-D-12-00156.1

Tomczak, M., & Tomczak, E. (2014). The need to report effect size estimates revisited. An overview of some recommended measures of effect size. Trends in Sport Sciences, 21(1), 19-25.