ReAL‑E (Remote Adult Language – Experiment)

ReAL‑E is a set of simple, online tasks designed so adults (ages 19–64) can contribute to language science from home, using their own computer and a pair of headphones. The goal is to make data collection more inclusive, efficient, and scalable—so we can answer bigger questions about communication and learning (Lancaster et al., 2025).

What is ReAL‑E? A short, at‑home battery for speech, language, and reading.

Why build ReAL‑E?

Traditional, in‑person testing takes time, space, travel, and staff—barriers that can shrink who gets to participate. By moving carefully selected tasks online, we can reach more people, collect data faster, and enable large‑scale studies (including genetics) that would be impractical otherwise.

Why online? Broader reach, lower cost, faster studies.

What participants do

ReAL‑E currently includes:

Speech-motor tasks: Syllable repetition (DDK), Real‑word repetition, and Nonword repetition
Oral and Written Language tasks: Irregular word spelling, Word definitions, Timed nonword reading, and Following directions

How we built it

We identified tasks from the literature, adapted them for web delivery, and asked experts to review face validity before running user experience sessions. Technically, tasks were coded in jsPsych and hosted on cognition.run, with screening surveys in REDCap.

Pipeline: literature → expert review → UX testing → revisions

Expert feedback. On a 5‑point scale across multiple criteria, experts gave supportive scores for several tasks (e.g., syllable repetition, nonword repetition, spelling) and flagged “following directions” as an area to improve. Overall averages: ~3.29 (Adaptation), 3.48 (Effectiveness), 3.76 (Face validity).

User feedback. People called the experience “interesting” and “kinda fun,” and asked for clearer navigation, a shorter “following directions” task, and more flexible typing (backspace in spelling). We also noted strong interest in mobile access.

Early pilot results. Early pilot testing showed healthy variability and sensible ranges across tasks (e.g., Real‑word repetition mean ≈ 92.33% ± 6.97; Irregular word spelling mean ≈ 78.00% ± 16.07; Following directions mean ≈ 77.86% ± 9.59) (Lancaster et al., 2025). These insights guided wording, timing, and interface tweaks.

Early performance snapshots from pilot testing

*Note.* DDK timing was referenced to classic time‑by‑count norms (Fletcher, 1972), while reading/language tasks used percent accuracy. We revised instructions (e.g., “repeat steadily”) and item sets based on user feedback.

What we’ve learned so far

Headphone quality: a pleasant surprise

When we compared home earbuds to lab‑quality headphones, we didn’t find meaningful performance differences for nonword or real‑word repetition—great news for at‑home testing. (Example: Nonword t(18)=−0.24, p=.810; Real‑word t(18)=1.12, p=.279.) Next up: looking at internet speeds and individual differences.

Headphones vs. earbuds: performance looks similar for repetition tasks

Do speech and language always travel together?

In a community sample, human scoring showed weak correlations among syllable repetition (DDK), irregular‑word spelling, and word definitions; automated scoring (edit distance and semantic similarity) confirmed minimal association (e.g., IWS–WD r≈−0.03). Together, results suggest a relative dissociation between speech‑motor, grapheme‑phoneme, and lexical‑semantic skills in typical adults.

Speech–language associations: small in typical adults

What’s next

Phase 2: Fully online, broader recruitment, reliability, and item selection.
Phase 3: Validity, scale‑up, and building/training automatic scoring pipelines.

Team & support

ReAL‑E is a collaboration across research, clinical, and technical teams, with support from COBRE 5 P20 GM109023‑08 (Project 13).

Posters & downloads

Do Self-Reported Language and Reading Skills Predict Performance on Remote Behavioral Tasks in Adults? Strait, Buttner, & Lancaster - Research Assisstant Flash Talks December 2025. Download the slides (PDF)

Impact of Headphone Quality on Remotely Delivered Repetition Tasks
Parks, Buttner, Fitzgerald, Bashford, & Lancaster — Conference poster March 2024. Download the poster (PDF)

Associations Between Speech and Language Performance in Adults
Lancaster, Farzana, Parks, Fitzpatrick, Buttner, Bashford, & Parde — Conference poster November 2023. Download the poster (PDF)

Purpose: Recent technological advancements in psychology have yielded multiple unsupervised online tools for measuring cognitive skills. However, these tools do not measure key features of communication abilities in adults, such as productive speech or oral language. To address this gap, we developed the Remote Adult Language Experiment (ReAL-E) using a three-step approach that incorporates domain identification, expert evaluation, and piloting testing. Method: Step 1 focused on domain and task identification by collaborating with experts and reviewing the literature. Step 2 included two rounds of evaluation by content and lay experts to assess the initial tool design, identify areas for refinement, and solicit feedback for desired features. Step 3 involved piloting the tool in a sample of adults recruited online. Results: The ReAL-E tool adapted three speech planning and production tasks and four oral and written language tasks into online versions. Feedback from content and lay experts led to significant improvements, including enhanced task instructions, presentation modifications, and the incorporation of features to address potential confounds. Pilot testing demonstrated that adults perform within expected ranges, and five out of seven tasks had good to excellent preliminary reliability. Conclusion: The development of the ReAL-E tool contributes to the field by offering a scalable, efficient method for assessing a broad range of communication skills in adults, paving the way for future research and clinical applications in developmental communication disorders. Our study underscores the collaborative effort and innovative approach required for creating a digital assessment tool that promises to expand research capabilities and improve diagnostic processes in the discipline.