Explaining AI Without Code: A User Study on Explainable AI

TL;DR

No-code ML platforms often sacrifice explainability, potentially undermining transparency and trust. The authors integrate three complementary XAI techniques—Partial Dependence Plots, Permutation Feature Importance, and KernelSHAP—into DashAI and validate the approach with a human-centered user study (N=20). Results show high task success ($\geq 80\%$) across explainability tasks; novices rate explanations as useful, accurate, and trustworthy, while experts demand deeper sufficiency and completeness, with explanations also boosting perceived predictability and confidence, particularly for novices. The work demonstrates that XAI for no-code systems can be accessible to beginners yet diagnostically meaningful for experts, underscoring the importance of adaptive explanations and outlining directions for extending such approaches to larger models and interactive interpretability.

Abstract

The increasing use of Machine Learning (ML) in sensitive domains such as healthcare, finance, and public policy has raised concerns about the transparency of automated decisions. Explainable AI (XAI) addresses this by clarifying how models generate predictions, yet most methods demand technical expertise, limiting their value for novices. This gap is especially critical in no-code ML platforms, which seek to democratize AI but rarely include explainability. We present a human-centered XAI module in DashAI, an open-source no-code ML platform. The module integrates three complementary techniques, which are Partial Dependence Plots (PDP), Permutation Feature Importance (PFI), and KernelSHAP, into DashAI's workflow for tabular classification. A user study (N = 20; ML novices and experts) evaluated usability and the impact of explanations. Results show: (i) high task success ($\geq80\%$) across all explainability tasks; (ii) novices rated explanations as useful, accurate, and trustworthy on the Explanation Satisfaction Scale (ESS, Cronbach's $α$ = 0.74, a measure of internal consistency), while experts were more critical of sufficiency and completeness; and (iii) explanations improved perceived predictability and confidence on the Trust in Automation scale (TiA, $α$ = 0.60), with novices showing higher trust than experts. These findings highlight a central challenge for XAI in no-code ML, making explanations both accessible to novices and sufficiently detailed for experts.

Figures (6)

• Figure 1: Home view of DashAI. Users can access datasets, create experiments, run predictions, explore explainability tools, and manage plugins.
• Figure 2: Experiment and prediction modules in DashAI. Users configure and train models (a), and subsequently apply them to unseen data for prediction (b).
• Figure 3: Explainability module in DashAI. PDP (a) and PFI (b) provide global insights into model behavior, while SHAP (c) illustrates local instance-level explanations.
• Figure : (a) Novices ($N=10$)
• Figure : (a) Novices ($N=10$)