All research papersAI crop diagnosis

Assessing the Usability, Diagnostic Performance and Adoption Potential of Crop Cura AI-Based Diagnosis Tool Among Agriculture Students in Nnamdi Azikiwe University, Nigeria

Victor Onyekachukwu, Soetan Olatunbosun Jonathan, Christian Iwuagwu, Akerele Dare · 2026

Download PDF

Abstract

This study assessed the usability, perceived diagnostic performance, trust, practical usefulness and adoption potential of an artificial intelligence (AI)-based crop-diagnosis application among agriculture-related students in Nigeria. The analysis used 222 questionnaire and combined frequency distributions and percentages with an ordinary least squares multiple linear probability model using HC3 heteroskedasticity-robust standard errors. Most respondents had not previously used a farming application (81.5%), although 54.1% reported farming outside school. Initial usability was generally favourable: 85.1% rated setup as easy or very easy, and 81.5% rated navigation as easy or very easy. However, 47.3% experienced an application failure, freezing or photo-upload problem, while 68.5% reported that network connectivity or data costs affected use at least sometimes. Perceived diagnostic agreement was moderate to positive: 58.1% said the diagnosis matched their visual observations most times or always, while 9.9% reported that it rarely or never matched. Most respondents reported mostly or complete trust (76.1%), had applied the recommendation as provided or with an organic alternative (85.6%), and perceived clear or slight crop improvement (86.5%). Overall, 80.6% would recommend the application to other farmers. In the multivariable model, perceived diagnostic agreement (B = 0.094, p = .006) and trust (B = 0.170, p = .001) were positively associated with willingness to recommend; ease of navigation, technical failure, network/data burden and perceived crop improvement were not statistically significant at the 5% level. The findings indicate promising user acceptance but also substantial reliability and access concerns. Responsible deployment should therefore prioritise field validation on locally important crops, transparent communication of uncertainty, affordable and safe recommendations, low-data/offline functionality, and integration with agricultural extension services.

Authors

  • Victor OnyekachukwuCrop Cura Limited
  • Soetan Olatunbosun JonathanDepartment of Agricultural and Rural Development, Federal University of Agriculture, Abeokuta
  • Christian IwuagwuDepartment of Crop Sciences, Nnamdi Azikiwe University
  • Akerele DareDepartment of Crop Sciences and Horticulture

Cite this paper

Victor Onyekachukwu, Soetan Olatunbosun Jonathan, Christian Iwuagwu, Akerele Dare (2026). Assessing the Usability, Diagnostic Performance and Adoption Potential of Crop Cura AI-Based Diagnosis Tool Among Agriculture Students in Nnamdi Azikiwe University, Nigeria. Crop Cura Limited. https://www.cropcura.com/ResearchPaper/assessing-the-usability-diagnostic-performance-and-adoption-potential-of-crop-cura-ai-based-diagnosis-tool-among-agriculture-students-in-nnamdi-azikiwe-university-nigeria