•  
  •  
 

Abstract

Online business education creates persistent demands for timely instructional support, yet readily available analytics from custom educational assistants may reveal little about who used a tool, how it was used, or whether learning improved. This descriptive portfolio case study examines six custom educational assistants used in asynchronous undergraduate and graduate business courses at a large private nonprofit university. The analysis is intentionally bounded to documented portfolio characteristics: intended role, uploaded knowledge-file count, one named analytical setting, available access information, and cumulative bucketed chat values displayed as of August 13, 2026. The six displayed minimums sum to 190 chats; because the dashboard values are bucketed and user denominators are unavailable, they are not treated as actual usage shares, adoption rates, unique users, or learning outcomes. A revision-stage documentation audit separates observed configuration facts from untested judgments about grounding quality, capability, and access effects. The study uses the Chatbot LMS Integration Framework (CLIF) only as an interpretive lens where integration evidence is documented and otherwise records integration status as unknown. Findings show that the available portfolio record supports a transparent configuration inventory but does not support causal, student-level, or effectiveness claims. The article concludes with governance implications, a prospective measurement framework, and a pilot-evaluation design intended to produce stronger evidence while preserving privacy, human oversight, and academic responsibility.

Included in

Business Commons

Share

COinS