AI在教育研究領域的應用系列（十四）：善用Elicit AI研究助手進行系統性綜述與統合分析

Applications of AI in Educational Research Series (XIV): Leveraging the Elicit AI Research Assistant for Systematic Reviews and Meta-Analyses

本研究旨在探討Elicit AI研究助手在教育研究領域中進行系統性綜述與統合分析之應用成效與實際操作流程。Elicit為一款基於GPT-3語言模型開發的人工智慧工具，能透過自然語言輸入協助研究者進行文獻搜尋、摘要整理、研究比較與資料萃取，藉此提升研究效率與資料整合的精準度。本文採實作導向設計，詳細說明Elicit的操作介面、流程設計與智慧篩選機制，並透過範例資料展示其在教育領域研究中的應用模式。實測結果顯示，Elicit能顯著減少文獻篩選與整理所需時間，並自動辨識研究設計、樣本特性、實驗方法與主要結果，協助研究者更快速地完成初步篩選與主題分析。總而言之，Elicit在系統性文獻回顧與統合分析中可作為高效且可靠的輔助工具，適合教育研究者於文獻搜尋與初步篩選階段使用，將其視為輔助而非取代人為判斷，將能進一步提升學術研究的品質與可及性。

This study aims to examine the applicability and practical workflow of the Elicit AI Research Assistant in conducting systematic reviews and meta-analyses within the field of educational research. Elicit, an artificial intelligence tool developed based on the GPT-3 language model, enables researchers to perform literature searches, summarize key findings, compare studies, and extract data through natural language input, thereby improving research efficiency and the accuracy of data integration. Adopting a practice-oriented approach, this paper provides a detailed explanation of Elicit’s user interface, workflow design, and intelligent filtering mechanisms, and demonstrates its application through sample data in educational research contexts. The experimental results indicate that Elicit can significantly reduce the time required for literature screening and organization, automatically identify research designs, sample characteristics, methodologies, and key findings, and assist researchers in conducting preliminary screening and thematic analysis more efficiently. In summary, Elicit serves as an effective and reliable assistant tool for systematic reviews and meta-analyses. It is particularly suitable for educational researchers in the literature search and initial screening stages.