果樹生產的智慧農業創新傳佈研究 －以臺東地區農民為例

Abstract

本研究以創新傳佈理論為基礎，探討臺東地區果樹農民面對智慧農業科技的認知態度、擴散路徑與採用行為。近年臺灣農業面臨氣候變遷與缺工挑戰，智慧農業成為產業轉型的關鍵策略，然而偏鄉地區資源與資訊傳遞受限，其技術擴散模式有別於都會地區。 研究採量化問卷調查法，以臺東地區果樹生產者為對象進行便利抽樣，共回收210份有效問卷，並運用獨立樣本t檢定、單因子變異數分析與多元迴歸分析進行資料處理。實證結果指出，多數受測農民已具備環境感測或自動灌溉等初步設備的使用經驗。在人口統計變項方面，女性、中生代農民以及有參與在地農業組織的生產者，對智慧農業的創新認知與採用意願均較高。 在創新傳佈特性對採用意願的影響上，相對優勢與可觀察性呈現正向預測力，複雜度呈現反向預測力，其中以相對優勢的影響力最為關鍵；相容性與可試用性則未達統計顯著水準。實證結果證實，科技操作越簡便（複雜度越低、易用性越高），越能降低農民轉換技術的心理門檻，進而提升採用意願。研究依據感知分數進行實證類型分析，發現樣本具備前端群體高度集中（早期採用者佔44.29%）、後端群體極端萎縮（落後者佔4.29%）的非對稱偏態分佈。 基於上述發現，本研究建議相關推廣單位應著重展現量化的經濟效益，善用地方基層組織建立實體觀摩場域，並研擬差異化的輔導策略。同時，針對實證分佈中後端群體稀少的現象，未來研究應實施分層隨機抽樣或立意抽樣以補強散農樣本，並搭配質性混合研究設計，深入剖析後端保守群體的深層心理抗拒與制度性阻礙，以期建構更具包容性的偏鄉智慧農業推廣模式。

This study applies diffusion of innovations theory to examine fruit growers’ perceptions, diffusion pathways, and adoption intentions regarding smart agricultural technologies in Taitung, Taiwan. As Taiwan’s agriculture faces climate change and labor shortages, smart agriculture has become a key transformation strategy. However, technology diffusion in remote rural areas is limited by resources and access to information. A quantitative questionnaire survey using convenience sampling yielded 210 valid responses. Data were analyzed using independent-samples t-tests, one-way ANOVA, and multiple regression. Most respondents had experience with basic technologies such as environmental sensors or automated irrigation. Female farmers, middle-aged farmers, and those involved in local agricultural organizations reported more favorable perceptions and stronger adoption intentions. Relative advantage and observability positively predicted adoption intention, whereas complexity negatively predicted it. Relative advantage was the strongest predictor, while compatibility and trialability were not statistically significant. Thus, simpler and more user-friendly technologies can reduce psychological barriers to switching technologies and increase adoption intention. Classification based on perception scores revealed an asymmetrical distribution, with early adopters comprising 44.29% of the sample and laggards only 4.29%. The study recommends demonstrating measurable economic benefits, establishing demonstration sites through local organizations, and developing differentiated support strategies. Future research should use stratified random or purposive sampling to better represent unaffiliated farmers and mixed methods to examine psychological resistance and institutional barriers among conservative late adopters and laggards, thereby supporting a more inclusive model for promoting smart agriculture in remote rural areas.