衛生教育與行為科學介入研究成效之統合分析

The Meta-Analysis of the Effects of Intervention Studies Using Health Education and Behavioral Modification

本研究蒐集台灣地區1971年至2001年間，衛生教育和行為科學的介入研究之中文 文獻，運用統合分析的方法來評定其在健康相關行為上的介入成效。本研究所選取的介入研究應符合下列原則：1.測量健康相關行為2.研究中運用至少一種衛生教育或行為科學的介入方法，3.研究中提供控制組與對照組或前後測結果以供比較，4.研究中提供相關統計數字得以計算成效值，如頻率、百分比、圖表、t-test、F-test、Chi-Square、皮爾森積差相關、P值或平均值與標準差等。相同主題之介入文獻達三篇以上才納入本研究進行分析，研究樣本共計70篇論文。本研究由二名研究人員及主持人進行文獻閱讀和譯碼工作，一致性高達0.83。衛生 教育和行為科學介入對知識的改變成效最大(d=0.51)，為中等的正向效果，此成效於後後測時仍存在。對目標行為、自我效能和態度有介於弱和中等的正向效果，對於行為意向則只有弱效果，測量行為意向的研究則大多未測量目標行為，後後測也大致符合此一趨勢。研究結果發現對知識的介入成效最佳，行為次之，介入地點為學校或工作場所對知識有較高的成效值，多年期介入、與多元介入策略對目標行為有較高的成效值，建議以後的研究可以延長介入期程、於有組織的場所如學校、職場進行、使用多種介入策略與教學方法，介入效果會較佳。

The purpose of this study was to examine the overall effectiveness of health education and behavioral modification interventions studies on health-related behaviors from 1971 to 2001 in Taiwan area. Criteria for selecting a study were: 1). There was a dependent variable that measured health-related behaviors. 2). The methodology of study was appropriate and using education or behavioral modifications intervention designed to improve health-related behaviors. 3) Outcomes of the intervention were quantified and could be compared with a variance estimate of the outcome from a control group or pre-test condition. 4). An effect size could be calculated. Seventy retrieved research papers match by the criteria were included in this study. Multiple reviewers screen abstracts and retrieved studies. The agreement rate between two coders was 0.83. According to Cohen's criteria, there was a medium effect in knowledge intervention at post-test. Meanwhile, the intervention effects were between medium and small for targeted behaviors, self-efficacy and attitude. However, there was only a small intervention effect for behavioral intension. The results suggest that interventions produced desired effectiveness on knowledge and targeted behaviors. Intervention sites, longitudinal and multiple strategies of intervention were associated with better outcome. Future studies can employ longitudinal design, apply intervention programs to organized settings, and use multiple strategies and approaches in order to enhance the effectiveness of interventions.