這篇論文的研究目的是透過一個新的民調方法，聯合分析(cojoint analysis)，來檢視選民如何評估預算偏好。明確的說，在假設其他支出項目及稅率不變下，我們評估選民如何在四個支出項目(教育、福利和健康、治安、及環保和天然資源；上做取捨(tradeO由)。傳統問卷通常以一種像願望清單(wishlist)的方式提出不切實際及不完整的問題謙民眾選擇，因此傳統問卷實際上測量到的只是民眾的慾望(desires)而非他們真正的偏好Ltruepre比rences)。相對來說，聯合分析強迫受訪者在眾多因素中根據其相對的重要性而做出取捨，因此我們可以更精確的收集到選民政策偏好。我們的研究結果發現從聯合分析與傳統問卷所得到的結果差距很大：有時在某些政策領域，選民的政策偏好竟完全相反。這個研究結果的涵義在於：許多研究使用從傳統民調得到的結果來解釋、分析、評估、及預測各種的政治及政策議題。如果這些從傳統民調得到的結果是不正確或不完整的，這些基於傳統民調報告的研究結果可能減弱其說服力、產生誤導的訊息、甚至得到錯誤的結論。更者，如果政府官員基於傳統民調得到的結果來制定政策，這些政策將無法反映人民真實的政策偏好。民主政治要能反映民意的這項特色也將大打折扣。最後，聯合分析在本研究中的信度(reliability)均經過不同方式的測試且獲得令人滿意的結果。

　　　　　The purpose of this study is to examine how voters assess budgetary preferences byapplying a method called conjoint analysis. Specifically, we evaluate how voters maketradeoffs among four spending areas: education, welfare and health, crime and publicsafety, and natural resources and environmental control while holding other programsand tax rate constant. Traditional survey questions usually are addressed in a "wishlist" type of format with unrealistic choices among incomplete sets of budget options,thus they actually measure people's "desires" rather than their "true preferences." Incontrast, conjoint analysis asks respondents to make more "real world" decisions byforcing them to trade off competing interests and hence could offer researchers a way tocollect more accurate information about voters' true preferences. Our findings reveal that voters' policy preferences from traditional surveys andconjoint estimates are quite different; sometimes voters' preference orders for aparticular spending area even go in completely opposite direction. The implicationfrom our findings is that ifpolicymakers make policy decisions based on inaccurate andincomplete information from traditional surveys, these policies cannot reflect peoples'true preferences and hence the quality of democracy might be discounted. Furthermore, in order to illustrate the applicability of conjoint analysis, simulationshave been performed to predict the 98' gubernatorial election in Michigan. Lastly,reliability tests demonstrate that our conjoint estimates yield satisfactory results.