跨機關服務協力合作網絡的有效形成，在官僚體制漸漸無法發揮績效的情況下，更顯得重要。為了探討在機關間協力合作的種種問題，本研究以公共網絡管理文獻建構跨機關間協力合作的模型。公共網絡管理的文獻，主要在探索公共管理者的行為與角色，以及機關間的協力合作行為，如何貢獻於服務傳遞協力合作網絡績效，而這兩個研究問題，也存在於本論文的觀察單位-臺北市政府戶政業務之中。資料來源主要是以問卷方式調查臺北市政府負責戶政業務戶政人員的主觀認知性資料。研究方法則是透過描述性統計、相關係數、信度考驗、確認性因素分析(Confirmatory Factor Analysis, CFA)、迴歸分析、迴歸診斷與結構方程式（Structural Equation Model, SEM）。 本研究的問卷抽樣數共524份，有效問卷為338份，回收率為64.5％。在統計分析上，無論在因素負荷量、信度考驗、迴歸模型解釋比率，以及結構方程式的模型適配度皆符合或接近標準。而多元迴歸分析（OLS）與結構方程式（SEM）亦顯示機關之間的資源分享與協助並無助於戶政服務傳遞協力協力合作網絡間的績效，這也解釋部分理論文獻對協力協力合作網絡過於樂觀的質疑。本研究亦發現網絡管理文獻所建構的特定因素如核心機關（民政局）的協調，以及策略與結構的協調與整合是有助於戶政服務傳遞協力協力合作網絡績效。在研究貢獻上，期許檢驗研究問題與研究假設，以及這些規範性理論應用至政府實務的深度。 / The efficient formation for interagency service delivery network is comparatively important when the bureaucracy falis to make it well in public affairs. This research employs the literatures of public network management to construct the model of interagency collaborations. Two research are discussed in public network management and our research with the observed unit of the household registration of Taipei City Government: how the behaviors and roles of public network managers and collaborative behaviors among public agencies contribute to the network performance of interagency service delivery network. The data sources are draw from subjective perceptive data of public managers and public administrators who response for household registration services in Taipei City Government. The research method employs descriptive statistics, correlations, reliability, confirmatory factor analysis, regression analysis, regressive dignosis and structural equation model. The response rates in this research are 64.5%(338/524). In statistics, the the factor loadings, reliability, R2 for the regression model, and the goodness of the fit for structural equation model(SEM) also show that the resources-sharing among the agencies does not benefit the netwotk performance in household service-delivery network, which doubt the celebratory viewpoints from some public network management literatures. This research also finds some factors such as the coordination of network core administrative agency (the Bureau of Civil Affairs), the integration and coordination of strategies and structures, are helpful for network performance in collaborative network of household service-delivery network. For the contributions of research, this research expects to examine our research questions and research hypotheses and then applies these normative theories to the depths of government practices.

2016年的英國脫歐公投，最終以51.89%比48.11%的差距，決定了英國脫歐的命運。各界紛紛揣測各種可能造成民眾投下脫歐一票的原因，包含個人經濟狀況、政黨認同、受民粹操控、反菁英、對歐洲認同等。本文針對既有研究中尚存在之空缺，聚焦「議題投票取向」對選民投票抉擇的影響，探究脫歐陣營對資訊的扭曲是否對選民的投票抉擇有顯著影響。論文就脫歐派針對國民健保、移民、脫歐後的英國對外經貿、失業率、勞工權益保障與是否能無條件持續享有歐盟單一市場好處這六項議題，透過「二分勝算對數模型」進行分析，從個體層次行為來瞭解資訊扭曲對選民在脫歐公投中投票行為的影響。 研究結果顯示，在移民潮來襲、民眾高喊反全球化與選民不安全感高漲的大環境背景下，脫歐陣營對選民所關心之議題的煽動與對資訊的操控使選民在做出投票抉擇時受到明顯誤導。此外，模型結果亦顯示，選民的個人社經背景與心理態度認知同樣對選民投票行為產生影響。總結而言，本研究透過微觀層次的分析，針對促使選民投下脫歐一票的因素做出深入探討，並以「資訊扭曲」作為重要變數，補充議題投票相關研究之不足。 / In the Brexit referendum held in June 2016, 51.89 per cent of the voter voted Leave and 48.11 per cent voted Remain. The result sealed the fate of the UK as an outsider of the European Union. While the reasons behind voter’s choice to leave the EU remain disputed, individual voter’s socio-economic status, party identification, populism, anti-elite mood, and identity toward European were among the most frequently listed factors. In order to fill the gap that existing researchs has left, this thesis focuses on the significance of “issue voting” in voting behavior. It discusses if the distortion of information by the Leave campaign had a significant impact on people’s voting choice. It uses the “logistic regression model” to analyse six issues brought about by the Leave campaign. They include the NHS, immigration, trading arrangements with other nations, unemployment, working conditions for British workers, and unconditional maintaining of all the benefits from the EU. By focusing on the individual level, I try to investigate how the distortion of information has impacted upon people’s voting behavior in the Brexit referendum. The study shows that against the background of an increased flow of immigrants and heightened anti-globalization mood, voters had a strong sense of insecurity. Under such circumstances, the Leave campaign exploited the opportunity to incite voters on issues they cared most about, manipulated information during the demagogic campaign, and misled people on their voting decision. In addition, the model also demonstrates that voters’ socioeconomic status and mental cognizance have the same effect on people’s voting behavior. In the nutshell, this thesis uses micro-level analysis to investigate voter behavior in the Brexit referendum. It brings in “information distortion” as a key variable in explaining voter behavior, a variable largely neglected in the existing literature on issue voting.