IEEE 802.16 是一支援都會型無線網路的協定，IEEE 802.16支援PMP模式(點對多點)和網狀模式兩種。在網狀模式中，所有節點的構成仿如ad-hoc方式，並依據在控制性子框中的排程資訊來計算下次遞送時間。在資料傳送之前，會有一段設定連線的時間。這段時間，每一個節點都必須跟鄰節點競爭，以取得廣播它的排程資訊給鄰節點的機會。這樣的行為跟它過去的歷史無關。換句話說，它具有”時間同質性”而適合以隨機程序來模擬。在這篇論文中，我們將用排隊程序來建立排程行為的模型，然後以馬可夫鏈來估計它的平均延遲時間，也就是一節點持續地競爭直到贏為止的這段等待時間。 / The IEEE 802.16 standard is a protocol for wireless metropolitan networks. IEEE 802.16 MAC protocol supports both of PMP (point to multipoint) and Mesh mode. In the mesh mode, all nodes are organized in a fashion similar ad-hoc and calculate their next transmission time based on the scheduling information performed in the control subframe. Before data transmission for a certain node, there is a period of time to setup the connection. During this period, each node has to compete with each other for the opportunity to advertise scheduling messages to its neighbors. This behavior does not depend on past history. In other words, it is a “Time Homogeneous” and suitable for being modeled by stochastic process. In this thesis, we will model this scheduling behavior by queuing process, and apply the Markov Chain to estimate its average delay time which a node keep waiting until it win the competition.

為了降低無線網路基地台後端之backhaul成本及解決Ad hoc網路涵蓋面積問題，無線網狀網路WMNs(Wireless Mesh Networks)因此應運而生。WMNs網路上的節點裝置與Ad hoc網路上的行動裝置對電量消耗及負載的需求是非常不同的，所以在Ad hoc網路上可使用之路由協定在WMNs網路上是無法直接適用的。 在Pure Ad hoc網路上考量Power之MMBCR(Min-Max Battery Cost Routing)及考量Loading之CSLAR(Contention Sensitive Load Aware Routing)等路由協定並沒有考量到WMNs網路上不同元件間的不同特性。有鑑於此，我們著重在Hybrid WMNs網路環境上，並提出了在Mesh Clients及Routers上同時考量Power及Loading的路徑演算法，我們稱之為POLAR。實驗結果顯示我們的路由協定能夠提昇整體的網路效能及延長網路存活時間。 / In order to reduce the backhaul cost and solve Ad hoc network coverage problem, WMNs (Wireless Mesh Networks) arise at the historic moment. The requirements on power efficiency and loading are much different between mesh nodes of WMNs and mobile hosts of ad hoc networks. The routing protocol used in Ad hoc networks would be not suitable in WMN networks. The power-aware routing in MMBCR (Min-Max Battery Cost Routing) and load-aware routing in CSLAR (Contention Sensitive Load Aware Routing) used in pure Ad hoc networks don‘t consider the different characteristics of the components in WMNs. In view of this, we focus on the Hybrid WMNs environment, and propose a combined POwer-Aware with Load-Aware Routing algorithm (Called POLAR) along mesh clients and routers. The experimental results show that our routing protocol can enhance the network efficiency and lengthen the network live time.

本篇論文使用了LHP的近似方法去評價擔保債權憑證，並推導出漸進極值因子模型，又稱單因子copula模型，單因子copula模型被廣泛運用在CDO之風險管理與一些風險因子模擬之應用，但由於2008年之金融海嘯造成市場標準模型Gaussian copula model會有評價上的誤差，所以為了能在市場不穩定時能更精確的求算出分券價差，我們必須找到一個更簡單且快速捕捉到市場不穩定性的模型。在這篇論文中，我們引用了Anna Schloesser在2009年所提出以NIG copula model為基礎的兩個延伸，讓模型更穩健和且擁有良好的性質去進行模擬，NIG Regime-Switch 模型有兩大特色: (i)可以用一致的方法去評價不同到期日的分券，放寬了同一分券必須是相同到期日的假設，和(ii)有不同的相關係數狀態，對於金融風暴來說，狀態轉換可以有效地降低市場不穩定所帶來的評價誤差。本文也對不同模型下的CDO進行風險分析與避險，分券的期望損失廣泛被信評公司視為一項審定信用評等重要的風險衡量指標，但是並無法真實反映出擔保債權憑證分券之間相對風險之大小，因此本文採用期望損失率的觀念，利用期望損失佔本金的比例來比較各分券之相對風險，且本文也求算出CDO之避險參數，讓投資人了解對合成行擔保債權憑證分券避險時所需之避險部位，分券持有人也可依據所要規避的風險類型，選擇市場上現有的信用違約交換指數或是單一資產之信用違約交換(single-name credit default swap)來進行避險。 / This paper presents the Large Homogeneous Portfolio (LHP) approach to the pricing of CDOs and we derive the one-factor copula model. It is popular that the one-factor copula models are very useful for risk management and measurement applications involving the generation of scenarios for the complete universe of risk factors. However, since the financial crisis in 2008 induces some errors in the valuation by Gaussian copula model, which is originally adopted by credit rating firms, it is necessary to have a simple and fast model that can capture the market unstableness. In this paper we apply two extensions of the NIG copula model, which are first present by Anna Schloesser (2009), since they make the model well defined and powerful for scenario simulation. The NIG Regime-Switch copula model allows for two important features: (i) tranches with different maturities modeled in a consistent way, and (ii) different correlation regimes. The regime-switching component of the NIG copula model is especially important in view of the financial crisis. This paper also targets on different models to conduct risk analysis and hedging strategy. The expected loss of tranches is widely used by credit rating organizations as one of the important indicators for risk measurement. However, it can’t reflect the relative risk level between CDO’s tranches. Therefore, our research adopts the concept of expected loss rate, which use the proportion of expected loss to total principal amount to compare the relative risk of each tranche. Moreover, when we want to hedge the spread risk of synthetic CDO tranches, the holders of tranches can choose the existing CDS index or the single-name CDS based on different risks types to hedge. The employment of the NIG Regime-Switch copula model not only has more precise estimation for the spread of tranches but also possess more stable hedge ratio to hedge.

研究目的：本研究的目的，是以縱貫研究的方式探討自閉症類疾患兒童自學齡前期至學齡期的臨床診斷穩定度、自閉症狀嚴重度變化、適應行為型態與其變化、內外化行為。 研究方法：本研究納入28名自閉症類疾患與18位非自閉症類疾患，於三個時間點進行評估：時間點一平均生理年齡為2歲半、時間點二為4歲、時間點三為9歲半。本研究分析三個時間點之臨床診斷穩定度以及後兩個時間點之評估資料。症狀嚴重度方面，以自閉症診斷會談問卷修訂版之晤談評估；適應行為方面，以文蘭適應行為量表第二版評估；認知功能方面，時間點二時以穆林發展量表評估，時間點三時使用魏氏兒童智力量表第四版評估；內外化行為方面，時間點三時以阿肯巴克實證衡鑑系統中的兒童行為檢核表評估。 研究結果：本研究結果顯示，自閉症類疾患從學齡前期至學齡期的臨床診斷穩定度高，若分亞型來看，自閉性疾患的診斷穩定度比其他未註明之廣泛性發展疾患高。症狀嚴重度方面，學齡期時自閉症類疾患兒童的溝通、侷限及重複刻板行為有顯著改善，社交行為亦有改善趨勢。適應行為方面，學齡前期及學齡期皆未發現自閉症類疾患與非自閉症類疾患之表現型態有差異，兩組在社會化領域的表現皆較其他領域來得差。自閉症類疾患的適應行為於學齡期時，年齡當量顯著增加，但社會化適應行為標準分數反而退步；進一步以認知功能分組檢視，則發現高功能自閉症類疾患在日常生活技巧與社會化領域的進步幅度顯著較多。內外化行為方面，學齡期時自閉症類疾患個案的退縮/憂鬱分量尺與社會問題分量尺之分數高於非自閉症類疾患，且自閉症類疾患在退縮/憂鬱、社會問題、注意力問題等三個分量尺上，皆有超過一半的個案達到邊緣範圍以上之分數；未發現內外化行為與自閉症狀嚴重度之間的相關。 結論：自閉症類疾患兒童到了學齡期大部分仍維持臨床診斷，顯示其自閉特徵仍然穩定存在，不過症狀嚴重度的整體趨勢會隨著年紀改善；但若以個體差異來看，每位個案進步的程度呈現高異質性，每位個案症狀改變的程度可能不同。自閉症類兒童的社會化適應行為從學齡前期至學齡期皆為其弱項，雖適應行為會隨著年齡進步，但社會化適應行為仍追不上同儕程度；高功能自閉症類疾患兒童的適應行為在日常生活技巧與社會化領域的進步幅度顯著較多，可能與認知功能有關。內外化行為方面，自閉症類疾患兒童的退縮/憂鬱問題、社會問題、注意力問題較多，但與症狀嚴重度無關。 / Purposes: The purpose of this longitudinal study was to examine the stability of clinical diagnosis, symptom severity, adaptive function and enternal/external behavior from toddler to school age in children with autism spectrum disorders (ASDs). Methods: Twenty eight children diagnosed with ASDs and eighteen children with Non-ASDs participated at time 1 (mean chronological age was 2.5 years old), and then followed at time 2 (mean chronological age was 4 years old) and at time 3 (mean chronological age was 9.5 years old). This study examines the stability of clinical diagnosis from time 1 to time 3, and analyze the measurement data between time 2 and time 3. Autism Diagnostic Interview-Revised (ADI-R) was used to measure autistic symptom severity at time 2 and time 3; Vineland Adaptive Behavior Scales-II (VABS-II) was used to measure daily adaptive behavior at time 2 and time 3; Mullen Scale of Early Learning (MSEL) or Wechsler intelligence scale for children-Fourth Edition (WISC-IV) was used to measure cognitive function at two time points; Child Behavior Checklist for Ages 6-18 (CBCL/6-18) was used to measure enternal and external behavior at time 3. Results: The stability of clinical diagnosis was high from toddler age to school age, and the stability of Autistic Disorder was higher than Pervasive Developmental Disorder Not Otherwise Specified. The ADI-R scores of communication domain, repetitive and stereotyped behavior domain were decreased at school age, while reciprocal social interaction domain was also declined but no significant difference. All domains of VABS-II standard score were no significant difference between ASDs and Non-ASDs at two time point, and the score of social domain was worst in both two groups. The VABS-II standard score of social domain was reduced although the age equivalent was increased at school age, and the standard score of other domains were no significant difference between the two time points but all age equivalents were improved at school age. Dividing ASDs to high-function ASDs and low-function ASDs to analyze the growth rate of adaption behaviors, the results indicated that the growth rate of adaption behaviors were higher in high-function ASDs at daily skill domain and social domain. Children with ASDs had significant higher scores on Withdrawn/Depressed scale and Social Problem scale than Non-ASDs in CBCL, and over 50 percent of ASDs scored above the borderline range at Withdrawn/Depressed scale, Social Problem scale and Attention Problem scale. In addition, there were no significant correlations between CBCL and ADI-R. Conclusions: In summary, the study showed that most of ASDs sustained the clinical diagnosis from preschool age to school age, however, the autistic symptom severity was declined. Additionally, the improvement of symptom severity of individual ASDs was high variability. The social adaptive behavior was worst in both toddler age and school age in ASDs. Although these children with ASD developed some of adaptive behaviors in the school age, the magnitude of improvement couldn’t equal to the change of their chronological age, especially in socialization domain. The growth rate of daily skill domain and social domain were higher in high-function ASDs than low-function ASDs, might associated with cognitive function. More Children with ASDs had withdrawn/depressed problem, social problem and attention problem, but there were not related to autistic symptom severity.

公司法之臨時管理人制度係為因應董事會無法行使職權，將致公司有受損害之虞，考量股東權益之保障以及交易秩序之維護，由司法機關介入，為公司選任臨時管理人，以暫時代行董事會及董事長職權。然在實務運作面上，臨時管理人除了作為董事缺位、公司內部因股權結構或董事派別因素所致無法決定事務等困境之解決手段外，尚有藉由選任臨時管理人以遂行法律或訴訟程序，或在公司經營權爭奪之時作為穩定局勢之用，亦有將之視為拯救董事會嚴重瀆職之方式。另一方面，在構成要件之解釋上，則多以該規定之立法說明為依歸。惟細究臨時管理人制度，其本質為司法介入公司自治領域，故在適用前提上應更為審慎。 本文以美國模式作為我國選任臨時管理人之實質要件與權限責任的參考對象，將股東自治可能性視為現行法下是否適用臨時管理人制度之重要前提。具體類型則包括股東僵局、董事僵局以及董事遭禁止行使職權之定暫時狀態假處分等「董事會不能行使職權」之事由；至於「董事會不為行使職權」之事由，則需考量我國臨時管理人之制度功能有無包括債務清理、重建功能，以及選任臨時管理人之同時是否將與現任董事間造成權限重疊問題。此外，就「致公司有受損害之虞」之要件，本文嘗試提出幾項判準茲以適用。至於臨時管理人之權限，基於其暫時性代行職權之性質，應作目的性限縮解釋，相對地，就其義務及責任內容亦應有所調整。 末以，我國公司法臨時管理人制度之規範不足，相關之修正聲浪未曾停歇。本文建議加入股東自治可能之選任要件，並賦予法院裁定准許臨時管理人採取打破公司僵局或防止公司受損等必要行為之權限，同時將臨時管理人之地位明文化，並規範其對公司、第三人所應負之責任程度。此外，尚將其報酬請求權之基礎及決定方式一併明定，最後，關於解任臨時管理人之程序及效力，則依董事會是否得以行使職權分別提出立法建議。

甲基安非他命是一種被廣泛濫用的非法神經興奮劑，而且使用之後常伴隨著精神疾病的發生，動物研究也顯示，施打甲基安非他命所引起的神經毒性不僅會造成多巴胺神經元及血清素神經元的損傷，也引起認知功能和社交行為的缺失，同時對於產生迷幻作用的5-HT2A受體作用劑的行為反應增強。N,N,N-trimethylglycine (TMG)和N,N-dimethylglycine (DMG)是甘胺酸的甲基化衍生物，由於這兩種藥物具有治療神經系統疾病的潛力，因此本研究的目的為評估TMG及DMG是否可以預防或改善小鼠在甲基安非他命的暴露下所導致的行為缺失包括新位置辨識測試，新物體辨識測試，社交行為互動測試以及使用5-HT2A受體作用劑DOI 誘導小鼠頭部抽搐(head twitch )的行為。實驗方式為腹腔注射給予雄性ICR小鼠甲基安非他命，一天注射四劑（4 × 5mg/kg），每劑間隔兩小時。實驗一，小鼠在暴露甲基安非他命，先確認行為改變後，給予腹腔注射TMG及DMG (10或30 mg/kg)連續七天，評估TMG及DMG的治療效果。實驗二在施打每劑甲基安非他命30分鐘前給予TMG及DMG (100 mg/kg)，七天後進行行為評估，實驗三，評估TMG及DMG個別及混合劑量的治療效果，小鼠給予甲基安非他命之後，先確認行為改變，再給予腹腔注射TMG及DMG (20、5+5或是10+10 mg/kg) 連續七天，七天後進行行為測試。實驗四，檢測TMG及DMG的治療效果是否藉由活化NMDA受體glycine binding site，小鼠給予甲基安非他命七天之後，腹腔注射TMG及DMG (20 mg/kg)並在給予TMG及DMG前30分鐘給予glycine binding site 拮抗劑7-chlorokynurenic acid (7-CK) (1 mg/kg)，連續給藥七天，七天後進行行為評估。實驗結果發現連續給予七天TMG及DMG在個別劑量及混合劑量中都能夠恢復甲基安非他命所造成的認知功能缺損，社交退縮和降低由DOI 誘導小鼠頭部抽搐行為表現，以及在紋狀體中酪氨酸羥化酶的蛋白質表達減少情況。而前給予7-CK則阻斷TMG及DMG對甲基安非他命所造成的認知功能缺損，社交退縮的改善作用，但是對TMG及DMG對DOI 誘導小鼠頭部抽搐的行為的改善作用影響較小，顯示TMG及DMG可能都是經由活化NMDA 受體的glycine binding site改善甲基安非他命所造成的認知功能缺損，社交退縮，這些發現表示，TMG及DMG具有治療甲基安非他命成癮者所造成的精神分裂等異常症狀的潛力。 / Methamphetamine (METH) is a widely abused illicit psychostimulant. METH use is commonly associated with psychosis. A neurotoxic regimen of METH, which damages the dopaminergic and serotonergic neurons, causes cognitive dysfunction, social interaction deficits, and supersensitivity to hallucinogen in mice. N,N,N-trimethylglycine (TMG) and N,N-dimethylglycine (DMG) are methyl derivatives of amino acid glycine and naturally occur as intermediate metabolites in choline-to-glycine metabolism. Growing evidence shows that both compounds have potential to treat some neurological disorders. The aim of this study was to examine the protective and therapeutic effects of TMG and DMG on METH-induced behavioral aberrations. The novel location recognition test (NLRT), the novel objective recognition test (NORT), the social interaction and the hallucinogenic 2, 5-dimethoxy-4-iodoamphetamine (DOI)-induced head twitch response were evaluated. Male ICR mice received one day drug treatment with four injections of METH (4 × 5 mg/kg, i.p.) or saline at 2h interval. First, TMG or DMG (10 or 30 mg/kg, i.p.) were separately administered once daily for seven consecutive days after the behavioral impairment was confirmed in METH-treated mice. Seven days after final injection of TMG and DMG, the behavioral tests were monitored. Secondly, the preveting effects of TMG and DMG were examined by TMG and DMG (100 mg/kg, i.p.) pretreatment, 30 min prior to each dose of METH. Third, the lower dose (20 mg/kg) and combined effects of TMG and DMG (5+5 or 10+10 mg/kg i.p.) were evaluated. Fourth, in order to determine if the improving effects of TMG and DMG are mediated by NMDA receptor glycine binding site, the glycine binding site antagonist 7-CK (1 mg/kg, i.p.) was administered 30 min prior to each dose of TMG and DMG (20 mg/kg, i.p.), TMG and DMG dose-dependently improved, but not prevented the METH-induced cognition deficits, social withdrawal and hypersensitivity to hallucinogen with additional effect. Pretreatment of 7-CK, reversed the improving effects of TMG and DMG on behavioral deficits after METH exposure, yet had minor effect on hypersensitivity to hallucinogen. These results demonstrate that TMG and DMG might activate the glycine binding site of NMDA receptor to improve METH-induced cognition deficits and social withdrawal. TMG and DMG may be the novel therapeutic agents for psychiatric disorders related to METH abuse.

我國於97年7月1日設立智慧財產法院，專責辦理智慧財產案件，智慧財產民事保全程序之假扣押、假處分及定暫時狀態處分亦包括在內，為瞭解智慧財產法院對於假扣押、假處分及定暫時狀態處分案件所採取之見解，本文蒐集智慧財產法院自97年7月1日設立後至103年4月30日止之假扣押、假處分及定暫時狀態處分裁定，將相關裁定之准許理由、駁回理由加以歸納分類，進行系統化之整理，依不同類型之權利如專利權、商標權、著作權等分類，予以統計分析，並提出問題與建議。

本研究自語言的想像和傳承起興，探討新聞報導時間的變化。主採敘事學觀點，輔以語言學相關概念綜論新聞文本時間敘事的特色，並進而論及其時間想像及歷史感。 文本取材自人民日報和聯合報共十一則報導，逐篇進行言說模式及窗口分析，其結果說明新聞報導的歷史感非只憑敘述過去事件，更須著重過去和現在素材的連接，不同的連接形式，即敘事時間次序、頻率、節奏的變化各有其意義，並以指示語及各種資訊狀態縫合文本時間的跳躍和變動。而文本中的可然世界(作者或故事人物的精神活動之敘述)則在時間交錯處編織出可供讀者掌握過去、現在、未來時間的各種想像。 / This study, originated from imagination and inheritance of language, aims at examining features of narrative time. Taken a narratologic perspective with concepts borrowed from linguistics, this research discusses the narrativity in news text and also further explores temporarity and sense of history in news. A total of 11 news stories selected from the People’s Daily (Beijing) and the United Daily News (Taipei) were then analyzed by using the methods of discourse analysis and window analysis respectively. The results showed that the sense of history in news was more than simple description of the past events. In fact, it (the sense of history) could only be emphasized by connecting materials of both the past and the present. Further, with different forms of time connection, such as frequency, order, and tempo, news texts could be understood with the use of deixis and information status despite time fluctuation and change. Finally the possible world delineated in any news text would construct the readers’ imagination and allow them to hold the past, the present and the future all together. Keywords: time imagination, narrative theory, deixis, information status, discursive structure, here and now

本文旨在探討我國上市公司除了盈餘外，資本支出是否具有增額資訊內涵。研究內容分為兩個部份，第一部份在使用狀態空間樣型探討資本支出是否有助於預測未來盈餘；第二部份根據本文推導之模式探討資本支出資訊與股標報酬之關聯。〔分為年度研究與長期研究（兩年及三年）兩個部份〕。研究結果發現： 1.除了盈餘資訊外，資本支出無法幫助吾人預測未來盈餘。 2.無論係長期或年度研究，均呈現股價領先財務報表期間現象。在長期研究中，本文發現大公司之股價有post announcement drift現象。 3.盈餘、資本支出資訊（或財務報表資訊）對股價之解釋能力視股市係屬多空頭市場而定。多頭市場解釋能力較高。 4.長期研究財務報表對股價解釋能力高於年度研究。 5.盈餘水準、資本支出水準均具增額資訊內涵，不論長期或年度研究。 6.非預期土地投資與非預期廠房設備投資對股價均具解釋能力。 7.未預期資本支出反應係數之影響因素： (1)就成長機會而言： 股票市價╱權益比愈大之公司，營收成長率愈高之公司未預期資本支出反應係數愈大。 (2)就系統風險而言： β值愈大之公司，未預期資本支出反應係數愈大。 (3)就資本支出報酬率而言： 盈餘水準、盈餘持續度愈大之公司，未預期資本支出反應係數愈大；而自有資金比率愈低之公司，未預期資本支出反應係數愈大；小公司之資本支出反應係數較大；研發水準愈高之公司，未預期資本支出反應係數愈大。 (4)就資本支出受益年限而言： 本文以產業進入障礙為資本支出受益年限之替代變數。研究結果發現各行業之資本支出反應係數與產業進入障礙正相關。 (5)資本支出型態 規模成長型公司之資本支出反應係數大於汰舊換新型公司。 / This study aims to examine the incremental information content of capital expenditures of Taiwan listed companies. Taiwan listed companies generally have intensive capital expenditure rather then research and development costs in order to sustain the growth of their performance. Thus, this study suspects that the level of capital expenditures could help predict future earnings upon which capital expenditure could incrementally explain the earnings/return relationship. Empirically, this study first investigates the relationship between current capital expenditure and future earnings. Second, in order to select the optimal earnings/return windows, this study simulates the returns window for large and small firms over various long windows. Third, this study extends Collins and Kothari (1989) and Feltham and Ohlson (1995) to investigate whether the capital expenditure would contain an incremental information content in terms of earnings/return relationship. The findings of this study can be summarized as follows. 1.Besides earnings itself, the capital expenditure cannot well predict future earnings. 2.No matter what is temporal or cross sectional study, price leads the realization of earnings. In addition, the large firm sample group demonstrates the phenomenon of post-earnings drift. 3.The capital expenditure has more explanatory power to earnings/return relationship in the bull market than in the bear market. 4.Earnings and capital expenditure level have incremental information contents in terms of earnings/return relationship. 5.Both unexpected property and unexpected plant investments have explanatory power to the stock price. 6.The determinants of capital expenditure response coefficient, including growth opportunity, systematic risk, returns on capital expenditure, beneficial period of capital expenditure, and types of capital expenditure can increase the explanatory power of earnings/return relationship.

本研究採用門檻迴歸模型 (Threshold Autoregression Model)，試圖找出董監事等內部人之申報轉讓比率、持股比率及質押比率等門檻值，進而分析門檻值以內及以外，指標對於代理變數：融資成長率、營收成長率以及本益比與加權指數報酬率的影響程度與方向。本研究實證結果發現： 一、在申報轉讓比率方面： 當申報轉讓比率低於門檻值，存在所謂的群聚效果。當申報轉讓比率高於門檻值時，市場動能與加權指數報酬率無顯著關係，投資人於此階段進行投資決策時應該要謹慎小心。 二、在持股比率方面： 在持股比率低於門檻值時，加權指數報酬率對於前期營收成長率表現的修正幅度較大，意謂著董監事等內部人根據其對未來營收資訊掌握的優勢，反應其對營收資訊的真實性，而藉由持股轉讓的行為，使加權指數大幅度的修正。 三、在質押比率方面： 不管高於或低於門檻值，均無法利用董監事等內部人質押比率為門檻變數來分析本益比效果對加權指數報酬率的影響。造成其檢定失效的原因，可能是樣本小且模型受到極端值的影響所造成。