109 年 國立暨南國際大學財務金融學系碩士班《統計學》

📄 試題原卷 免費註冊後即可對照原始考卷 PDF免費註冊

第 1 題

Approximate the following binomial probabilities by the use of normal approximation. Twenty percent of students who finish high school do not go to college. What is the probability that in a sample of 80 high school students
(a) exactly 10 will not go to college? (5%)
(b) 70 or more will go to college? (5%)

登入後即可作答並保存紀錄。

這一題的完整詳解

此題考驗對二項分配進行常態近似的應用。當樣本數 nn 夠大,且 npnp 與 n(1−p)n(1-p) 都大於某個常數(通常是 5 或 10)時,二項分配可以使用常態分配來近似。

給定條件:
樣本數 n=80n = 80
學生不升學的機率 p=0.20p = 0.20
則學生升學的機率 1−p=0.801-p = 0.80

二項分配的參數:
平均數 μ=np=80×0.20=16\mu = np = 80 \times 0.20 = 16
標準差 σ=np(1−p)=80×0.20×0.80=12.8≈3.5777\sigma = \sqrt{np(1-p)} = \sqrt{80 \times 0.20 \times 0.80} = \sqrt{12.8} \approx 3.5777

由於 np=16>10np = 16 > 10 且 n(1−p)=80×0.80=64>10n(1-p) = 80 \times 0.80 = 64 > 10,我們可以使用常態近似。

(a) 計算「恰好 10 位學生不升學」的機率。
使用常態近似時,需要進行連續性修正 (continuity correction)。對於「恰好 kk」的事件,我們將其範圍視為 [k−0.5,k+0.5][k-0.5, k+0.5]。
因此,我們要計算的機率為 P(9.5≤X≤10.5)P(9.5 \le X \le 10.5),其中 XX 服從近似的常態分配 N(16,12.8)N(16, 12.8)。
將 9.59.5 和 10.510.5 轉換為標準常態分數 zz:
z1=9.5−μσ=9.5−163.5777≈−6.53.5777≈−1.8168z_1 = \frac{9.5 - \mu}{\sigma} = \frac{9.5 - 16}{3.5777} \approx \frac{-6.5}{3.5777} \approx -1.8168

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 2 題

Winter visitors are extremely important to the economy of Southwest Florida. Hotel occupancy is an often-reported measure of visitor volume and visitor activity. Hotel occupancy data for February in two consecutive years are as follows.

Current YearPrevious Year
Occupied Rooms14701458
Total Rooms17501800

(a) Formulate the hypothesis test that can be used to determine if there has been an increase in the proportion of rooms occupied over the one-year period. Using a 0.05 level of significance, what is your hypothesis test conclusion? (10%)
(b) What is the 95% confidence interval estimate of the change in occupancy for the one-year period? Do you think area officials would be pleased with the results? (10%)

登入後即可作答並保存紀錄。

這一題的完整詳解

此題考驗對比例檢定與比例信賴區間的應用。我們需要比較兩個不同時期的旅館入住率,並進行假設檢定與信賴區間估計。

首先,計算兩個時期的入住率(樣本比例):
目前年度入住率:p^1=14701750=0.84\hat{p}_1 = \frac{1470}{1750} = 0.84
前一年度入住率:p^2=14581800=0.81\hat{p}_2 = \frac{1458}{1800} = 0.81

(a) 比例檢定
我們想檢定的是「目前年度的入住率是否高於前一年度」。

  1. 虛無假設 (H0H_0) 與對立假設 (H1H_1)
    令 p1p_1 為目前年度的真實入住率,令 p2p_2 為前一年度的真實入住率。
    H0:p1≤p2H_0: p_1 \le p_2 (入住率沒有增加,或下降)
    H1:p1>p2H_1: p_1 > p_2 (入住率有所增加)

  2. 顯著水準
    α=0.05\alpha = 0.05

  3. 檢定統計量
    由於我們要檢定兩個獨立樣本比例的差異,且樣本數足夠大(可透過檢查 np^n\hat{p} 和 n(1−p^)n(1-\hat{p}) 的值來判斷,在此處 n1p^1=1470>10n_1\hat{p}_1 = 1470 > 10, n1(1−p^1)=280>10n_1(1-\hat{p}_1) = 280 > 10, n2p^2=1458>10n_2\hat{p}_2 = 1458 > 10, n2(1−p^2)=342>10n_2(1-\hat{p}_2) = 342 > 10),我們可以使用 zz-檢定。
    在虛無假設為真時,我們假設兩個比例相等,因此使用合併比例 p^pool\hat{p}_{pool} 來計算標準誤。
    合併樣本比例:p^pool=n1p^1+n2p^2n1+n2=1470+14581750+1800=29283550≈0.8248\hat{p}_{pool} = \frac{n_1\hat{p}_1 + n_2\hat{p}_2}{n_1 + n_2} = \frac{1470 + 1458}{1750 + 1800} = \frac{2928}{3550} \approx 0.8248
    標準誤:SEpool=p^pool(1−p^pool)(1n1+1n2)=0.8248(1−0.8248)(11750+11800)SE_{pool} = \sqrt{\hat{p}_{pool}(1-\hat{p}_{pool})(\frac{1}{n_1} + \frac{1}{n_2})} = \sqrt{0.8248(1-0.8248)(\frac{1}{1750} + \frac{1}{1800})}
    SEpool=0.8248×0.1752×(11750+11800)≈0.1444×(0.0005714+0.0005556)≈0.1444×0.001127≈0.0001627≈0.012756SE_{pool} = \sqrt{0.8248 \times 0.1752 \times (\frac{1}{1750} + \frac{1}{1800})} \approx \sqrt{0.1444 \times (0.0005714 + 0.0005556)} \approx \sqrt{0.1444 \times 0.001127} \approx \sqrt{0.0001627} \approx 0.012756
    檢定統計量:z=p^1−p^2SEpool=0.84−0.810.012756=0.030.012756≈2.3518z = \frac{\hat{p}_1 - \hat{p}_2}{SE_{pool}} = \frac{0.84 - 0.81}{0.012756} = \frac{0.03}{0.012756} \approx 2.3518

  4. 臨界值與決策規則
    由於是對立假設 H1:p1>p2H_1: p_1 > p_2 進行右尾檢定,在 α=0.05\alpha = 0.05 的顯著水準下,臨界值 zαz_{\alpha} 為

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 3 題

Two Wharton professors analyzed 1,613,234 putts by golfers on the Professional Golfers Association (PGA) Tour and found that 983,764 of the putts were made and 629,470 of the putts were missed. Further analysis showed that for putts that were made, 64.0% of the time the player was attempting to make a par putt and 18.8% of the time the player was attempting to make a birdie putt. And, for putts that were missed, 20.3% of the time the player was attempting to make a par putt and 73.4% of the time the player was attempting to make a birdie putt.

(a) What is the probability that a PGA Tour player makes a putt? (5%)
(b) Suppose that a PGA Tour player has a putt for par. What is the probability that the player will make the putt? (5%)
(c) Suppose that a PGA Tour player has a putt for birdie. What is the probability that the player will make the putt? (5%)

登入後即可作答並保存紀錄。

這一題的完整詳解

此題考驗對條件機率與全機率公式的應用。題目提供了關於 PGA 選手推桿的總體數據,並要求計算不同情境下的機率。

我們定義以下事件:
MM: 推桿成功 (Made)
MissMiss: 推桿失敗 (Missed)
ParPar: 嘗試推 Par 推桿 (Attempting Par putt)
BirdieBirdie: 嘗試推 Birdie 推桿 (Attempting Birdie putt)

給定數據:
總推桿數 (Total putts) = 1,613,234
成功推桿數 (Made putts) = 983,764
失敗推桿數 (Missed putts) = 629,470

注意:983,764 + 629,470 = 1,613,234,數據吻合。

另外,題目給了推桿結果(成功/失敗)與推桿目標(Par/Birdie)之間的條件機率。
對於成功的推桿:
P(Par∣M)=0.640P(Par | M) = 0.640
P(Birdie∣M)=0.188P(Birdie | M) = 0.188
注意:0.640+0.188=0.8280.640 + 0.188 = 0.828。這表示還有 1−0.828=0.1721 - 0.828 = 0.172 的成功推桿是嘗試其他目標(例如 Bogey 或 Chip-in 等),或者題目只關注這兩種情況。我們暫時假設這兩種情況是主要的,但要留意此處的機率和不為 1。

對於失敗的推桿:
P(Par∣Miss)=0.203P(Par | Miss) = 0.203
P(Birdie∣Miss)=0.734P(Birdie | Miss) = 0.734
注意:0.203+0.734=0.9370.203 + 0.734 = 0.937。這表示還有 1−0.937=0.0631 - 0.937 = 0.063 的失敗推桿是嘗試其他目標。

(a) 計算 PGA 選手成功推桿的機率。
這是直接從總體數據中可以計算出來的。
P(M)=成功推桿數總推桿數=983,7641,613,234P(M) = \frac{\text{成功推桿數}}{\text{總推桿數}} = \frac{983,764}{1,613,234}
P(M)≈0.6098P(M) \approx 0.6098

(b) 計算 PGA 選手有 Par 推桿時,成功推桿的機率。
這需要用到貝氏定理或全機率公式。我們已知 P(Par∣M)=0.640P(Par | M) = 0.640 和 P(Birdie∣M)=0.188P(Birdie | M) = 0.188。
題目問的是「Suppose that a PGA Tour player has a putt for par. What is the probability that the player will make the putt?」,這實際上是在問 P(M∣Par)P(M | Par)。

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 4 題

The Global Business Travel Association reported the domestic airfare for business travel for the current year and the previous year. Below is a sample of 12 flights with their domestic airfares shown for both years. Use a 0.05 level of significance and test for a difference between the population mean for the current year and the population mean for the previous year. (10%)

Current YearPrevious YearCurrent YearPrevious Year
345315635585
526463710650
420462605545
216206517547
285275570508
405432610580

登入後即可作答並保存紀錄。

這一題的完整詳解

此題考驗獨立樣本 tt-檢定的應用。我們需要比較兩個不同年份的商務旅行機票平均價格,並進行假設檢定。

我們有兩組獨立樣本:今年 (Current Year) 和去年 (Previous Year) 的商務旅行機票價格。

樣本數據:
今年 (Current Year): n1=12n_1 = 12, 數據為 {345, 526, 420, 216, 285, 405, 635, 710, 605, 517, 570, 610}
去年 (Previous Year): n2=12n_2 = 12, 數據為 {315, 463, 462, 206, 275, 432, 585, 650, 545, 547, 508, 580}

我們假設兩組樣本的變異數相等(可以進行 FF-檢定來驗證,但在此處題目未要求,我們暫時假設相等)。

  1. 虛無假設 (H0H_0) 與對立假設 (H1H_1)
    令 μ1\mu_1 為今年商務旅行的平均機票價格,μ2\mu_2 為去年商務旅行的平均機票價格。
    題目要求檢定「是否有差異」,這意味著是雙尾檢定。
    H0:μ1=μ2H_0: \mu_1 = \mu_2 (平均機票價格沒有差異)
    H1:μ1≠μ2H_1: \mu_1 \neq \mu_2 (平均機票價格有差異)

  2. 顯著水準
    α=0.05\alpha = 0.05

  3. 計算樣本平均數與樣本變異數

    今年 (Current Year):
    ∑x1=345+526+420+216+285+405+635+710+605+517+570+610=5844\sum x_1 = 345+526+420+216+285+405+635+710+605+517+570+610 = 5844
    xˉ1=584412=487\bar{x}_1 = \frac{5844}{12} = 487
    計算平方和:∑x12=3452+5262+...+6102=3,435,564\sum x_1^2 = 345^2 + 526^2 + ... + 610^2 = 3,435,564
    s12=∑x12−n1xˉ12n1−1=3435564−12×487211=3435564−12×23716911=3435564−284602811=58953611≈53594.18s_1^2 = \frac{\sum x_1^2 - n_1\bar{x}_1^2}{n_1-1} = \frac{3435564 - 12 \times 487^2}{11} = \frac{3435564 - 12 \times 237169}{11} = \frac{3435564 - 2846028}{11} = \frac{589536}{11} \approx 53594.18

    去年 (Previous Year):
    ∑x2=315+463+462+206+275+432+585+650+545+547+508+580=5568\sum x_2 = 315+463+462+206+275+432+585+650+545+547+508+580 = 5568
    xˉ2=556812=464\bar{x}_2 = \frac{5568}{12} = 464
    計算平方和:∑x22=3152+4632+...+5802=3,147,190\sum x_2^2 = 315^2 + 463^2 + ... + 580^2 = 3,147,190

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 5 題

Multiple regression analysis was used to study the relationship between a dependent variable Y, and three independent variables X1, X2 and, X3. The following is a partial result of the regression analysis involving 20 observations.

CoefficientStandard Error
Intercept20.005.00
X1X_115.003.00
X2X_28.005.00
X3X_3-18.000.00

Analysis of Variance

SourceDFSSMSF
Regression??80?
Error?320

(a) Fill in all the blanks in the above ANOVA table. (4%)
(b) Compute the coefficient of determination. (5%)
(c) Perform a tt test and determine whether or not β1,β2,β3\beta_1, \beta_2, \beta_3 are significantly different from zero (α=0.05\alpha=0.05). (6%)
(d) At α=0.05\alpha=0.05, perform an FF test and determine whether or not the regression model is significant. (5%)

登入後即可作答並保存紀錄。

這一題的完整詳解

此題考驗對複迴歸分析結果的判讀,包含 ANOVA 表的填寫、判定係數計算、個別迴歸係數的 tt-檢定,以及整體迴歸模型的 FF-檢定。

基本資訊:
樣本數 n=20n = 20
自變數個數 k=3k = 3 (X1,X2,X3X_1, X_2, X_3)

迴歸係數與標準誤:
β^0=20.00\hat{\beta}_0 = 20.00, SE(β^0)=5.00SE(\hat{\beta}_0) = 5.00
β^1=15.00\hat{\beta}_1 = 15.00, SE(β^1)=3.00SE(\hat{\beta}_1) = 3.00
β^2=8.00\hat{\beta}_2 = 8.00, SE(β^2)=5.00SE(\hat{\beta}_2) = 5.00
β^3=−18.00\hat{\beta}_3 = -18.00, SE(β^3)=0.00SE(\hat{\beta}_3) = 0.00

ANOVA 表:

SourceDFSSMSF
RegressionkkSSRSSRMSRMSRFcalcF_{calc}
Errorn−k−1n-k-1SSESSEMSEMSE
Totaln−1n-1SSTSST

(a) 填寫 ANOVA 表的空格

  • Regression DF: 自變數的個數 k=3k = 3。
  • Error DF: n−k−1=20−3−1=16n - k - 1 = 20 - 3 - 1 = 16。
  • Regression SS (SSRSSR): 已知 MSR=80MSR = 80 且 DFRegression=3DF_{Regression} = 3。MSR=SSRDFRegressionMSR = \frac{SSR}{DF_{Regression}},所以 SSR=MSR×DFRegression=80×3=240SSR = MSR \times DF_{Regression} = 80 \times 3 = 240。
  • Error SS (SSESSE): 已知 SSE=320SSE = 320。
  • Total SS (SSTSST): SST=SSR+SSE=240+320=560SST = SSR + SSE = 240 + 320 = 560。
  • MSE: MSE=SSEDFError=32016=20MSE = \frac{SSE}{DF_{Error}} = \frac{320}{16} = 20。
  • F-statistic (FcalcF_{calc}): Fcalc=MSRMSE=8020=4F_{calc} = \frac{MSR}{MSE} = \frac{80}{20} = 4。

填寫後的 ANOVA 表:

SourceDFSSMSF
Regression3240804
Error1632020
Total19560

(b) 計算判定係數 (R2R^2)
判定係數 R2R^2 定義為 R2=SSRSSTR^2 = \frac{SSR}{SST} 或 R2=1−SSESSTR^2 = 1 - \frac{SSE}{SST}。
使用已填寫的 ANOVA 表數據:
R2=SSRSST=240560=2456=37≈0.4286R^2 = \frac{SSR}{SST} = \frac{240}{560} = \frac{24}{56} = \frac{3}{7} \approx 0.4286

(c) tt-檢定個別迴歸係數
對於每個迴歸係數 β^i\hat{\beta}_i,我們進行檢定 H0:βi=0H_0: \beta_i = 0 vs H1:βi≠0H_1: \beta_i \neq 0。檢定統計量為 t=β^iSE(β^i)t = \frac{\hat{\beta}_i}{SE(\hat{\beta}_i)}。自由度為 Error DF,即 16。顯著水準 α=0.05\alpha = 0.05。

  • 對於 β1\beta_1:
    t1=β^1SE(β^1)=15.003.00=5.00t_1 = \frac{\hat{\beta}_1}{SE(\hat{\beta}_1)} = \frac{15.00}{3.00} = 5.00
    臨界值 tα/2,df=t0.025,16≈2.120t_{\alpha/2, df} = t_{0.025, 16} \approx 2.120。
    因為 ∣t1∣=5.00>2.120|t_1| = 5.00 > 2.120,所以拒絕 H0H_0。β1\beta_1 顯著異於零。

  • 對於 β2\beta_2:

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 6 題

According to Thomson Financial, through January 25, 2006, the majority of companies reported profits had beaten estimates. A sample of 162 companies showed 104 beat estimates, 29 matched estimates, and 29 fell short.

(a) Determine the margin of error and provide a 95% confidence interval for the proportion that beat estimates. (5%)
(b) How large a sample is needed if the desired margin of error is 0.05? Use α=0.05\alpha = 0.05. (5%)

登入後即可作答並保存紀錄。

這一題的完整詳解

此題考驗對比例的信賴區間估計與樣本數需求的計算。

給定數據:
總樣本數 n=162n = 162
成功(擊敗預期)的公司數 x=104x = 104
匹配預期的公司數 = 29
未達預期的公司數 = 29

(a) 計算 95% 信賴區間與誤差範圍
我們要估計的是「擊敗預期的公司比例」(pp)。
樣本比例 p^=xn=104162≈0.641975\hat{p} = \frac{x}{n} = \frac{104}{162} \approx 0.641975

對於 95% 信賴區間,α=0.05\alpha = 0.05,α/2=0.025\alpha/2 = 0.025。
查標準常態分配表,對應的 zz 值為 z0.025≈1.96z_{0.025} \approx 1.96。

信賴區間的公式為:p^±zα/2p^(1−p^)n\hat{p} \pm z_{\alpha/2} \sqrt{\frac{\hat{p}(1-\hat{p})}{n}}
誤差範圍 (Margin of Error, MOE) =zα/2p^(1−p^)n= z_{\alpha/2} \sqrt{\frac{\hat{p}(1-\hat{p})}{n}}

計算標準誤:
1−p^=1−0.641975=0.3580251 - \hat{p} = 1 - 0.641975 = 0.358025
p^(1−p^)=0.641975×0.358025≈0.22986\hat{p}(1-\hat{p}) = 0.641975 \times 0.358025 \approx 0.22986
p^(1−p^)n=0.22986162≈0.0014189\frac{\hat{p}(1-\hat{p})}{n} = \frac{0.22986}{162} \approx 0.0014189
p^(1−p^)n≈0.0014189≈0.037668\sqrt{\frac{\hat{p}(1-\hat{p})}{n}} \approx \sqrt{0.0014189} \approx 0.037668

計算誤差範圍:
MOE =1.96×0.037668≈0.07383= 1.96 \times 0.037668 \approx 0.07383

計算 95% 信賴區間:
信賴區間 =p^±MOE=0.641975±0.07383= \hat{p} \pm MOE = 0.641975 \pm 0.07383
信賴區間 =[0.641975−0.07383,0.641975+0.07383]= [0.641975 - 0.07383, 0.641975 + 0.07383]

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 7 題

Two different designs on a new line of winter jackets for the coming winter are available for your manufacturing plants. Your profit (in thousands of dollars) will depend on the taste of the consumers when winter arrives. The probability of the three possible different tastes of the consumers and the corresponding profits are presented in the following table.

ProbabilityTasteDesign ADesign B
0.2more180520
0.5conservative230310
0.3more liberal350270

(a) What are the expected value and standard deviation for profit when Design A is chosen? (5%)
(b) What are the expected value and standard deviation for profit when Design B is chosen? (5%)
(c) Compute the correlation coefficient for the profits of Design A and Design B. (5%)

登入後即可作答並保存紀錄。

這一題的完整詳解

此題考驗對離散隨機變數期望值、變異數(進而計算標準差)的計算,以及相關係數的概念。

我們有兩個設計 (Design A, Design B),其利潤(以千美元為單位)取決於三種可能的消費者品味 (Taste)。三種品味的機率是獨立於設計的。

令 XAX_A 為選擇 Design A 時的利潤,令 XBX_B 為選擇 Design B 時的利潤。
令 P1=0.2P_1 = 0.2, P2=0.5P_2 = 0.5, P3=0.3P_3 = 0.3 為三種品味的機率。
Let A1=180,A2=230,A3=350A_1 = 180, A_2 = 230, A_3 = 350 be the profits for Design A under tastes 1, 2, 3 respectively.
Let B1=520,B2=310,B3=270B_1 = 520, B_2 = 310, B_3 = 270 be the profits for Design B under tastes 1, 2, 3 respectively.

(a) Design A 的期望值與標準差
期望值 E(XA)E(X_A):
E(XA)=∑i=13Pi×AiE(X_A) = \sum_{i=1}^{3} P_i \times A_i
E(XA)=(0.2×180)+(0.5×230)+(0.3×350)E(X_A) = (0.2 \times 180) + (0.5 \times 230) + (0.3 \times 350)
E(XA)=36+115+105=256E(X_A) = 36 + 115 + 105 = 256

變異數 Var(XA)Var(X_A):
Var(XA)=E(XA2)−[E(XA)]2Var(X_A) = E(X_A^2) - [E(X_A)]^2
首先計算 E(XA2)E(X_A^2):
E(XA2)=∑i=13Pi×Ai2E(X_A^2) = \sum_{i=1}^{3} P_i \times A_i^2
E(XA2)=(0.2×1802)+(0.5×2302)+(0.3×3502)E(X_A^2) = (0.2 \times 180^2) + (0.5 \times 230^2) + (0.3 \times 350^2)
E(XA2)=(0.2×32400)+(0.5×52900)+(0.3×122500)E(X_A^2) = (0.2 \times 32400) + (0.5 \times 52900) + (0.3 \times 122500)
E(XA2)=6480+26450+36750=69680E(X_A^2) = 6480 + 26450 + 36750 = 69680

Var(XA)=69680−(256)2=69680−65536=4144Var(X_A) = 69680 - (256)^2 = 69680 - 65536 = 4144

標準差 σXA\sigma_{X_A}:
σXA=Var(XA)=4144≈64.374\sigma_{X_A} = \sqrt{Var(X_A)} = \sqrt{4144} \approx 64.374

(b) Design B 的期望值與標準差
期望值 E(XB)E(X_B):
E(XB)=∑i=13Pi×BiE(X_B) = \sum_{i=1}^{3} P_i \times B_i
E(XB)=(0.2×520)+(0.5×310)+(0.3×270)E(X_B) = (0.2 \times 520) + (0.5 \times 310) + (0.3 \times 270)
E(XB)=104+155+81=340E(X_B) = 104 + 155 + 81 = 340

變異數 Var(XB)Var(X_B):
Var(XB)=E(XB2)−[E(XB)]2Var(X_B) = E(X_B^2) - [E(X_B)]^2
首先計算 E(XB2)E(X_B^2):
E(XB2)=∑i=13Pi×Bi2E(X_B^2) = \sum_{i=1}^{3} P_i \times B_i^2
E(XB2)=(0.2×5202)+(0.5×3102)+(0.3×2702)E(X_B^2) = (0.2 \times 520^2) + (0.5 \times 310^2) + (0.3 \times 270^2)
E(XB2)=(0.2×270400)+(0.5×96100)+(0.3×72900)E(X_B^2) = (0.2 \times 270400) + (0.5 \times 96100) + (0.3 \times 72900)

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

其他考古題