應(yīng)用回歸分析第三版·何曉群-第三章所有習(xí)題答案WORD

1、文檔可能無(wú)法思考全面，請(qǐng)瀏覽后下載! 應(yīng)用回歸分析第三章習(xí)題3.1基本假定：（1） 諸非隨機(jī)變量，rank（x）=p+1，X為滿(mǎn)秩矩陣（2） 誤差項(xiàng)（3）3.23.33.4并不能這樣武斷地下結(jié)論。與回歸方程中的自變量數(shù)目以及樣本量n有關(guān)，當(dāng)樣本量n與自變量個(gè)數(shù)接近時(shí)，易接近1，其中隱含著一些虛假成分。因此，并不能僅憑很大的就模型的優(yōu)劣程度。3.5首先，對(duì)回歸方程的顯著性進(jìn)行整體上的檢驗(yàn)F檢驗(yàn)7 / 8接受原假設(shè)：在顯著水平下，表示隨機(jī)變量y與諸x之間的關(guān)系由線(xiàn)性模型表示不合適拒絕原假設(shè)：認(rèn)為在顯著性水平下，y與諸x之間有顯著的線(xiàn)性關(guān)系第二，對(duì)單個(gè)自變量的回歸系數(shù)進(jìn)行顯著性檢驗(yàn)。接受原假設(shè)：認(rèn)為

2、=0，自變量對(duì)y的線(xiàn)性效果并不顯著3.6原始數(shù)據(jù)由于自變量的單位往往不同，會(huì)給分析帶來(lái)一定的困難；又由于設(shè)計(jì)的數(shù)據(jù)量較大，可能會(huì)以為舍入誤差而使得計(jì)算結(jié)果并不理想。中心化和標(biāo)準(zhǔn)化回歸系數(shù)有利于消除由于量綱不同、數(shù)量級(jí)不同帶來(lái)的影響，避免不必要的誤差。3.73.83.9由上兩式可知，其考慮的都是通過(guò)在總體中所占比例來(lái)衡量第j個(gè)因素的重要程度，因而與是等價(jià)的。3.10【沒(méi)整出來(lái)】3.11（1）計(jì)算可知，y與x1 x2 x3 的相關(guān)關(guān)系是：Correlations貨運(yùn)總量y工業(yè)總產(chǎn)值x1農(nóng)業(yè)總產(chǎn)值x2居民非商品支出x3貨運(yùn)總量yPearson Correlation1.556.731*.724*Si

3、g. (2-tailed).095.016.018N10101010工業(yè)總產(chǎn)值x1Pearson Correlation.5561.113.398Sig. (2-tailed).095.756.254N10101010農(nóng)業(yè)總產(chǎn)值x2Pearson Correlation.731*.1131.547Sig. (2-tailed).016.756.101N10101010Correlations貨運(yùn)總量y工業(yè)總產(chǎn)值x1農(nóng)業(yè)總產(chǎn)值x2居民非商品支出x3貨運(yùn)總量yPearson Correlation1.556.731*.724*Sig. (2-tailed).095.016.018N10101010

4、工業(yè)總產(chǎn)值x1Pearson Correlation.5561.113.398Sig. (2-tailed).095.756.254N10101010農(nóng)業(yè)總產(chǎn)值x2Pearson Correlation.731*.1131.547Sig. (2-tailed).016.756.101N10101010居民非商品支出x3Pearson Correlation.724*.398.5471Sig. (2-tailed).018.254.101N10101010居民非商品支出x3Pearson Correlation.724*.398.5471Sig. (2-tailed).018.254.101N1

5、0101010*. Correlation is significant at the 0.05 level (2-tailed).則相關(guān)關(guān)系矩陣如下：（2）CoefficientsaModelUnstandardized CoefficientsStandardized CoefficientstSig.BStd. ErrorBeta1(Constant)-348.280176.459-1.974.096工業(yè)總產(chǎn)值x13.7541.933.3851.942.100農(nóng)業(yè)總產(chǎn)值x27.1012.880.5352.465.049居民非商品支出x312.44710.569.2771.178.284a

6、. Dependent Variable: 貨運(yùn)總量y（3）擬合優(yōu)度檢驗(yàn)Model SummarybModelRR SquareAdjusted R SquareStd. Error of the EstimateChange StatisticsDurbin-WatsonR Square ChangeF Changedf1df2Sig. F Change1.898a.806.70823.442.8068.28336.0151.935a. Predictors: (Constant), 居民非商品支出x3, 工業(yè)總產(chǎn)值x1, 農(nóng)業(yè)總產(chǎn)值x2b. Dependent Variable: 貨運(yùn)總量

7、y決定系數(shù)R2=0.708 R=0.898較大所以認(rèn)為擬合度較高（4）對(duì)回歸方正作整體顯著性檢驗(yàn)ANOVAbModelSum of SquaresdfMean SquareFSig.1Regression13655.37034551.7908.283.015aResidual3297.1306549.522Total16952.5009ANOVAbModelSum of SquaresdfMean SquareFSig.1Regression13655.37034551.7908.283.015aResidual3297.1306549.522Total16952.5009a. Predict

8、ors: (Constant), 居民非商品支出x3, 工業(yè)總產(chǎn)值x1, 農(nóng)業(yè)總產(chǎn)值x2b. Dependent Variable: 貨運(yùn)總量yF=8.283 取=0.05時(shí) P=0.015<0.05所以認(rèn)為回歸方程在整體上擬合的好（5）對(duì)每個(gè)回歸系數(shù)作顯著性檢驗(yàn)CoefficientsaModelUnstandardized CoefficientsStandardized CoefficientstSig.BStd. ErrorBeta1(Constant)-348.280176.459-1.974.096工業(yè)總產(chǎn)值x13.7541.933.3851.942.100農(nóng)業(yè)總產(chǎn)值x27.

9、1012.880.5352.465.049居民非商品支出x312.44710.569.2771.178.284a. Dependent Variable: 貨運(yùn)總量y=0.05時(shí)，x3并未通過(guò)顯著性檢驗(yàn)（6）將x3剔除后，重新建立回歸方程并做回歸方程的顯著性檢驗(yàn)：CoefficientsaModelUnstandardized CoefficientsStandardized CoefficientstSig.BStd. ErrorBeta1(Constant)-459.624153.058-3.003.020工業(yè)總產(chǎn)值x14.6761.816.4792.575.037農(nóng)業(yè)總產(chǎn)值x28.971

10、2.468.6763.634.008a. Dependent Variable: 貨運(yùn)總量yCoefficientsaModelUnstandardized CoefficientsStandardized CoefficientstSig.BStd. ErrorBeta1(Constant)-459.624153.058-3.003.020工業(yè)總產(chǎn)值x14.6761.816.4792.575.037農(nóng)業(yè)總產(chǎn)值x28.9712.468.6763.634.008ANOVAbModelSum of SquaresdfMean SquareFSig.1Regression12893.19926446

11、.60011.117.007aResidual4059.3017579.900Total16952.5009a. Predictors: (Constant), 農(nóng)業(yè)總產(chǎn)值x2, 工業(yè)總產(chǎn)值x1b. Dependent Variable: 貨運(yùn)總量y由上兩表可知，回歸方程總體上，并且每一個(gè)回歸系數(shù)均通過(guò)了顯著性檢驗(yàn)（7）CoefficientsaModelUnstandardized CoefficientsStandardized CoefficientstSig.95.0% Confidence Interval for BBStd. ErrorBetaLower BoundUpper Bound1(Constant)-348.280176.459-1.974.096-780.06083.500工業(yè)總產(chǎn)值x13.7541.933.3851.942.100-.9778.485農(nóng)業(yè)總產(chǎn)值x27.1012.880.5352.465.049.05314.149居民非商品支出x312.44710.569.2771.178.284-13.41538.310a. Dependent Variable: 貨運(yùn)總量yx1:(-0.997,8.485) x2:(0.053,14.149) x3:(-13.415,38.310)（8）（9）（175.4748，292.5545）（10）由于