Spss上机练习题（6）目的：1．多元回归分析中变量的选择2．多重共线性的诊断习题1、为了分析工业增加值率（x1）、总资产贡献率（x2）、资产负债率（x3）、流动资产周转次数（x4）、工业成本费用利润率（x5）、全员劳动生产率（x6）与产品销售率（x7）之间的关系，我们搜集了2001年部分地区全部国有及规模以上非国有工业企业主要经济效益指标资料，如下表（摘自《中国统计年鉴2002》）：地区工业增加值率（%）总资产贡献率（%）资产负债率（%）流动资产周转次数（次/年）工业成本费用利润率（%）全员劳动生产率[元/(人·年)]产品销售率（%）北京25.827.2855.291.484.636953998.17天津24.768.0158.281.686.595898498.02河北33.038.8563.251.795.264557698.05山西35.785.5663.791.073.042798897.06内蒙古37.055.7958.441.332.443777097.93辽宁28.036.1358.691.43.264817097.33吉林31.397.8362.631.384.984730696.94黑龙江51.0618.3459.021.4926.696810497.61上海28.398.8346.461.566.619539398.83江苏25.069.4260.031.993.885717797.06浙江23.7812.0755.312.066.285112796.97安徽31.638.1761.411.533.733854898.16福建29.728.5256.761.914.475406096.99江西30.346.2566.371.311.423062397.6山东30.9311.4560.282.126.615545597.77河南32.528.1365.681.64.083722597.89湖北33.128.1860.721.64.634913897.6湖南33.499.366.551.473.253908198.72广东26.638.756.011.894.496457197.62广西32.357.2261.181.463.783856497.18海南29.757.1259.931.313.865543395.31重庆28.736.6362.721.212.283660697.91四川34.36.9462.811.273.894033098.01贵州33.966.962.80.943.083564995.73云南50.314.7454.141.248.998203898.61西藏53.085.2223.820.7418.883250089.32陕西34.357.166.011.165.193981397.17甘肃31.175.0964.811.130.973448596.99青海374.6269.760.862.735095093.32宁夏30.855.2657.531.191.633766196.78新疆41.6610.3560.181.4612.428619399.08分别用向前、向后和逐步回归法建立多元回归模型，写出回归方程，并对回归方程的显著性进行检验。2.试通过国内总产值、最终消费和储蓄去预测进口总额：下表收集了我国内地15年的相关数据（单位：亿元），试做多重共线性诊断。（数据摘自《中国统计年鉴2002》）。年份国内总产值最终消费储蓄年终余额进口总额19783605.62239.1210.6187.419804551.32976.1399.5298.819858792.157731622.61257.8199018319.511365.27119.82574.3199121280.413145.99241.63398.7199225863.715952.111759.44443.3199334500.720182.115203.55986.2199446690.72679621518.89960.1199558510.53363529662.311048.1199668330.440003.938520.8411557.4199774894.243579.446279.811806.5199879003.346405.953407.4711626.1199910780949722.759621.813736.52000119650.254616.6664332.4186392001132310.958952.673762.420164.2