中文摘要：

被测量的量是连续随机变量时，所得到的测量数据就带有一定程度的误差．当用这些有误差的数据进行经典参数估计时，误差就会传递到参数估计中，从而对估计的有效性产生一定程度的影响．研究这些影响，对如何达到有效估计起到指导性的作用．本文主要研究了在均匀分布下，通过理论计算以及模拟研究了四舍五人数据对期望、方差估计的影响，给出了x＊-为期望肛的无偏估计的条件；利用同样的方法研究出了x—U（0，θ）时，X8（n）是θ的相合估计的条件．

英文摘要：

When the variable being measured is a continuous random variable, the obtained data has a certain degree of error. Using these error data to classical parameter estimation, on this time the error will be passed to the parameter estimation, then impact the effect of the parameter estimation. Researching this effects has a great significance for effective estimation. Under the uniform distribution, using the actual calculation and simulation to study the effect, which comes from using the rounded data to the classically estimate the mean, variance. Also gives the condition about that X＊- is u the unbiased estimation. In addition, under X - U （0, θ）, using the actual calculation and simulation to study that X＊（n） is θ the condition of the consistent estimation.