CAJ | 학술논문

对实验报告的评分需要评价实验数据的准确度和精密度，如若直接用全体平均值代替真实值来评价准确度，往往会带来误差。从全体实验测量值所服从的分布出发，通过对数变换将其转化为偏斜度为零的正态分布，此时的全体平均值就极为接近真实值，也就可依据离全体平均值越近越优的原则评分。对于精密度，则以相对平均偏差越小越优的原则评分。最后等权重考虑准确度和精密度，为每位学生的实验报告打分。具体的实例表明了该方法的有效性。
대실험보고적평분수요평개실험수거적준학도화정밀도，여약직접용전체평균치대체진실치래평개준학도，왕왕회대래오차。종전체실험측량치소복종적분포출발，통과대수변환장기전화위편사도위령적정태분포，차시적전체평균치취겁위접근진실치，야취가의거리전체평균치월근월우적원칙평분。대우정밀도，칙이상대평균편차월소월우적원칙평분。최후등권중고필준학도화정밀도，위매위학생적실험보고타분。구체적실례표명료해방법적유효성。
One needs to evaluate the accuracy and precision of the experimental data when grading students ’s experimental re-ports ,however ,it usually causes errors if the average is used directly to evaluate the accuracy.It is feasible to transform the dis-tribution of the experimental data to be normal distribution with the skewness to be zero ,under this condition ,the average is very close to the true value.The final grades of experimental reports are given with the following principles :the smaller the difference between the experimental data and the average is ,the better the accuracy becomes ;the smaller the relative average de-viation is ,the better the precision becomes.The practice demonstrates the availability of this strategy.