Research On Test Equating Of Examination In Enterprise Human Resources Professional

Based on the classical test theory,we explore the test equating of examination enterprise human resources professional(Level 3)in September 2015 and November2015.In this research,we use the method which combines common-item non-equivalent groups design with random groups design.In order to choose the best and the most stable equating method for the examination in Enterprise Human Resources Professional,we take two methods.One is standard errors and the other is Cross validation,as the standard,they compare these equating methods such as the Tucker linear method,Levine observed scores equating method and Levine true score equating method.Then,we use the best method to design the equivalent fraction conversion tables,which will provide some references for the related research in the future.The correlation coefficient between the anchor items and two examinations are0.835 and 0.759.There is a significant correlation between anchor test and two examinations.It shows that the anchor test is scientific and it can be used to test the participants.In linear equating method,when the weight is 1,intercept is-3.264 and slope is 1.051 when we use the Tucker linear method.Intercept is-8.719 and slope is1.188 when we use the Levine observed scores equating method.When the weight is0.501,intercept is-3.774 and slope is 1.063 when we use the Tucker linear method.Intercept is-8.52 and slope is 1.183 when we use the Levine observed scores equating method.Intercept is-10.1 and slope is 1.219 when we use the Levine true score equating method.According to the method of standard error,the value of standard error of the Tucker linear method is the minimum when the weight is 1,which is 0.741.The second minimum value of standard error is the Tucker linear method when the weight is 0.501,which is 0.856.Followed by the two methods are the Levine observed scores equating method when the weight is 0.501 and the Levine observed scores equating method when the weight is 1.The value of standard error of the two methods are 2.086 and2.137.The value of standard error of the Levine true score equating method is the maximum,which is 2.431.According to the method of Cross validation,the value of T of the Tucker linear method when the weight is 0.5 is the minimum,which is 1.347.Followed by this method are the Tucker linear method when the weight is 1,the Levine observed scores equating method when the weight is 1 and the Levine observed scores equating method when the weight is 0.5.The value of T of the three methods are 2.523,2.973 and 4.199.The value of T of the Levine true scores equating method is the maximum,which is 5.799.The research for the test equating to the examination in Enterprise Human Resources Professional shows the following conclusions.1.As a way to evaluate the advantages and disadvantages of different equivalent methods,Cross validation is Practical and effective.2.In comparing,no matter the standard errors are used or the Cross validation are used,the result is the following.The Tucker linear method is the best and the most stable method,the Levine is worse than the Tucker linear,while the Levine true scores equating method is the worst method.The Tucker linear method when the weight is 1 is the best method to the Examination in Enterprise Human Resources Professional.3.Different weight has little effect on the result of test equating.