Two experiments examine developmental changes of error type in serial recall tasks.Experiment 1 takes respectively as subjects 20 primary school students of grade one, grade three, and grade five, chooses three categories of Chinese characters (normal Chinese characters, confused Chinese characters in phonology and similar Chinese characters in shape) as experimental materials, and takes the proportion of correctness and proportion of errors as dependent variables. The results of experiment 1 show (l)The proportion of correctness of primary school students increases with the grade advancing ,and the difference of proportion of correctness between grade one and grade five is significant. (2)In serial recall task, four types of error are found, that is omissions, intrusions, movements and repetitions. The differences of proportion of erorrs between grade one and grade three and between grade one and grade five are significant.The proportion of movements is the highest ,on the contrary, the proportion of repetitions is the lowest. (3)Four types of error do not develop in balance, proportions of errors of omissions and movements decrease with grade. The differences of omissions between each grade are significant. The difference of movements between grade one and grade five is significant. There is no difference on intrusions and repetitions. (4) Errors of confused Chinese characters in phonology are more than normal Chinese characters and similar Chinese characters in shape.The differences in proportion of errors of omissions, intrusions and movements between confused Chinese characters in phonology and the other two categories of Chinese characters are significant.(5)Proportion of correctness of each grade in each position increases with age.The position effects are different between four types of error, and the position effects of omissions and movements are significant.Experiment 2 takes respectively as subjects 20 primary school students of grade one, grade three, and grade five, half of whom in each grade are high-score students, the other half of whom are low-score students. In experiment 2, experimental materials and dependent variables are the same as experiment l.The results of experiment 2 show (1) The proportion of correctness of high-score students and low-score students increases with age. The proportion of correctness of high-score students is higher than that of low-score students. The difference of proportion of correctness between high-score students and low-score students is significant.(2)The proportion of errors of low-score students is higher than that of high-score students. The omissions' proportion of errors of low-score students is higher than that of high-score students, and the difference is significant. The intrussions' proportion errors of low-score students is higher than that of high-score students, and the difference is significant.The movements' proportion of errors of low-score students is higher than that of high-score students, and the differences between grade one and grade five are significant. There is no difference in repititions between high-score students andlow-score students. (3)The proportion of correctness of high-score students is higher than low-score students in three categories of Chinese characters. For both high-score and low-score students, The proportion of errors of confused Chinese characters in phonology are apparently more than normal Chinese characters and similar Chinese characters in shape, and the differences are significant.The research indicates that the each-sided development of STM is uneven balanced although the overall ability of STM for the primary school students increases with the grade advancing. The inhibition of output and the speed of output develop the fastest; but the suppression of dealing with the irrelevant information and the development of the encoding ability of the temporary characteristics develop relatively late. It shows that intrusions and repetitions have no differences in terms of grades advancing while the pr... |