纵向Rasch模型在大学新生适应性追踪研究中的应用

对1952名大学新生进行适应性调查，其中285人接受了2次以上的追踪调查，所得的多级评分重复测量数据采用纵向Rasch模型进行统计分析。研究应用SAS的GLIMMIX过程对多层Rasch模型参数估计作了新的尝试。结果表明：（1）新生在第一学年内，学习和情绪适应总体呈上升趋势，人际适应呈下降趋势；（2）不同个体入学时的适应状况差异显著，但是随时间变化的趋势、快慢相同；（3）学习适应分量表的项目稳定性较好，而人际、情绪适应的部分项目难度存在时间效应。研究结果对新生辅导具有启示意义

Adjustment of Freshnmn College Students:A longitudinal Study using Longituding Rasch Model

University attendance is a critical transition for many late adolescents. Many researchers have argued that successful adjustment, particularly during the first year, can predict students’ academic success, mental health and personal development. Thus, it is imperative for researchers to understand how adjustment problems emerge over time and vary across different students and develop prevention and how intervention efforts aid a smooth and productive transition. However, most of the traditional longitudinal analyses just focus on the general trend of change in people with time but ignore the individual differences. Moreover, few of the longitudinal studies consider the stability of instruments. The present longitudinal study aims to better understand the overall trend and individual differences of how academic, communication, and emotional adjustments of freshmen change over time, concurrently with an assessment of the invariance of the item location. With an initial sample of 1940 participants recruited in 6 universities, the participants in the follow-up study sampled from only 1 university of the initial sample in convenience. New subjects were included in while some were lost during the following 3 waves of measurement. The final sample going to analysis comprised 1952 freshmen who were used for at least one measurement. The adjustment of freshmen had been assessed 4 times during the first year, in Oct 2005, in Feb 2005, in Apr 2006, and in Jun 2006 using the Adjustment Questionnaire of College Freshman (AQCF). With SAS GLIMMIX procedure performing the analysis, longitudinal Rasch model was used to fit the polytomous repeated-measures data. The following results were obtained: (1) Although the majority of students were well adjusted in academic study, emotion, and communication, academic transition is the most challenging for freshmen. (2) Academic and emotional adjustments had increased whereas communication adjustment declined during the first year in general. For academic adjustment has negative linear time effect and positive quadratic time effect, the change of academic adjustment over time appears a U-shape, with the lowest score at the end of the first semester. (3) There was significant variation in adjustment among freshmen on the entrance to college, but the trend of change with time was the same for all students.(4) The item locations of Academic scale were invariant over time, whereas some item locations on emotion and communication scales had changed linearly with time. Student development professionals should pay special attention to students with poor adjustment. When they provide useful guidance to those students according to their specific needs during their first year, it should be noted that academic transition was the most challenging one and the academic adjustment may drop at the end of the first semester. When researchers use the AQCF questionnaire, the stability of emotion and communication sub-scales should be considered. Although longitudinal Rasch Model is useful in many ways, it still needs more investigation to popularize it