追踪研究方法在国内心理研究中的应用述评

以1982~2012年中国期刊网收录的88例追踪研究为对象,从应用现状、设计特征、数据处理三方面分析和评估追踪研究方法在国内心理研究的应用情况及存在的问题。结果显示,2005年之前追踪研究方法应用增长缓慢,2005年开始呈显著增长趋势,研究对象以未成年及成年早期群体为主。主要采用固定样本追踪设计,大部分研究测量2-3次、样本量在10~300之间、持续时间在3年内。61例有缺失的研究中,38例用删除法处理缺失;主要运用 HLM、方差分析、t检验和SEM分析追踪数据。相当部分研究存在测量次数少、样本量较小、持续时间短、被试缺失严重及数据处理方法相对陈旧问题。追踪研究方法的应用应注意,根据理论模型和研究有效性要求确定设计类型和设计特征,根据数据特征选择缺失处理方法和追踪数据分析方法。     

加速追踪设计的方法和应用

加速追踪设计(ALD)是一种选择相邻多个群组同时进行短期追踪研究, 获得在测量上有重叠的多个群组追踪数据, 对多个群组数据进行合并建构一条在时间跨度上较长的发展趋势或增长曲线的方法。ALD结合真追踪和横断设计的特征, 既保持真追踪设计的大部分优点, 克服真追踪研究中由于重测效应和被试缺失导致的问题, 又尝试分离年龄、群组和历史时间效应, 在发展心理研究有重要应用。已有研究探讨ALD的数据分析方法、ALD的有效性及设计特征。未来研究应关注拓展设计条件下ALD的适应性, 探索非线性假设或群组效应显著时的数据分析方法和ALD中缺失数据处理问题。     

认知诊断缺失数据处理方法的比较：零替换、多重插补与极大似然估计法

数据缺失在测验中经常发生, 认知诊断评估也不例外, 数据缺失会导致诊断结果的偏差。首先, 通过模拟研究在多种实验条件下比较了常用的缺失数据处理方法。结果表明：(1)缺失数据导致估计精确性下降, 随着人数与题目数量减少、缺失率增大、题目质量降低, 所有方法的PCCR均下降, Bias绝对值和RMSE均上升。(2)估计题目参数时, EM法表现最好, 其次是MI, FIML和ZR法表现不稳定。(3)估计被试知识状态时, EM和FIML表现最好, MI和ZR表现不稳定。其次, 在PISA2015实证数据中进一步探索了不同方法的表现。综合模拟和实证研究结果, 推荐选用EM或FIML法进行缺失数据处理。     

基于增长模型的非随机缺失数据处理:选择模型和极大似然方法

对含有非随机缺失数据的潜变量增长模型,为了考察基于不同假设的缺失数据处理方法:极大似然(ML)方法与DiggleKenward选择模型的优劣,通过Monte Carlo模拟研究,比较两种方法对模型中增长参数估计精度及其标准误估计的差异,并考虑样本量、非随机缺失比例和随机缺失比例的影响。结果表明,符合前提假设的Diggle-Kenward选择模型的参数估计精度普遍高于ML方法;对于标准误估计值,ML方法存在一定程度的低估,得到的置信区间覆盖比率也明显低于Diggle-Kenward选择模型。     

LGM模型中缺失数据处理方法的比较:ML方法与Diggle-Kenward选择模型

追踪研究中缺失数据十分常见。本文通过Monte Carlo模拟研究,考察基于不同前提假设的Diggle-Kenward选择模型和ML方法对增长参数估计精度的差异,并考虑样本量、缺失比例、目标变量分布形态以及不同缺失机制的影响。结果表明:(1)缺失机制对基于MAR的ML方法有较大的影响,在MNAR缺失机制下,基于MAR的ML方法对LGM模型中截距均值和斜率均值的估计不具有稳健性。(2)DiggleKenward选择模型更容易受到目标变量分布偏态程度的影响,样本量与偏态程度存在交互作用,样本量较大时,偏态程度的影响会减弱。而ML方法仅在MNAR机制下轻微受到偏态程度的影响。     

计划缺失设计——通过有意缺失让研究更高效

缺失值是社会科学研究中非常普遍的现象。全息极大似然估计和多重插补是目前处理缺失值最有效的方法。计划缺失设计利用特殊的实验设计有意产生缺失值, 再用现代的缺失值处理方法来完成统计分析, 获得无偏的统计结果。计划缺失设计可用于横断面调查减少(或增加)问卷长度和纵向调查减少测量次数, 也可用于提高测量有效性。常用的计划缺失设计有三式设计和两种方法测量。     

2PL模型的两种马尔可夫蒙特卡洛缺失数据处理方法比较

马尔科夫蒙特卡洛（MCMC）是项目反应理论中处理缺失数据的一种典型方法。文章通过模拟研究比较了在不同被试人数,项目数,缺失比例下两种MCMC方法（M-H within Gibbs和DA-T Gibbs）参数估计的精确性,并结合了实证研究。研究结果表明,两种方法是有差异的,项目参数估计均受被试人数影响很大,受缺失比例影响相对更小。在样本较大缺失比例较小时,M-H within Gibbs参数估计的均方误差（RMSE）相对略小,随着样本数的减少或缺失比例的增加,DA-T Gibbs方法逐渐优于M-H within Gibbs方法     

缺失数据处理的期望-极大化算法与马尔可夫蒙特卡洛方法

缺失数据普遍存在于心理学研究中, 影响着统计推断。极大似然估计(MLE)与基于贝叶斯的多重借补(MI)是处理缺失数据的两类重要方法。期望-极大化算法(EM)是寻求MLE的一种强有力的方法。马尔可夫蒙特卡洛方法(MCMC)可以相对简易地实现MI, 而且可以适用于复杂情况下的缺失数据处理。结合研究的需要讨论了实现这两类方法的适用软件。     

认知诊断测评中缺失数据的处理：随机森林阈值插补法

认知诊断测评中缺失数据的处理是理论和实际应用者非常关注的研究主题。借鉴随机森林插补法(RFI)不依赖于缺失机制假设的特点,对已有的RFI方法进行改进,提出采用个人拟合指标(RCI)确定插补阈值的新方法:随机森林阈值插补方法(RFTI)。模拟研究表明,RFTI在插补正确率上明显高于RFI方法;与RFI和EM方法相比,RFTI在被试属性模式判准率和边际判准率上表现出明显优势,尤其是非随机缺失和混合缺失机制,以及缺失比例较高的条件下,其优势更加明显。但对项目参数的估计, RFTI方法与EM方法相比不具有优势。     

国内追踪数据分析方法研究与模型发展

追踪研究因其可以得到比横断研究更有说服力的变量关系论证, 在心理学等科学中具有重要地位。梳理国内以心理学为主的相关领域中追踪数据分析方法研究的发表现状、主要解决的研究问题和模型发展。追踪研究可以进行均值差异比较、分析多变量相互影响、描述总体发展趋势及差异和探究心理动态变化过程。近20年的研究热点和发展思路也集中在上述研究问题当中, 特别是总体发展趋势及差异、多变量相互影响、总体发展趋势与多变量相互影响的融合、追踪研究设计、缺失数据等议题上。最后, 比较国内外研究的差异, 并结合交叉学科对国内追踪研究未来发展做出展望。     

How to handle missing data: A comparison of different approaches

Many researchers face the problem of missing data in longitudinal research. Especially, high risk samples are characterized by missing data which can complicate analyses and the interpretation of results. In the current study, our aim was to find the most optimal and best method to deal with the missing data in a specific study with many missing data on the outcome variable. Therefore, different techniques to handle missing data were evaluated, and a solution to efficiently handle substantial amounts of missing data was provided. A simulation study was conducted to determine the most optimal method to deal with the missing data. Results revealed that multiple imputation (MI) using predictive mean matching was the most optimal method with respect to lowest bias and the smallest confidence interval (CI) while maintaining power. Listwise deletion and last observation carried backward also scored acceptable with respect to bias; however, CIs were much larger and sample size almost halved using these methods. Longitudinal research in high risk samples could benefit from using MI in future research to handle missing data. The paper ends with a checklist for handling missing data.     

Maximizing data quality and shortening survey time: Three-form planned missing data survey design

Simulation studies have shown the three-form planned missing data design efficiently collects high quality data while reducing participant burden. This methodology is rarely used in sport and exercise psychology. Therefore, we conducted a re-sampling study with existing sport and exercise psychology survey data to test how three-form planned missing data survey design implemented with different item distribution approaches effect constructs’ internal measurement structure and validity. Results supported the efficacy of the three-form planned missing data survey design for cross-sectional data collection. Sample sizes of at least 300 (i.e., 100 per form) are recommended for having unbiased parameter estimates. It is also recommended items be distributed across survey forms to have representation of each facet of a construct on every form, and that a select few of these items be included across all survey forms. Further guidelines for three-form surveys based upon the results of this resampling study are provided.     

Full Information Maximum Likelihood Estimation for Latent Variable Interactions With Incomplete Indicators

Researchers have developed missing data handling techniques for estimating interaction effects in multiple regression. Extending to latent variable interactions, we investigated full information maximum likelihood (FIML) estimation to handle incompletely observed indicators for product indicator (PI) and latent moderated structural equations (LMS) methods. Drawing on the analytic work on missing data handling techniques in multiple regression with interaction effects, we compared the performance of FIML for PI and LMS analytically. We performed a simulation study to compare FIML for PI and LMS. We recommend using FIML for LMS when the indicators are missing completely at random (MCAR) or missing at random (MAR) and when they are normally distributed. FIML for LMS produces unbiased parameter estimates with small variances, correct Type I error rates, and high statistical power of interaction effects. We illustrated the use of these methods by analyzing the interaction effect between advanced cancer patients’ depression and change of inner peace well-being on future hopelessness levels.     

A maximum likelihood approach to test validation with missing and censored dependent variables

A maximum likelihood approach is described for estimating the validity of a test (x) as a predictor of a criterion variable (y) when there are both missing and censoredy scores present in the data set. The missing data are due to selection on a latent variable (y _s ) which may be conditionally related toy givenx. Thus, the missing data may not be missing random. The censoring process in due to the presence of a floor or ceiling effect. The maximum likelihood estimates are constructed using the EM algorithm. The entire analysis is demonstrated in terms of hypothetical data sets.     

Accounting for Missing Correlation Coefficients in Fixed-Effects MASEM

Meta-analytic structural equation modeling (MASEM) is increasingly applied to advance theories by synthesizing existing findings. MASEM essentially consists of two stages. In Stage 1, a pooled correlation matrix is estimated based on the reported correlation coefficients in the individual studies. In Stage 2, a structural model (such as a path model) is fitted to explain the pooled correlations. Frequently, the individual studies do not provide all the correlation coefficients between the research variables. In this study, we modify the currently optimal MASEM-method to deal with missing correlation coefficients, and compare its performance with existing methods. This study is the first to evaluate the performance of fixed-effects MASEM methods under different levels of missing correlation coefficients. We found that the often used univariate methods performed very poorly, while the multivariate methods performed well overall.     

Asymptotic bias of normal-distribution-based maximum likelihood estimates of moderation effects with data missing at random

Moderation analysis is useful for addressing interesting research questions in social sciences and behavioural research. In practice, moderated multiple regression (MMR) models have been most widely used. However, missing data pose a challenge, mainly because the interaction term is a product of two or more variables and thus is a non-linear function of the involved variables. Normal-distribution-based maximum likelihood (NML) has been proposed and applied for estimating MMR models with incomplete data. When data are missing completely at random, moderation effect estimates are consistent. However, simulation results have found that when data in the predictor are missing at random (MAR), NML can yield inaccurate estimates of moderation effects when the moderation effects are non-null. Simulation studies are subject to the limitation of confounding systematic bias with sampling errors. Thus, the purpose of this paper is to analytically derive asymptotic bias of NML estimates of moderation effects with MAR data. Results show that when the moderation effect is zero, there is no asymptotic bias in moderation effect estimates with either normal or non-normal data. When the moderation effect is non-zero, however, asymptotic bias may exist and is determined by factors such as the moderation effect size, missing-data proportion, and type of missingness dependence. Our analytical results suggest that researchers should apply NML to MMR models with caution when missing data exist. Suggestions are given regarding moderation analysis with missing data.