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。     

时距长度和速度变化对进度条时距知觉的影响

进度条在人机交互界面中具有广泛应用。本研究运用单任务的预期式范式和口头估计法,探究速度变化对不同时距长度进度条的时距知觉的影响,为进度条的智能化设计提供参考。研究表明：（1）时距长度影响个体的时距估计,时距长度与时距估计比率分数成反比关系;（2）进度条的不同速度变化趋势影响进度条时距知觉,在持续时间相同的匀速、减速、加速和先加后减四种进度条类型中,加速进度条的时距知觉最短,而匀速进度条时距估计最长。     

中介效应的检验方法和效果量测量:回顾与展望

通过中介效应检验方法之间的比较和效果量指标之间的比较,建议放弃将总效应c显著作为中介效应检验的前提条件,放弃基于直接效应c'显著性的完全和部分中介的提法,推荐使用偏差校正的百分位Bootstrap法直接对中介效应ab进行检验,使用κ²、R_med²等中介效果量指标并报告效果量的置信区间。作为示例,用R软件的MBESS软件包对某消防员饮食健康调查进行了中介效应检验和效果量测量。随后展望了中介效应检验方法和效果量测量的拓展方向。     

使用似然比D2统计量的题目属性定义方法

题目属性的定义是实施认知诊断评价的关键步骤, 通过有丰富经验的领域专家对题目的属性进行定义是当前的主要方法, 然而该方法受到许多主观经验因素的影响。寻找客观的题目属性定义或验证方法可以为主观定义过程提供策略支持或对结果进行改进, 因此已经引起研究者们的关注。本研究构建了一种简单高效的题目属性定义方法, 研究使用似然比D2统计量从作答数据中估计题目属性的方法, 实现属性掌握模式、题目参数和题目属性向量的联合估计。模拟研究结果表明, 使用似然比D2统计量可以有效地识别题目的属性向量, 该方法一方面可以实现新编制题目属性向量的在线估计, 另一方面可以验证已经定义的题目属性向量的准确性。     

儿童数字线表征的发展——心理长度的影响

探讨儿童在数字线估计任务中心理长度的发展及其对数量表征模式的影响.选取7~9岁儿童共109名进行数字线估计任务测试,设置了10cm和20cm两种长度条件,要求儿童完成根据位置判断数字任务(position to number,PN任务).结果表明儿童在数字线PN任务中存在心理长度,且7岁儿童心理长度的范围超过10,随着儿童年龄的增长,他们的心理长度范围不断缩小;心理长度范围影响儿童的表征模式,随着心理长度范围的缩小,儿童的数字线表征出现从指数模式到线性模式的变化趋势;与表征模式的发展趋势一致,儿童估计的精确性随年龄增长逐渐提高.     

Does performance on the standard antisaccade task meet the co-familiality criterion for an endophenotype?

The “co-familiality” criterion for an endophenotype has two requirements: (1) clinically unaffected relatives as a group should show both a shift in mean performance and an increase in variance compared with controls; (2) performance scores should be heritable. Performance on the antisaccade task is one of several candidate endophenotypes for schizophrenia. In this paper we examine whether the various measures of performance on the standard version of the antisaccade task meet the co-familiality criterion for an endophenotype. The three measures of performance—reflexive saccade errors, latency of correct antisaccades, and gain—show a wide range of effect sizes and variance ratios as well as evidence of significant or near significant heterogeneity. The estimated mean effect sizes [Cohen’s d: error rate: 0.34 (SD: 0.29); latency: 0.33 (SD: 0.30); gain: 0.54 (SD: 0.38)] are significantly greater than 0, but the magnitude of the departures from 0 is relatively small, corresponding to modest effect sizes. The width of the 95% confidence intervals for the estimated effect sizes (error rate: 0.2–0.49; latency: 0.17–0.50; gain: 0.23–0.85) and the coefficients of variation in effect sizes (error rate: 85.3%; latency: 90.9%; gain: 68.4%) reflect heterogeneity in effect sizes. The effect sizes for error rate showed statistically significant heterogeneity and those for latency (P = .07) and gain (P = .09) showed a trend toward heterogeneity. These results indicate that the effect sizes are not consistent with a single mean and that the average effect size may be a biased estimate of the magnitude of differences in performance between relatives of schizophrenics and controls. Relatives of schizophrenics show a small but significant increase in variance in error rate, but the confidence interval is broad, perhaps reflecting the heterogeneity in effect size. The variance ratios for latency and gain did not differ in relatives of schizophrenics and controls. Performance, as measured by error rate, is moderately heritable. The data do not provide compelling support for a consistent shift in mean or variance in relatives of schizophrenia patients compared with nonpsychiatric controls, both of which are required for a major gene involved in co-familial transmission. This set of findings suggests that although intra-familial resemblance in antisaccade performance is due in part to genetic factors, it may not be related to a schizophrenia genotype. Based on the current literature, it would be premature to conclude that any of the measures of antisaccade performance unambiguously meets the co-familiality criterion for an endophenotype.     

Reducing Measurement Error in Student Achievement Estimation

The achievement level is a variable measured with error, that can be estimated by means of the Rasch model. Teacher grades also measure the achievement level but they are expressed on a different scale. This paper proposes a method for combining these two scores to obtain a synthetic measure of the achievement level based on the theory developed for regression with covariate measurement error. In particular, the focus is on ordinal scaled grades, using the SIMEX method for measurement error correction. The result is a measure comparable across subjects with smaller measurement error variance. An empirical application illustrates the method.     

A fast numerical algorithm for the estimation of diffusion model parameters

In this paper, we describe a new algorithmic approach for parameter estimation in Ratcliff's [(1978). A theory of memory retrieval. Psychological Review, 85 (2), 59-108] diffusion model. This problem, especially if inter-trial variabilities of parameters are included in the model, is computationally very expensive; the parameter estimation procedure often takes a long time even with today's high-speed computers. The algorithm described here makes the calculation of the cumulative distribution functions for predicted process durations computationally much less expensive. This improvement is achieved by solving the Kolmogorov backward equation numerically instead of employing the previously used closed form solution. Additionally, the algorithm can determine the optimum fit for one of the model parameters (the starting point z) directly, thereby reducing the dimension of the parameter search space by one. The resulting method is shown to be notably faster than the standard (closed-form solution) method for parameter estimation.     

Limited information estimation and testing of discretized multivariate normal structural models

Discretized multivariate normal structural models are often estimated using multistage estimation procedures. The asymptotic properties of parameter estimates, standard errors, and tests of structural restrictions on thresholds and polychoric correlations are well known. It was not clear how to assess the overall discrepancy between the contingency table and the model for these estimators. It is shown that the overall discrepancy can be decomposed into a distributional discrepancy and a structural discrepancy. A test of the overall model specification is proposed, as well as a test of the distributional specification (i.e., discretized multivariate normality). Also, the small sample performance of overall, distributional, and structural tests, as well as of parameter estimates and standard errors is investigated under conditions of correct model specification and also under mild structural and/or distributional misspecification. It is found that relatively small samples are needed for parameter estimates, standard errors, and structural tests. Larger samples are needed for the distributional and overall tests. Furthermore, parameter estimates, standard errors, and structural tests are surprisingly robust to distributional misspecification. This research was supported by the Department of Universities, Research and Information Society (DURSI) of the Catalan Government, and by grants BSO2000-0661 and BSO2003-08507 of the Spanish Ministry of Science and Technology.