Pub Date : 2011-04-15DOI: 10.1027/1614-2241/A000023
A. Ünlü
Schrepp (2005) points out and builds upon the connection between knowledge space theory (KST) and latent class analysis (LCA) to propose a method for constructing knowledge structures from data. Candidate knowledge structures are generated, they are considered as restricted latent class models and fitted to the data, and the BIC is used to choose among them. This article adds additional information about the relationship between KST and LCA. It gives a more comprehensive overview of the literature and the probabilistic models that are at the interface of KST and LCA. KST and LCA are also compared with regard to parameter estimation and model testing methodologies applied in their fields. This article concludes with an overview of KST-related publications addressing the outlined connection and presents further remarks about possible future research arising from a connection of KST to other latent variable modeling approaches.
{"title":"A Note on the Connection Between Knowledge Structures and Latent Class Models","authors":"A. Ünlü","doi":"10.1027/1614-2241/A000023","DOIUrl":"https://doi.org/10.1027/1614-2241/A000023","url":null,"abstract":"Schrepp (2005) points out and builds upon the connection between knowledge space theory (KST) and latent class analysis (LCA) to propose a method for constructing knowledge structures from data. Candidate knowledge structures are generated, they are considered as restricted latent class models and fitted to the data, and the BIC is used to choose among them. This article adds additional information about the relationship between KST and LCA. It gives a more comprehensive overview of the literature and the probabilistic models that are at the interface of KST and LCA. KST and LCA are also compared with regard to parameter estimation and model testing methodologies applied in their fields. This article concludes with an overview of KST-related publications addressing the outlined connection and presents further remarks about possible future research arising from a connection of KST to other latent variable modeling approaches.","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"7 1","pages":"63-67"},"PeriodicalIF":3.1,"publicationDate":"2011-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292879","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2011-04-15DOI: 10.1027/1614-2241/A000022
Edixon J. Chacón, Jesús M Alvarado, C. Santisteban
The LISREL8.8/PRELIS2.81 program can carry out ordinal factorial analysis (OFA command), with full information maximum likelihood methods, in a data set containing n samples obtained by simulation. Nevertheless, when the replication number is greater than 1, an error command is produced, which impedes reaching solutions that can execute normal (NOR) and logistic (POM) functions. This paper proposes a new procedure of data simulation in PRELIS-LISREL. This procedure permits the generation of n replications and the calculation of the goodness of fit (GOF) indices in the item response theory (IRT) models for each replication, thus allowing the execution of the OFA command for Monte Carlo simulations. The solutions using underlying variable (weighted least squares (WLS) estimation method) and IRT approaches are compared.
{"title":"A simulation procedure for the generation of samples to evaluate goodness of fit indices in item response theory models.","authors":"Edixon J. Chacón, Jesús M Alvarado, C. Santisteban","doi":"10.1027/1614-2241/A000022","DOIUrl":"https://doi.org/10.1027/1614-2241/A000022","url":null,"abstract":"The LISREL8.8/PRELIS2.81 program can carry out ordinal factorial analysis (OFA command), with full information maximum likelihood methods, in a data set containing n samples obtained by simulation. Nevertheless, when the replication number is greater than 1, an error command is produced, which impedes reaching solutions that can execute normal (NOR) and logistic (POM) functions. This paper proposes a new procedure of data simulation in PRELIS-LISREL. This procedure permits the generation of n replications and the calculation of the goodness of fit (GOF) indices in the item response theory (IRT) models for each replication, thus allowing the execution of the OFA command for Monte Carlo simulations. The solutions using underlying variable (weighted least squares (WLS) estimation method) and IRT approaches are compared.","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"7 1","pages":"56-62"},"PeriodicalIF":3.1,"publicationDate":"2011-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2011-04-15DOI: 10.1027/1614-2241/A000021
Fiona Dziopa, K. Ahern
Q-methodology is a technique incorporating the benefits of both qualitative and quantitative research. Q-method involves Q-sorting, a method of data collection and factor analysis, to assess subjective (qualitative) information. The use of Q-sorting and factor analysis has often resulted in the misconception that Q-methodology involves psychometric or quantitative assessment, although Q as a methodology actually enables the systematic assessment of qualitative data. Misconceptions regarding Q have resulted in a heterogeneous collection of Q-applications in the extant literature, which has obscured the fundamental principles of Q-methodology. The purpose of this paper is to present a systematic review of Q-based research to investigate the criteria researchers have used to develop Q-studies. Published research studies between January 2008 and December 2008 that employed Q-techniques and methodology were assessed. Data were extracted and synthesized through the development and use of the Assessment and Revi...
{"title":"A Systematic Literature Review of the Applications of Q-Technique and Its Methodology","authors":"Fiona Dziopa, K. Ahern","doi":"10.1027/1614-2241/A000021","DOIUrl":"https://doi.org/10.1027/1614-2241/A000021","url":null,"abstract":"Q-methodology is a technique incorporating the benefits of both qualitative and quantitative research. Q-method involves Q-sorting, a method of data collection and factor analysis, to assess subjective (qualitative) information. The use of Q-sorting and factor analysis has often resulted in the misconception that Q-methodology involves psychometric or quantitative assessment, although Q as a methodology actually enables the systematic assessment of qualitative data. Misconceptions regarding Q have resulted in a heterogeneous collection of Q-applications in the extant literature, which has obscured the fundamental principles of Q-methodology. The purpose of this paper is to present a systematic review of Q-based research to investigate the criteria researchers have used to develop Q-studies. Published research studies between January 2008 and December 2008 that employed Q-techniques and methodology were assessed. Data were extracted and synthesized through the development and use of the Assessment and Revi...","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"7 1","pages":"39-55"},"PeriodicalIF":3.1,"publicationDate":"2011-04-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1027/1614-2241/A000021","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292827","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2011-01-01DOI: 10.1027/1614-2241/A000033
K. Schweizer
Probability-based and measurement-related hypotheses for confirmatory factor analysis of repeated-measures data are investigated. Such hypotheses comprise precise assumptions concerning the relationships among the true components associated with the levels of the design or the items of the measure. Measurement-related hypotheses concentrate on the assumed processes, as, for example, transformation and memory processes, and represent treatment-dependent differences in processing. In contrast, probability-based hypotheses provide the opportunity to consider probabilities as outcome predictions that summarize the effects of various influences. The prediction of performance guided by inexact cues serves as an example. In the empirical part of this paper probability-based and measurement-related hypotheses are applied to working-memory data. Latent variables according to both hypotheses contribute to a good model fit. The best model fit is achieved for the model including latent variables that represented seri...
{"title":"Probability-Based and Measurement- Related Hypotheses With Full Restriction for Investigations by Means of Confirmatory Factor Analysis An Example From Cognitive Psychology","authors":"K. Schweizer","doi":"10.1027/1614-2241/A000033","DOIUrl":"https://doi.org/10.1027/1614-2241/A000033","url":null,"abstract":"Probability-based and measurement-related hypotheses for confirmatory factor analysis of repeated-measures data are investigated. Such hypotheses comprise precise assumptions concerning the relationships among the true components associated with the levels of the design or the items of the measure. Measurement-related hypotheses concentrate on the assumed processes, as, for example, transformation and memory processes, and represent treatment-dependent differences in processing. In contrast, probability-based hypotheses provide the opportunity to consider probabilities as outcome predictions that summarize the effects of various influences. The prediction of performance guided by inexact cues serves as an example. In the empirical part of this paper probability-based and measurement-related hypotheses are applied to working-memory data. Latent variables according to both hypotheses contribute to a good model fit. The best model fit is achieved for the model including latent variables that represented seri...","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"7 1","pages":"157-164"},"PeriodicalIF":3.1,"publicationDate":"2011-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292945","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-09-08DOI: 10.1027/1614-2241/A000016
Emanuel Schmider, M. Ziegler, Erik Danay, Luzi Beyer, M. Bühner
Empirical evidence to the robustness of the analysis of variance (ANOVA) concerning violation of the normality assumption is presented by means of Monte Carlo methods. High-quality samples underlying normally, rectangularly, and exponentially distributed basic populations are created by drawing samples which consist of random numbers from respective generators, checking their goodness of fit, and allowing only the best 10% to take part in the investigation. A one-way fixed-effect design with three groups of 25 values each is chosen. Effect-sizes are implemented in the samples and varied over a broad range. Comparing the outcomes of the ANOVA calculations for the different types of distributions, gives reason to regard the ANOVA as robust. Both, the empirical type I error α and the empirical type II error β remain constant under violation. Moreover, regression analysis identifies the factor “type of distribution” as not significant in explanation of the ANOVA results.
{"title":"Is It Really Robust","authors":"Emanuel Schmider, M. Ziegler, Erik Danay, Luzi Beyer, M. Bühner","doi":"10.1027/1614-2241/A000016","DOIUrl":"https://doi.org/10.1027/1614-2241/A000016","url":null,"abstract":"Empirical evidence to the robustness of the analysis of variance (ANOVA) concerning violation of the normality assumption is presented by means of Monte Carlo methods. High-quality samples underlying normally, rectangularly, and exponentially distributed basic populations are created by drawing samples which consist of random numbers from respective generators, checking their goodness of fit, and allowing only the best 10% to take part in the investigation. A one-way fixed-effect design with three groups of 25 values each is chosen. Effect-sizes are implemented in the samples and varied over a broad range. Comparing the outcomes of the ANOVA calculations for the different types of distributions, gives reason to regard the ANOVA as robust. Both, the empirical type I error α and the empirical type II error β remain constant under violation. Moreover, regression analysis identifies the factor “type of distribution” as not significant in explanation of the ANOVA results.","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"6 1","pages":"147-151"},"PeriodicalIF":3.1,"publicationDate":"2010-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1027/1614-2241/A000016","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292813","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-06-30DOI: 10.1027/1614-2241/A000014
C. Atzmüller, Peter M Steiner
Vignette studies use short descriptions of situations or persons (vignettes) that are usually shown to respondents within surveys in order to elicit their judgments about these scenarios. By systematically varying the levels of theoretically important vignette characteristics a large population of different vignettes is typically available – too large to be presented to each respondent. Therefore, each respondent gets only a subset of vignettes. These subsets may either be randomly selected in following the tradition of the factorial survey or systematically selected according to an experimental design. We show that these strategies in selecting vignette sets have strong implications for the analysis and interpretation of vignette data. Random selection strategies result in a random confounding of effects and heavily rely on the assumption of no interaction effects. In contrast, experimental strategies systematically confound interaction effects with main or set effects, thereby preserving a meaningful in...
{"title":"Experimental Vignette Studies in Survey Research","authors":"C. Atzmüller, Peter M Steiner","doi":"10.1027/1614-2241/A000014","DOIUrl":"https://doi.org/10.1027/1614-2241/A000014","url":null,"abstract":"Vignette studies use short descriptions of situations or persons (vignettes) that are usually shown to respondents within surveys in order to elicit their judgments about these scenarios. By systematically varying the levels of theoretically important vignette characteristics a large population of different vignettes is typically available – too large to be presented to each respondent. Therefore, each respondent gets only a subset of vignettes. These subsets may either be randomly selected in following the tradition of the factorial survey or systematically selected according to an experimental design. We show that these strategies in selecting vignette sets have strong implications for the analysis and interpretation of vignette data. Random selection strategies result in a random confounding of effects and heavily rely on the assumption of no interaction effects. In contrast, experimental strategies systematically confound interaction effects with main or set effects, thereby preserving a meaningful in...","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"6 1","pages":"128-138"},"PeriodicalIF":3.1,"publicationDate":"2010-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-06-29DOI: 10.1027/1614-2241/A000012
H. Clarke, A. Kornberg, T. Scotto
Survey research on political efficacy is longstanding. In a number of countries efficacy has been measured using batteries of negatively worded “agree-disagree” statements. In this paper, we investigate the measurement properties of the Canadian variant of this traditional battery and compare its performance with an alternative, positively worded, battery. The research is based on data gathered by a random half-sample experiment administered in the 2004 Political Support in Canada national panel survey. Analyses of these data provide no evidence that negatively framing the statements designed to tap political efficacy is problematic. Rather, it appears that students of political efficacy would have been worse off if they had spent the past several decades conducting analyses employing positively worded variants of the traditional statements. Perhaps most important, scholars have not been misled by acquiescence bias depressing efficacious responses to the traditional battery. These experimental results ind...
{"title":"Accentuating the negative?: A political efficacy question-wording- experiment","authors":"H. Clarke, A. Kornberg, T. Scotto","doi":"10.1027/1614-2241/A000012","DOIUrl":"https://doi.org/10.1027/1614-2241/A000012","url":null,"abstract":"Survey research on political efficacy is longstanding. In a number of countries efficacy has been measured using batteries of negatively worded “agree-disagree” statements. In this paper, we investigate the measurement properties of the Canadian variant of this traditional battery and compare its performance with an alternative, positively worded, battery. The research is based on data gathered by a random half-sample experiment administered in the 2004 Political Support in Canada national panel survey. Analyses of these data provide no evidence that negatively framing the statements designed to tap political efficacy is problematic. Rather, it appears that students of political efficacy would have been worse off if they had spent the past several decades conducting analyses employing positively worded variants of the traditional statements. Perhaps most important, scholars have not been misled by acquiescence bias depressing efficacious responses to the traditional battery. These experimental results ind...","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"6 1","pages":"107-117"},"PeriodicalIF":3.1,"publicationDate":"2010-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292714","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-01-20DOI: 10.1027/1614-2241/A000001
L. A. van der Ark, Jeroen K. Vermunt
In this special issue you will find four papers on handling missing data. All papers have been presented at the 2007 Fall Meeting of Social Science Division of the Dutch Statistical Society (VVS-OR) in Tilburg, The Netherlands. Together, these four papers give an excellent overview of state of the art in missing data analysis. To date, in virtually all fields of the social sciences, researchers are required to deal sophistically with missing data. Ignoring the problem, for example, by simply removing all observations that contain missing data or thoughtlessly applying software that makes the problem go away may lead to seriously biased statistical results and wrong conclusions, and is no longer an option. Instead the researcher must consider the reasons why some of the data are missing and act accordingly. Given that in the social sciences most data are obtained from respondents who responded to tests, questionnaires, surveys, or stimuli in an experimental setting, the first option that comes to mind is approaching those respondents with missing scores again, ask them the reason for their nonresponse, and ask them to respond yet. Unfortunately, this is usually not a realistic option and the researcher must rely on statistical solutions. One way of dealing with missing data is to incorporate the mechanism that caused the missingness into the statistical modeling of the data. In the context of educational measurement, Goegebeur, De Boeck, and Molenberghs (2010) discuss test speededness, which refers to the phenomenon that respondents do not respond to certain items in the test or examination due to a lack of time. They clearly explain how speededness can be incorporated into the statistical model. Using this model-based approach, they show how to identify respondents whose scores were affected by speededness. Advantage of this approach is that it allows the researcher to deal with data that are not missing at random. In some situations, it will not be possible to translate the researcher’s theories on the missingness mechanism into a statistical model because such theories are too complex or not available. Probably the best known strategy to deal with missing data is to assume that the missing scores are missing at random and conduct (multiple) imputation: Replacing the missing scores in the data by plausible values. Two papers discuss imputation methods. First, Van Ginkel, Sijtsma, Van der Ark, and Vermunt (2010) investigated the occurrence of missing data and current practices of handling nonresponse in test and questionnaire data in personality psychology. They found that in the large majority of published research reporting missing data, either the handling of missing data was not discussed, cases with missing values were deleted, or ad hoc procedures were used. In order to improve the use of appropriate methods they proposed using Method Two-Way for handling missing data in test and questionnaire data. Method Two-Way is a multiple imputation t
在本期特刊中，您将找到四篇关于处理丢失数据的论文。所有论文已在荷兰蒂尔堡举行的荷兰统计学会(VVS-OR)社会科学部2007年秋季会议上发表。总之，这四篇论文给出了在缺失数据分析的艺术状态的一个很好的概述。迄今为止，在几乎所有的社会科学领域，研究人员都需要巧妙地处理缺失的数据。忽略这个问题，例如，通过简单地删除所有包含缺失数据的观察结果或轻率地应用使问题消失的软件可能导致严重偏颇的统计结果和错误的结论，并且不再是一种选择。相反，研究人员必须考虑一些数据丢失的原因，并采取相应的行动。考虑到在社会科学中，大多数数据都是从在实验环境中对测试、问卷、调查或刺激做出回应的受访者那里获得的，我想到的第一个选择是再次接近那些分数缺失的受访者，询问他们不回应的原因，并要求他们立即回应。不幸的是，这通常不是一个现实的选择，研究人员必须依靠统计解决方案。处理缺失数据的一种方法是将导致缺失的机制合并到数据的统计建模中。在教育测量的背景下，Goegebeur, De Boeck, and Molenberghs(2010)讨论了测试速度，它是指被调查者由于缺乏时间而对测试或考试中的某些项目不做出反应的现象。他们清楚地解释了如何将速度纳入统计模型。使用这种基于模型的方法，他们展示了如何识别得分受速度影响的受访者。这种方法的优点是它允许研究人员处理不是随机丢失的数据。在某些情况下，将研究人员关于缺失机制的理论转化为统计模型是不可能的，因为这些理论过于复杂或不可用。处理缺失数据的最佳策略可能是假设缺失的分数是随机缺失的，并进行(多重)imputation:用可信的值替换数据中缺失的分数。两篇论文讨论了归算方法。首先，Van Ginkel, Sijtsma, Van der Ark, and vermont(2010)调查了人格心理学中测试和问卷数据中缺失数据的发生和处理无反应的现行做法。他们发现，在绝大多数报告缺失数据的已发表研究中，要么没有讨论对缺失数据的处理，要么删除了缺失值的案例，要么使用了特别程序。为了提高方法的适用性，提出了采用方法双向法处理试验数据和问卷数据中的缺失数据。方法双向是一种容易理解和使用的多重输入。仿真研究表明，对于测试和问卷数据分析中经常使用的统计数据，Method two所获得的结果与技术上更先进的方法所获得的结果相当。在第二篇关于多重输入的论文中，Van Buuren(2010)讨论了完全条件规范来输入缺失值的分数。完全条件规范可以看作是技术上更高级的方法，在R和SPSS等软件包中都有。在一项模拟研究中，Van Buuren(2010)表明，在计算Cronbach 's alpha时，完全条件规范优于Method TwoWay。由于Van Ginkel et al.(2010)和Van Buuren(2010)的论文就Method Two-Way得出了不同的结论，我们认为一些编辑评论是为了解释不同的结果。我们认为这两篇论文都是高质量的，但侧重点不同。首先，Van Buuren(2010)的研究和Van Ginkel等人(2010)的研究中缺失数据的百分比不同。一方面，Van Buuren(2010)使用大缺失百分比(44-78%)比较了方法双向和完全条件规范，在极端情况下，技术上更先进的方法比简单的方法表现出更优越的性能。另一方面，Van Ginkel et al.(2010)表明，在实践中缺失的百分比要低得多(平均9%的响应向量至少有一个缺失观测值)，并参考了缺失百分比在1到20之间的研究，在典型情况下，简单而复杂的方法表现相似。此外，由于缺失率很高，更复杂的贝叶斯版本的双向方法(Van Ginkel, Van der Ark，
{"title":"New Developments in Missing Data Analysis","authors":"L. A. van der Ark, Jeroen K. Vermunt","doi":"10.1027/1614-2241/A000001","DOIUrl":"https://doi.org/10.1027/1614-2241/A000001","url":null,"abstract":"In this special issue you will find four papers on handling missing data. All papers have been presented at the 2007 Fall Meeting of Social Science Division of the Dutch Statistical Society (VVS-OR) in Tilburg, The Netherlands. Together, these four papers give an excellent overview of state of the art in missing data analysis. To date, in virtually all fields of the social sciences, researchers are required to deal sophistically with missing data. Ignoring the problem, for example, by simply removing all observations that contain missing data or thoughtlessly applying software that makes the problem go away may lead to seriously biased statistical results and wrong conclusions, and is no longer an option. Instead the researcher must consider the reasons why some of the data are missing and act accordingly. Given that in the social sciences most data are obtained from respondents who responded to tests, questionnaires, surveys, or stimuli in an experimental setting, the first option that comes to mind is approaching those respondents with missing scores again, ask them the reason for their nonresponse, and ask them to respond yet. Unfortunately, this is usually not a realistic option and the researcher must rely on statistical solutions. One way of dealing with missing data is to incorporate the mechanism that caused the missingness into the statistical modeling of the data. In the context of educational measurement, Goegebeur, De Boeck, and Molenberghs (2010) discuss test speededness, which refers to the phenomenon that respondents do not respond to certain items in the test or examination due to a lack of time. They clearly explain how speededness can be incorporated into the statistical model. Using this model-based approach, they show how to identify respondents whose scores were affected by speededness. Advantage of this approach is that it allows the researcher to deal with data that are not missing at random. In some situations, it will not be possible to translate the researcher’s theories on the missingness mechanism into a statistical model because such theories are too complex or not available. Probably the best known strategy to deal with missing data is to assume that the missing scores are missing at random and conduct (multiple) imputation: Replacing the missing scores in the data by plausible values. Two papers discuss imputation methods. First, Van Ginkel, Sijtsma, Van der Ark, and Vermunt (2010) investigated the occurrence of missing data and current practices of handling nonresponse in test and questionnaire data in personality psychology. They found that in the large majority of published research reporting missing data, either the handling of missing data was not discussed, cases with missing values were deleted, or ad hoc procedures were used. In order to improve the use of appropriate methods they proposed using Method Two-Way for handling missing data in test and questionnaire data. Method Two-Way is a multiple imputation t","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"6 1","pages":"1-2"},"PeriodicalIF":3.1,"publicationDate":"2010-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1027/1614-2241/A000001","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292586","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-01-20DOI: 10.1027/1614-2241/A000003
J. V. van Ginkel, K. Sijtsma, L. A. van der Ark, J. Vermunt
The focus of this study was the incidence of different kinds of missing-data problems in personality research and the handling of these problems. Missing-data problems were reported in approximately half of more than 800 articles published in three leading personality journals. In these articles, unit nonresponse, attrition, and planned missingness were distinguished but missing item scores in trait measurement were reported most frequently. Listwise deletion was the most frequently used method for handling all missing-data problems. Listwise deletion is known to reduce the accuracy of parameter estimates and the power of statistical tests and often to produce biased statistical analysis results. This study proposes a simple alternative method for handling missing item scores, known as two-way imputation, which leaves the sample size intact and has been shown to produce almost unbiased results based on multi-item questionnaire data.
{"title":"Incidence of Missing Item Scores in Personality Measurement, and Simple Item-Score Imputation","authors":"J. V. van Ginkel, K. Sijtsma, L. A. van der Ark, J. Vermunt","doi":"10.1027/1614-2241/A000003","DOIUrl":"https://doi.org/10.1027/1614-2241/A000003","url":null,"abstract":"The focus of this study was the incidence of different kinds of missing-data problems in personality research and the handling of these problems. Missing-data problems were reported in approximately half of more than 800 articles published in three leading personality journals. In these articles, unit nonresponse, attrition, and planned missingness were distinguished but missing item scores in trait measurement were reported most frequently. Listwise deletion was the most frequently used method for handling all missing-data problems. Listwise deletion is known to reduce the accuracy of parameter estimates and the power of statistical tests and often to produce biased statistical analysis results. This study proposes a simple alternative method for handling missing item scores, known as two-way imputation, which leaves the sample size intact and has been shown to produce almost unbiased results based on multi-item questionnaire data.","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"6 1","pages":"17-30"},"PeriodicalIF":3.1,"publicationDate":"2010-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57293131","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-01-20DOI: 10.1027/1614-2241/A000004
S. Buuren
Imputation of incomplete questionnaire items should preserve the structure among items and the correlations between scales. This paper explores the use of fully conditional specification (FCS) to impute missing data in questionnaire items. FCS is particularly attractive for items because it does not require (1) a specification of the number of factors or classes, (2) a specification of which item belongs to which scale, and (3) assumptions about conditional independence among items. Imputation models can be specified using standard features of the R package MICE 1.16. A limited simulation shows that MICE outperforms two-way imputation with respect to Cronbach’s α and the correlations between scales. We conclude that FCS is a promising alternative for imputing incomplete questionnaire items.
{"title":"Item Imputation Without Specifying Scale Structure","authors":"S. Buuren","doi":"10.1027/1614-2241/A000004","DOIUrl":"https://doi.org/10.1027/1614-2241/A000004","url":null,"abstract":"Imputation of incomplete questionnaire items should preserve the structure among items and the correlations between scales. This paper explores the use of fully conditional specification (FCS) to impute missing data in questionnaire items. FCS is particularly attractive for items because it does not require (1) a specification of the number of factors or classes, (2) a specification of which item belongs to which scale, and (3) assumptions about conditional independence among items. Imputation models can be specified using standard features of the R package MICE 1.16. A limited simulation shows that MICE outperforms two-way imputation with respect to Cronbach’s α and the correlations between scales. We conclude that FCS is a promising alternative for imputing incomplete questionnaire items.","PeriodicalId":18476,"journal":{"name":"Methodology: European Journal of Research Methods for The Behavioral and Social Sciences","volume":"6 1","pages":"31-36"},"PeriodicalIF":3.1,"publicationDate":"2010-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"57292689","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：