The Sensitivity of Confirmatory Maximum Likelihood Factor Analysis to Violations of Measurement Scale and Distributional Assumptions

1987 ◽  
Vol 24 (2) ◽  
pp. 222-228 ◽  
Author(s):  
Emin Babakus ◽  
Carl E. Ferguson ◽  
Karl G. Jöreskog
Keyword(s):  
Factor Analysis ◽  
Maximum Likelihood ◽  
Ordinal Data ◽  
Large Scale ◽  
Goodness Of Fit ◽  
Convergence Rates ◽  
Measurement Scale ◽  
Standard Errors ◽  
Model Parameters ◽  
Distributional Assumptions

A large-scale simulation design was used to study the sensitivity of maximum likelihood (ML) factor analysis to violations of measurement scale and distributional assumptions in the input data. Product-moment, polychoric. Spearman's rho, and Kendall's tau- b correlations computed from ordinal data were used to estimate a single-factor model. The resulting ML estimates were compared on the bases of convergence rates and improper solutions, accuracy of the loading estimates, fit statistics, and estimated standard errors. The LISREL maximum likelihood solution algorithm was used to estimate model parameters. The polychoric correlation procedure was found to provide the most accurate estimates of pairwise correlations and factor loadings but performed worst on all goodness-of-fit criteria. LISREL overestimated all standard errors, thus not reflecting the effects of standardization as previously assumed. When the data were categorized, the polychoric correlations led to the best estimates of the standard errors.

scholarly journals Fitting Large Factor Analysis Models With Ordinal Data

2018 ◽  
Vol 79 (3) ◽  
pp. 417-436 ◽  
Author(s):  
Christine DiStefano ◽  
Heather L. McDaniel ◽  
Liyun Zhang ◽  
Dexin Shi ◽  
Zhehan Jiang
Keyword(s):  
Factor Analysis ◽  
Ordinal Data ◽  
Goodness Of Fit ◽  
Weighted Least Squares ◽  
Standard Errors ◽  
Parameter Estimates ◽  
Fit Indices ◽  
Ordered Categories ◽  
Size Number ◽  
Analysis Models

A simulation study was conducted to investigate the model size effect when confirmatory factor analysis (CFA) models include many ordinal items. CFA models including between 15 and 120 ordinal items were analyzed with mean- and variance-adjusted weighted least squares to determine how varying sample size, number of ordered categories, and misspecification affect parameter estimates, standard errors of parameter estimates, and selected fit indices. As the number of items increased, the number of admissible solutions and accuracy of parameter estimates improved, even when models were misspecified. Also, standard errors of parameter estimates were closer to empirical standard deviation values as the number of items increased. When evaluating goodness-of-fit for ordinal CFA with many observed indicators, researchers should be cautious in interpreting the root mean square error of approximation, as this value appeared overly optimistic under misspecified conditions.

scholarly journals Preliminary development of the social media disinformation scale (SMDS-12) and its association with social media addiction and mental health: COVID-19 as a pilot case study (Preprint)

2021 ◽  
Author(s):  
Noomen Guelmami ◽  
Maher ben Khalifa ◽  
Nasr Chalghaf ◽  
Jude Dzevela Kong ◽  
Tannoubi Amayra ◽  
...  
Keyword(s):  
Mental Health ◽  
Social Media ◽  
Factor Analysis ◽  
Internet Addiction ◽  
Goodness Of Fit ◽  
Three Dimensions ◽  
Measurement Scale ◽  
Dimensional Structure ◽  
The Internet ◽  
Wide Range

BACKGROUND In recent years, online disinformation has increased. An infodemic has spread around the COVID-19 pandemic. Since January 2020, the culprits and antidotes to disinformation have been digital and social media. OBJECTIVE Our study aimed to develop and test the psychometric properties of the SMDS-12 measurement scale which assesses the consumption, confidence, and sharing of information related to covid-19 by social media users. METHODS A total of 874 subjects recruited over two exploratory (n = 179, Mean age = 29.34, SD = 7.98) and confirmatory (n = 695, Mean age = 31.22, SD = 11.63) periods, completed thesocial media disinformation scale (SMDS-12),the Internet addiction test (IAT), the COVID-19 fear scale, and the perceived stress questionnaire.The 12-item scale (SMDS-12 ) was initially tested by exploratory factor analysis. RESULTS The test supported the three-dimensional structure, in addition, no items were removed from the measurement scale. Subsequently, confirmatory factor analysis confirmed the robustness of the measure by referring to a wide range of goodness-of-fit indices that met the recommended standards. The reliability of the instrument examined by means of three internal consistency indices demonstrated that the three dimensions of the instrument are reliable.The correlation between the instrument's dimensions with the internet addiction scale and mental health factors showed positive associations. CONCLUSIONS The scale is eligible for measuring the credibility of disinformation and can be adapted to measure the credibility of social media disinformation in other contexts.

scholarly journals EVALUATING MEASUREMENT PROPERTIES OF ATTITUDINAL ITEMS RELATED TO LEARNING SCIENCE IN TAIWAN FROM TIMSS 2007

2014 ◽  
Vol 13 (6) ◽  
pp. 856-869
Author(s):  
Pey-Yan Liou
Keyword(s):  
Factor Analysis ◽  
Ordinal Data ◽  
Large Scale ◽  
Empirical Studies ◽  
Integrated Approach ◽  
Measurement Properties ◽  
Item Response Model ◽  
Science Study ◽  
Timss 2007 ◽  
Mathematics And Science

In the recent empirical studies utilizing existing items and derived variables of international large-scale assessment (ILSA) data, the three major methodological deficiencies, including the use of a single item to define a construct, the statistical properties of ordinal data, and the fitness of the measurement structure for different scenarios, are examined. To overcome these issues, this study proposes an integrated approach to evaluating items and constructing derived variables in a given situation. Exploratory factor analysis, confirmatory factor analysis, and the item response model are utilized to evaluate student attitudinal items and derived variables from the Trends in International Mathematics and Science Study (TIMSS) 2007 Taiwanese data. The results suggest that the three-factor model composed of 12 items is optimal for the data, not the default factor structure in the database. The implications of evaluating items and creating derived variables from ILSA data for the education research community are also discussed. Key words: attitudinal items, factor analysis, Trends in International Mathematics and Science Study.

scholarly journals The ArcTan Lomax Distribution with Properties and Applications

2021 ◽  
pp. 117-125
Author(s):  
Arun Kumar Chaudhary ◽  
Vijay Kumar
Keyword(s):  
Survival Analysis ◽  
Maximum Likelihood ◽  
Maximum Likelihood Estimation ◽  
Goodness Of Fit ◽  
Continuous Distribution ◽  
Likelihood Estimation ◽  
Model Parameters ◽  
Lomax Distribution ◽  
And Mathematics

Here, in this paper, a continuous distribution called ArcTan Lomax distribution with three-parameter has been introduced along with some relevant properties of statistics and mathematics pertaining to the distribution. With the help of three established estimations methods including maximum likelihood estimation (MLE), estimation of the presented distribution’s model parameters is done. Also with the help of a real set of data, the distribution’s goodness-of-fit is examined in contrast to some established models in survival analysis.

Sign in / Sign up

Export Citation Format

Share Document