scholarly journals Machine Learning-Based Estimation and Goodness-of-Fit for Large-Scale Confirmatory Item Factor Analysis

2021 ◽  
Author(s):  
Christopher John Urban ◽  
Daniel J. Bauer
Keyword(s):  
Machine Learning ◽  
Factor Analysis ◽  
Parameter Estimation ◽  
Sample Size ◽  
Large Scale ◽  
Goodness Of Fit ◽  
Synthetic Data ◽  
Latent Factors ◽  
Item Factor Analysis ◽  
Fit Index

We investigate novel parameter estimation and goodness-of-fit (GOF) assessment methods for large-scale confirmatory item factor analysis (IFA) with many respondents, items, and latent factors. For parameter estimation, we extend Urban and Bauer's (2021) deep learning algorithm for exploratory IFA to the confirmatory setting by showing how to handle user-defined constraints on loadings and factor correlations. For GOF assessment, we explore new simulation-based tests and indices. In particular, we consider extensions of the classifier two-sample test (C2ST), a method that tests whether a machine learning classifier can distinguish between observed data and synthetic data sampled from a fitted IFA model. The C2ST provides a flexible framework that integrates overall model fit, piece-wise fit, and person fit. Proposed extensions include a C2ST-based test of approximate fit in which the user specifies what percentage of observed data can be distinguished from synthetic data as well as a C2ST-based relative fit index that is similar in spirit to the relative fit indices used in structural equation modeling. Via simulation studies, we first show that the confirmatory extension of Urban and Bauer's (2021) algorithm produces more accurate parameter estimates as the sample size increases and obtains comparable estimates to a state-of-the-art confirmatory IFA estimation procedure in less time. We next show that the C2ST-based test of approximate fit controls the empirical type I error rate and detects when the number of latent factors is misspecified. Finally, we empirically investigate how the sampling distribution of the C2ST-based relative fit index depends on the sample size.

scholarly journals A Deep Learning Algorithm for High-Dimensional Exploratory Item Factor Analysis

2020 ◽  
Author(s):  
Christopher John Urban ◽  
Daniel J. Bauer
Keyword(s):  
Factor Analysis ◽  
Deep Learning ◽  
Maximum Likelihood ◽  
Sample Size ◽  
Learning Algorithm ◽  
Marginal Maximum Likelihood ◽  
Latent Factors ◽  
Estimation Procedures ◽  
Item Factor Analysis ◽  
Fast Procedure

Marginal maximum likelihood (MML) estimation is the preferred approach to fitting item response theory models in psychometrics due to the MML estimator's consistency, normality, and efficiency as the sample size tends to infinity. However, state-of-the-art MML estimation procedures such as the Metropolis-Hastings Robbins-Monro (MH-RM) algorithm as well as approximate MML estimation procedures such as variational inference (VI) are computationally time-consuming when the sample size and the number of latent factors are very large. In this work, we investigate a deep learning-based VI algorithm for exploratory item factor analysis (IFA) that is computationally fast even in large data sets with many latent factors. The proposed approach applies a deep artificial neural network model called a variational autoencoder for exploratory IFA. An importance sampling technique to help the variational estimator better approximate the MML estimator is explored. We provide a real data application that recovers results aligning with psychological theory across random starts. Via simulation studies, we empirically demonstrate that the variational estimator is consistent (although factor correlation estimates exhibit some bias) and yields similar results to MH-RM in less time. Our simulations also suggest that the proposed approach performs similarly to and is potentially faster than constrained joint maximum likelihood estimation, a fast procedure that is consistent when the sample size and the number of items simultaneously tend to infinity.

scholarly journals Urdu Translation and Validation of 11-Item Measure to Assess Borderline Personality Features in Pakistani Adolescents

SAGE Open ◽  
2021 ◽  
Vol 11 (1) ◽  
pp. 215824402098615
Author(s):  
Humaira Bibi ◽  
Syeda Farhana Kazmi
Keyword(s):  
Factor Analysis ◽  
Psychometric Properties ◽  
Convergent Validity ◽  
Goodness Of Fit ◽  
Borderline Personality ◽  
Two Dimensions ◽  
Personality Features ◽  
Fit Index ◽  
Borderline Personality Features ◽  
Translation And Validation

The current study includes Urdu translation and validation of Borderline Personality Features Scale–11 (BPFS-11) in two phases. Phase 1 included forward and back translation of BPFS-11, and Phase 2 included establishment of psychometric properties for BPFS-11. For this purpose, 930 adolescents were selected from different hospitals, schools, and colleges. The reliability value of the scale was .72. Exploratory factor analysis revealed factor structure with four principal dimensions; besides confirmatory factor analysis, goodness-of-fit indices indicated good fit of model to data, and two dimensions of scale and factors showed good values of internal consistency. The obtained value for goodness-of-fit index was .995, for adjusted goodness-of-fit index was .989, for comparative fit index was .998, for incremental fit index was .998, and for root mean square error of approximation (RMSEA) value was .019. Good values of composite reliability and convergent validity were measured for both dimensions of the scale. The analysis of criterion-related validity showed significant positive correlation of BPFS-11 with Affective Lability Scale, Deliberate Self-Harm Inventory, and neuroticism scale of Big Five Inventory. Significant differences were found between scores of individuals having borderline personality disorder and scores of normal individuals. The results of the current study indicated that BPFS-11 is short and easily administered diagnostic tool that has good psychometric properties and can be helpful for diagnosis of borderline personality features in adolescents. It can enhance the understanding of the participants regarding the statements of the scale for Urdu natives.

scholarly journals Critical Multicultural Education Competencies Scale: A Scale Development Study

2016 ◽  
Vol 5 (3) ◽  
pp. 51
Author(s):  
Yasemin Acar-Ciftci
Keyword(s):  
Factor Analysis ◽  
Critical Race Theory ◽  
Multicultural Education ◽  
Goodness Of Fit ◽  
Critical Multicultural Education ◽  
Confirmatory Factor ◽  
Education Theory ◽  
Deductive Method ◽  
Coefficient Of Concordance ◽  
Fit Index

<p>The purpose of this study is to develop a scale in order to identify the critical mutlicultural education competencies of teachers. For this reason, first of all, drawing on the knowledge in the literature, a new conceptual framework was created with deductive method based on critical theory, critical race theory and critical multicultural education theory, which includes dimensions of awareness, knowledge, attitude and skill. In accordance with this framework, experimental form consisting of 56 items was submitted to experts for consideration. In accordance with the responses of the experts, content validity rate of the items was identified and the items which were below. 80 level were excluded from the study. The pilot study form consisting of 45 items, was applied to teachers who work preschools, primary and secondary school and the data which was obtained from 421 teachers in total were analyzed. Through the Exploratory Factor Analysis (EFA), a structure consisting of “Awareness”, “Attitude’’, “Knowledge” and “Skill” and 42 items was reached. The relationship between sub-dimensions of the scale was examined and it was observed that the factors were positively and significantly correlated with each other. In this case, it was concluded that scale supports the theory. After the analysis, it was confirmed that the sub-dimensions were the components of a structure called critical multicultural education competency and that together they form a higher structure. It was determined that the goodness of fit index of the model is quite high. Confirmatory Factor Analysis also confirmed the results of EFA. The internal coefficient of concordance was determined as .845 for the whole scale.</p>

scholarly journals Resampled dimensional reduction for feature representation in machine learning

2021 ◽  
Author(s):  
Herdiantri Sufriyana ◽  
Yu Wei Wu ◽  
Emily Chia-Yu Su
Keyword(s):  
Machine Learning ◽  
Parameter Estimation ◽  
Prediction Model ◽  
Sample Size ◽  
Dimensional Reduction ◽  
Latent Variables ◽  
Feature Representation ◽  
Estimated Parameters ◽  
Representation Technique ◽  
Selection Of

Abstract We aimed to provide a resampling protocol for dimensional reduction resulting a few latent variables. The applicability focuses on but not limited for developing a machine learning prediction model in order to improve the number of sample size in relative to the number of candidate predictors. By this feature representation technique, one can improve generalization by preventing latent variables to overfit data used to conduct the dimensional reduction. However, this technique may warrant more computational capacity and time to conduct the procedure. The key stages consisted of derivation of latent variables from multiple resampling subsets, parameter estimation of latent variables in population, and selection of latent variables transformed by the estimated parameters.

A validation of the performance management scale

2018 ◽  
Vol 35 (6) ◽  
pp. 1253-1267 ◽  
Author(s):  
Khahan Na-nan ◽  
Kanokporn Chaiprasit ◽  
Peerapong Pukkeeree
Keyword(s):  
Factor Analysis ◽  
Performance Evaluation ◽  
Performance Management ◽  
Goodness Of Fit ◽  
Mean Square ◽  
Evaluation Performance ◽  
Content Type ◽  
Performance Review ◽  
And Performance ◽  
Fit Index

Purpose The purpose of this paper is to develop a performance management (PM) scale questionnaire that encompasses the pre-requisite, performance planning, performance evaluation, performance review, and performance application dimensions of PM. Design/methodology/approach In the study, the 33 questionnaire questions were first validated using exploratory factor analysis (EFA) and then by confirmatory factor analysis (CFA) along the three performance dimensions. The research sample consists of 330 entrepreneurs. The factor analysis results confirm the validity of the questionnaire as a reliable entrepreneur PM evaluation tool, as evidenced by the composite reliability of 0.845 and the average variance extracted of 0.532. Findings All constructs revealed the acceptable internal consistency reliability. A good model fit was found for the measurement model using several fit index like χ2=449.983, degree of freedom=415, p-value (p)=0.114, goodness of fit index=0.927, adjusted goodness of fit index=0.901, root mean square error of approximation=0.016, and root of mean square residuals=0.032. Research limitations/implications The PM model was examined using EFA and CFA only. A sample with only SMEs entrepreneurs and large sample size and sample area can be used in future research. Practical implications This research paper is an endeavor to explore only the reliability and validity of the PM model. Thus all the five dimension, namely “pre-requisite” “performance planning,” “performance evaluation,” “performance review,” and “performance application” proved out of be reliable and validated when it will be tested in case of SMEs’ high-growth sectors and high-impact sectors. Originality/value The main contribution of this research is that all factors have a good fit and acceptable reliability value; each factor can be measured individually depending on the nature of the research.

The Self-Transcendence Scale

2011 ◽  
Vol 30 (3) ◽  
pp. 147-159 ◽  
Author(s):  
Gørill Haugan ◽  
Toril Rannestad ◽  
Helge Garåsen ◽  
Randi Hammervold ◽  
Geir Arild Espnes
Keyword(s):  
Factor Analysis ◽  
Confirmatory Factor Analysis ◽  
Goodness Of Fit ◽  
Factor Model ◽  
Well Being ◽  
The Self ◽  
Mean Square ◽  
Confirmatory Factor ◽  
Self Transcendence ◽  
Fit Index

Purpose: Self-transcendence, the ability to expand personal boundaries in multiple ways, has been found to provide well-being. The purpose of this study was to examine the dimensionality of the Norwegian version of the Self-Transcendence Scale, which comprises 15 items. Background: Reed’s empirical nursing theory of self-transcendence provided the theoretical framework; self-transcendence includes an interpersonal, intrapersonal, transpersonal, and temporal dimension. Design: Cross-sectional data were obtained from a sample of 202 cognitively intact elderly patients in 44 Norwegian nursing homes. Results: Exploratory factor analysis revealed two and four internally consistent dimensions of self-transcendence, explaining 35.3% (two factors) and 50.7% (four factors) of the variance, respectively. Confirmatory factor analysis indicated that the hypothesized two- and four-factor models fitted better than the one-factor model (c x2, root mean square error of approximation, standardized root mean square residual, normed fit index, nonnormed fit index, comparative fit index, goodness-of-fit index, and adjusted goodness-of-fit index). Conclusions: The findings indicate self-transcendence as a multifactorial construct; at present, we conclude that the two-factor model might be the most accurate and reasonable measure of self-transcendence. Implications: This research generates insights in the application of the widely used Self-Transcendence Scale by investigating its psychometric properties by applying a confirmatory factor analysis. It also generates new research-questions on the associations between self-transcendence and well-being.

scholarly journals Novel GIS Based Machine Learning Algorithms for Shallow Landslide Susceptibility Mapping

Sensors ◽  
2018 ◽  
Vol 18 (11) ◽  
pp. 3777 ◽  
Author(s):  
Ataollah Shirzadi ◽  
Karim Soliamani ◽  
Mahmood Habibnejhad ◽  
Ataollah Kavian ◽  
Kamran Chapi ◽  
...  
Keyword(s):  
Machine Learning ◽  
Sample Size ◽  
Prediction Accuracy ◽  
Goodness Of Fit ◽  
Learning Algorithm ◽  
Machine Learning Algorithms ◽  
Landslide Susceptibility Mapping ◽  
Sample Sizes ◽  
Promising Alternative ◽  
Statistical Measures

The main objective of this research was to introduce a novel machine learning algorithm of alternating decision tree (ADTree) based on the multiboost (MB), bagging (BA), rotation forest (RF) and random subspace (RS) ensemble algorithms under two scenarios of different sample sizes and raster resolutions for spatial prediction of shallow landslides around Bijar City, Kurdistan Province, Iran. The evaluation of modeling process was checked by some statistical measures and area under the receiver operating characteristic curve (AUROC). Results show that, for combination of sample sizes of 60%/40% and 70%/30% with a raster resolution of 10 m, the RS model, while, for 80%/20% and 90%/10% with a raster resolution of 20 m, the MB model obtained a high goodness-of-fit and prediction accuracy. The RS-ADTree and MB-ADTree ensemble models outperformed the ADTree model in two scenarios. Overall, MB-ADTree in sample size of 80%/20% with a resolution of 20 m (area under the curve (AUC) = 0.942) and sample size of 60%/40% with a resolution of 10 m (AUC = 0.845) had the highest and lowest prediction accuracy, respectively. The findings confirm that the newly proposed models are very promising alternative tools to assist planners and decision makers in the task of managing landslide prone areas.

scholarly journals Measurement of Stigma in Men Who Have Sex with Men in Hanoi, Vietnam: Assessment of a Homosexuality-Related Stigma Scale

2013 ◽  
Vol 2013 ◽  
pp. 1-9 ◽  
Author(s):  
Huy Ha ◽  
Michael W. Ross ◽  
Jan M. H. Risser ◽  
Huong T. M. Nguyen
Keyword(s):  
Factor Analysis ◽  
Degrees Of Freedom ◽  
Goodness Of Fit ◽  
Discriminant Validity ◽  
Cross Sectional ◽  
Stigma Scale ◽  
Convergent And Discriminant Validity ◽  
Related Stigma ◽  
Sex With Men ◽  
Fit Index

Objective. To develop and assess a homosexuality-related stigma scale among men who have sex with men (MSM) in Hanoi, Vietnam. Methods. We conducted a cross-sectional study using respondent-driven sampling in Hanoi, Vietnam, in 2011. We used a cross-validation approach. Factor analysis was performed, and interitem correlation matrices were constructed to identify the latent factor structures, examine the goodness of fit, and assess convergent and discriminant validity of the determined scales. Internal consistency checks were performed in split samples and whole sample, and separately for each determined factor. Results. The findings were consistent in split samples. Three homosexuality-related stigma factors were identified: enacted homosexual stigma, perceived homosexual stigma, and internalized homosexual stigma. The fit indices of the confirmatory factor analysis in both split samples supported the hypothesized three-factor structures (in subsamples A and B: χ2/degrees of freedom ratio = 1.77 and 1.59, nonnormed fit index = 0.92 and 0.94, comparative fit index = 0.93 and 0.95, and the root mean square of approximation = 0.06 and 0.05, resp.). The interitem correlation supported the convergent and discriminant validity of the scales. The reliability of the three scales indicated good consistency (Cronbach’s alpha: 0.79–0.84) across split samples and for the whole data. Conclusion. Our scales have good psychometric properties for measuring homosexuality-related stigma. These comprehensive and practical tools are crucial not only to assess stigma against MSM and its consequence, but also to guide the development of interventions targeting MSM, as well as to evaluate the efficacy of existing stigma reduction efforts in Vietnam and other countries with similar settings.

scholarly journals Development of a questionnaire assessing nursing staff’s knowledge, attitude, and practice on the prevention of the nosocomial infection in elderly patients: testing reliability and validity†

2021 ◽  
Vol 8 (2) ◽  
pp. 141-151
Author(s):  
Fu-Lin Cai ◽  
Xiu-Feng Chen ◽  
Yong-Xin Wang
Keyword(s):  
Factor Analysis ◽  
Nosocomial Infection ◽  
Elderly Patients ◽  
Goodness Of Fit ◽  
Reliability And Validity ◽  
Mean Square ◽  
Attitude And Practice ◽  
Confirmatory Factor ◽  
Fit Index ◽  
Knowledge Attitude And Practice

Abstract Objective To develop a questionnaire assessing nursing staff’s knowledge, attitude, and practice on the prevention of the nosocomial infection in elderly patients and test its reliability and validity. Methods After the drafted questionnaire was developed, two rounds of Delphi survey were conducted by consulting experts to improve the questionnaire. Subsequently, 700 copies of the questionnaire were distributed to nursing staff to assess its reliability and validity. Results Exploratory factor analysis (EFA) identifies 3 aspects, namely knowledge, attitude, and practice, with a total of 38 items. The Cronbach’s α coefficients of the questionnaire and each of the aspects are 0.85, 0.80, 0.886, and 0.77 (>0.7), respectively. Confirmatory factor analysis (CFA) of each of the aspects are c2/df = 3.99, 2.26, and 3.32; Goodness-of-fit index (GFI) = 0.91, 0.97, and 0.92; Root mean square error of approximation (RMSEA) = 0.06, 0.04, and 0.05; Comparative fit index (CFI) = 0.91, 0.96, and 0.90. Conclusions Through this study, it can be ascertained whether the developed questionnaire enjoys sound reliability and validity in assessing nursing staff’s knowledge, attitude, and practice on preventing the nosocomial infection in elderly patients and thus has certain application value.

scholarly journals THE DEVELOPMENT OF STUDENTS’ INTRAPERSONAL INSTRUMENT OF ELECTRICAL ENGINEERING PROGRAM VOCATIONAL SCHOOLS IN BALI

2017 ◽  
Vol 3 (1) ◽  
pp. 1-9
Author(s):  
I Gede Ratnaya ◽  
Gaguk Margono
Keyword(s):  
Factor Analysis ◽  
Goodness Of Fit ◽  
Electrical Engineering ◽  
P Value ◽  
Vocational High School ◽  
Content Validation ◽  
Chi Square ◽  
Engineering Program ◽  
Product Engineering ◽  
Fit Index

This research aimed to develop an instrument to measure the intrapersonal students’ of skill Electrical Engineering Program at Vocational High School in Bali by using Likert scale. This capability is important in regulating and monitoring personal’s goals during vocational education at SMK. This instrument has been tested to 110 students at Engineering Utilization of Electricity Program in the entire province of Bali. The validation of the instrument through the content validation by the experts, the validation grains measure with Momment Product engineering, and the validation of the construct done by factor analysis. Factor analysis uses the confirmation method Maximum Likelihood (ML) with the conformity or suitability obtained Chi Square amounted to 113,8 (p-value = 0,2622) and Goodness of Fit Index (GFI) by 0.92.

Sign in / Sign up

Export Citation Format

Share Document