Katerina Marcoulides

Dr. Katerina Marcoulides

orcid

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 September 2019 in Structural Equation Modeling: A Multidisciplinary Journal
Verified
Journal Article
Closed access
KMKaterina Marcoulides
Katerina Marcoulides
NFNjål Foldnes
Njål Foldnes
SGSteffen Grønneberg
Steffen Grønneberg
The assessment of model fit has received widespread interest by researchers in the structural equation modeling literature for many years. Various model fit test statistics have been suggested for conducting this assessment. Selecting an appropriate test statistic in order to evaluate model fit, however, can be difficult as the selection depends on the distributional characteristics of the sampled data, the magnitude of the sample size, and/or the proposed model features. The purpose of this paper is to present a selection procedure that can be used to algorithmically identify the best test statistic and simplify the whole assessment process. The procedure is illustrated using empirical data along with an easy to use computerized implementation.

24 February 2021 in Structural Equation Modeling: A Multidisciplinary Journal
Verified
Journal Article
Closed access
WLWalter L. Leite
Walter L. Leite
ZSZuchao Shen
Zuchao Shen
KMKaterina Marcoulides
Katerina Marcoulides
CFCharles L. Fisk
Charles L. Fisk
JHJeffrey Harring
Jeffrey Harring
Studies using structural equation modeling (SEM) to evaluate theories against observed data rely on multiple sources of evidence to support a proposed model, such as fit indices, variance explained, and comparison of alternative models. Additional evidence can be obtained by evaluating the model results’ sensitivity to an omitted confounder. The phantom variable approach to SEM sensitivity analysis requires manual specification of sensitivity parameters. This study improves on the phantom variable approach by employing the ant colony optimization algorithm to automatically search for sensitivity parameters, if any, that would lead to a change in the study’s conclusions. The proposed method is implemented in the package SEMsens for the R statistical software, and demonstrated with a sensitivity analysis of a model of the complex relation between working memory and writing.

13 July 2023 in Personality and Individual Differences
Verified
Journal Article
Open access
NTNathan T. Torunsky
Nathan T. Torunsky
SKSara Knauz
Sara Knauz
IVIris Vilares
Iris Vilares
KMKaterina Marcoulides
Katerina Marcoulides
WKWilma Koutstaal
Wilma Koutstaal
Alexithymia is a clinically relevant personality trait characterized by poor emotional awareness and associated with several psychological and physical health concerns. Individuals with high alexithymia tend to engage in experiential avoidance and this may mediate psychological distress. However, little is known about what specific processes of experiential avoidance are involved, and the nature of the relation between alexithymia, experiential avoidance, and psychological distress remains unclear at a latent construct level. To examine this relationship at the latent construct level, a representative sample of 693 U.S. adults completed alexithymia (TAS-20, BVAQ, PAQ), general distress (DASS-21), multi-dimensional experiential avoidance (MEAQ), and general health (PROMIS-G-10) questionnaires. Structural equation modeling revealed that alexithymia significantly predicted experiential avoidance (β = 0.966, t = 82.383, p < .01), experiential avoidance significantly predicted general distress (β = 0.810, t = 2.017, p < .05), and experiential avoidance fully mediated the relationship between alexithymia and general distress (βindirect = -0.159, t = -0.398, p > .05). Correlations between alexithymia and experiential avoidance subfactors revealed a strong relationship to the repression and denial subfactor. Experiential avoidance is a promising target for clinical interventions, though longitudinal research is necessary to elucidate how the relationship between alexithymia and experiential avoidance unfolds over time.

28 July 2020 in International Journal of Research & Method in Education
Verified
Journal Article
Closed access
KMKaterina Marcoulides
Katerina Marcoulides
KYKe-Hai Yuan
Ke-Hai Yuan
Multilevel structural equation models (MSEM) are typically evaluated on the basis of goodness of fit indices. A problem with these indices is that they pertain to the entire model, reflecting simultaneously the degree of fit for all levels in the model. Consequently, in cases that lack model fit, it is unclear which level model is misspecified. Although some researchers have suggested that model fit be assessed separately at all levels, this is not always ideal as it depends on the choice of fit indices, cutoff points, and magnitude of the intraclass correlation. Equivalence testing has recently been recommended as a superior way to evaluate the goodness of fit of models. The approach has also been heralded as a way to advance the inferential nature of structural equation modeling as a confirmatory tool. The purpose of this paper is to illustrate the use of equivalence testing for assessing multilevel model fit in educational research. Although the described procedures can be easily generalized to models with more than two levels, for simplicity we explicitly consider only two-level models.

18 March 2022 in Structural Equation Modeling: A Multidisciplinary Journal
Verified
Journal Article
Closed access
ZJZeyuan Jing
Zeyuan Jing
HKHuan KUANG
Huan KUANG
WLWalter L. Leite
Walter L. Leite
KMKaterina Marcoulides
Katerina Marcoulides
CFCharles L. Fisk
Charles L. Fisk
Model specification is a crucial aspect of structural equation modeling (SEM), since a misspecified model may lead to biased parameter estimation and result in inaccurate conclusions. We propose the Hybrid Ant Colony Optimization Algorithm (hACO), an improved metaheuristic algorithm to conduct model specification searches in SEM. This data mining algorithm combines aspects of the Ant Colony Optimization algorithm with the Tabu search algorithm to increase both accuracy and efficiency. A Monte Carlo simulation study showed that the hACO algorithm provided accurate and efficient SEM specification searches across all designed simulation conditions. The hACO algorithm can help applied researchers conduct specification searches while avoiding potential model misspecifications.

25 August 2021 in The Journal of Sex Research
Verified
Journal Article
Closed access
JKJan-Louw Kotze
Jan-Louw Kotze
PFPatricia A. Frazier
Patricia A. Frazier
KHKayla A. Huber
Kayla A. Huber
KMKaterina Marcoulides
Katerina Marcoulides
KLKatherine A. Lust
Katherine A. Lust
Sexual harassment and its negative consequences continue to affect a large percentage of higher education students in the US. Previous research has identified a limited number of harassment risk factors, and has generally not examined them in combination. In this study, an expanded set of individual, relationship, and community-level risk factors were examined using hurdle models and classification and regression tree (CART) analyses to identify key risk factors for peer and faculty/staff sexual harassment. Secondary data analysis was conducted using data from a sample of 9,285 students from 18 two-year and four-year schools in Minnesota. CART analyses indicated that, for peer sexual harassment, being younger; consuming alcohol more than once a month; attending a four-year school; being transgender, genderqueer, self-identified, or a cisgender woman; and having experienced bullying were the most important risk factors for peer harassment on campus. For faculty/staff harassment, being gay, lesbian, bisexual, questioning, or having a self-identified sexual orientation was the most important risk factor. These and other risk factors were significant in the hurdle models. More research is needed to understand why these factors are associated with harassment. Limitations and implications for prevention programming at higher education institutions are discussed.