An exploratory analysis of the latent structure of process data via action sequence autoencoders

Computer simulations have become a popular tool for assessing complex skills such as problem-solving. Log files of computer-based items record the human–computer interactive processes for each respondent in full. The response processes are very diverse, noisy, and of non-standard formats. Few generic methods have been developed to exploit the information contained in process data. In this paper we propose a method to extract latent variables from process data. The method utilizes a sequence-to-sequence autoencoder to compress response processes into standard numerical vectors. It does not require prior knowledge of the specific items and human–computer interaction patterns. The proposed method is applied to both simulated and real process data to demonstrate that the resulting latent variables extract useful information from the response processes.     

Characteristics of incarcerated women's participation in groups: Exploring the Programme for the International Assessment of Adult Competencies United States prison dataset

This study employs logistic regression and latent class analyses to explore the relationship between incarcerated women's (n = 252) participation in nonacademic prison groups/skills classes and their demographic (e.g., education) and prison context (e.g., time to release) characteristics, using the 2014 Programme for the International Assessment of Adult Competencies United States prison dataset. Time to release was a significant factor in readiness class participation. Time to release and race/ethnicity were significant factors in addiction group participation. Education level had the largest effect size across all groups/classes explored. We discuss counseling implications for using group/class screening to connect with incarcerated women, especially those with intersecting and historically marginalized identities.     

The association between personality and cognitive ability: Going beyond simple effects

To examine the relationship between the Big Five and cognitive ability, we investigated whether we could replicate in a heterogeneous population sample the positive association between cognitive ability and Openness and Emotional Stability and its negative association with Conscientiousness. Besides analyzing the pure associations, we shed further light on sources of these associations by investigating potential moderating effects of education and labor force participation. Our results clearly replicate the previously found positive association between cognitive ability and Emotional Stability and Openness and the negative relationship between Conscientiousness and cognitive ability. The correlation between cognitive ability and Openness was found to be moderated by educational attainment, the negative association between Conscientiousness and cognitive ability was moderated by labor force participation.