Of Strange Loops and Real Effects: Five Theses on Autofiction/the Autofictional

This chapter explores the advantages of understanding the autofictional as a flexible matrix with scalable parameters. It puts forward five theses: (1) The fact that so many scholars have tried to work with the term “autofiction” indicates an obvious need for the “autofictional” to grasp what is vibrant between life and text. (2) The autofictional is a scalable and latent dimension in all autobiographical writing. Therefore, autofiction is not a separate genre in addition to autobiography and the novel. (3) Imagination and the use of the supernatural may support autobiographical reference. (4) Autofiction produces real-life effects. (5) Autofiction oscillates between fictionality and factuality. Although it brings one or other aspect to the foreground, all of them persist and continue to, more or less, resonate together.

Pseudotime analysis reveals exponential trends in DNA methylation aging with mortality associated timescales

The epigenetic trajectory of DNA methylation profiles has a nonlinear relationship with time, reflecting rapid changes in DNA methylation early in life that progressively slow. In this study, we use pseudotime analysis to determine these trajectories. Unlike epigenetic clocks that constrain the functional form of methylation changes with time, pseudotime analysis orders samples along a path based on similarities in a latent dimension to provide an unbiased trajectory. We show that pseudotime analysis can be applied to DNA methylation in human blood and brain tissue and find that it is highly correlated with the epigenetic states described by the Epigenetic Pacemaker. Moreover, we show that the pseudotime nonlinear trajectory can be modeled using a sum of two exponentials with coefficients that are close to the timescales of human age-associated mortality. Thus, for the first time, we can identify age-associated molecular changes that appear to track the exponential dynamics of mortality risk.

Learning Aggregation Functions

Learning on sets is increasingly gaining attention in the machine learning community, due to its widespread applicability. Typically, representations over sets are computed by using fixed aggregation functions such as sum or maximum. However, recent results showed that universal function representation by sum- (or max-) decomposition requires either highly discontinuous (and thus poorly learnable) mappings, or a latent dimension equal to the maximum number of elements in the set. To mitigate this problem, we introduce LAF (Learning Aggregation Function), a learnable aggregator for sets of arbitrary cardinality. LAF can approximate several extensively used aggregators (such as average, sum, maximum) as well as more complex functions (e.g. variance and skewness). We report experiments on semi-synthetic and real data showing that LAF outperforms state-of-the-art sum- (max-) decomposition architectures such as DeepSets and library-based architectures like Principal Neighborhood Aggregation, and can be effectively combined with attention-based architectures.

Effects of eight neuropsychiatric copy number variants on human brain structure

Many copy number variants (CNVs) confer risk for the same range of neurodevelopmental symptoms and psychiatric conditions including autism and schizophrenia. Yet, to date neuroimaging studies have typically been carried out one mutation at a time, showing that CNVs have large effects on brain anatomy. Here, we aimed to characterize and quantify the distinct brain morphometry effects and latent dimensions across 8 neuropsychiatric CNVs. We analyzed T1-weighted MRI data from clinically and non-clinically ascertained CNV carriers (deletion/duplication) at the 1q21.1 (n = 39/28), 16p11.2 (n = 87/78), 22q11.2 (n = 75/30), and 15q11.2 (n = 72/76) loci as well as 1296 non-carriers (controls). Case-control contrasts of all examined genomic loci demonstrated effects on brain anatomy, with deletions and duplications showing mirror effects at the global and regional levels. Although CNVs mainly showed distinct brain patterns, principal component analysis (PCA) loaded subsets of CNVs on two latent brain dimensions, which explained 32 and 29% of the variance of the 8 Cohen’s d maps. The cingulate gyrus, insula, supplementary motor cortex, and cerebellum were identified by PCA and multi-view pattern learning as top regions contributing to latent dimension shared across subsets of CNVs. The large proportion of distinct CNV effects on brain morphology may explain the small neuroimaging effect sizes reported in polygenic psychiatric conditions. Nevertheless, latent gene brain morphology dimensions will help subgroup the rapidly expanding landscape of neuropsychiatric variants and dissect the heterogeneity of idiopathic conditions.

Assessing the Influence of Reversed Items and Force-Choice on the Work and Meaning Inventory [Evaluación de la influencia de los ítems invertidos y de elección forzosa en el Inventario de trabajo significativo]

Response biases are issues in inventories in positive organizational psychology. The study aims to control the response bias in the assessment of meaning of work through two methods: reversed key items and forced-choice format. The sample consisted of 351 professionals; women constituted 60.0 % of the sample. The participants answered two versions of the instrument for meaning of work: Likert-type items and forced-choice. For both versions, the unifactorial model was the most appropriate for the data available. The results indicate that the random intercepts model fit the Likert data (CFI = .92), as well as the forced-choice model (CFI = .97). Besides, the latent dimension of the forced-choice version did not correlate with acquiescence index (r < .08; p > .05), and approximately 20 % of the variance of the items might be due to the method (Likert or forced-choice). The present study illustrates the importance of response bias control in self-report instruments. ResumenLos sesgos de respuesta son problemas en los inventarios de la psicología organizacional positiva. El estudio tiene como objetivo controlar el sesgo de respuesta en la eva­luación del trabajo significativo a través de dos métodos: ítems clave invertidos y formato de elección forzosa. La muestra estuvo formada por 351 profesionales; las muje­res constituyeron el 60.0 % de la muestra. Los participan­tes respondieron dos versiones del instrumento de signifi­cado del trabajo: ítems tipo Likert y elección forzosa. Para ambas versiones, el modelo unifactorial fue el más apro­piado para los datos disponibles. Los resultados indican que el modelo de intersecciones aleatorias se ajusta a los datos Likert (CFI = .92), así como al modelo de elección forzada (CFI = .97). Además, la dimensión latente de la versión de elección forzada no se correlacionó con el ín­dice de aquiescencia (r < .08; p > .05), y aproximada­mente el 20 % de la varianza de los ítems podría deberse al método (Likert o forzado). elección). El presente estu­dio ilustra la importancia del control del sesgo de res­puesta en los instrumentos de autoinforme.

Using Committee Hearings to Place Lobbyists, Experts and Other Outside Agents into Legislative Preference Space

Theoretical expectations regarding the legislative influence of outside agents, such as lobbyists, agency officials or policy experts, often depend on the relationship between legislators' and agents' preferences. Non-elected agents, however, typically will have preferences defined on a dimension that is different from that of elected legislators. I develop a bridging method that accommodates when the agent preference space shifts and rotates relative to the legislator roll-call preference space, and that identifies distances across the two dimensions in meaningful units necessary to test institutional hypotheses. In my application to Medicare hearings, I show that the agent preference space has an orthogonal rotation anchored by a quality-cost latent dimension, and that legislators heavily condition their questioning of agents on preference distance in a way consistent with informational models of lobbying.

Hyperbolic matrix factorization reaffirms the negative curvature of the native biological space

Past research in systems biology has taken for granted the Euclidean geometry of biological space. This has not only drawn parallels to other fields but has also been convenient due to the ample statistical and numerical optimization tools available to address the core task and downstream machine learning problems. However, emerging theoretical studies now demonstrate that biological databases exhibit hierarchical topology, characterized by heterogeneous degree distribution and a high degree of clustering, thus contradicting the flat geometry assumption. Namely, since the number of nodes in hierarchical structures grows exponentially with node depth, the biological networks naturally reside in a hyperbolic space where the circle circumference and disk area are the exponential functions of the radius. To test these claims and assess potential benefits of the applications grounded in the above hypothesis, we have developed a mathematical framework and an accompanying computational procedure for matrix factorization and implied biological relationship inference in hyperbolic space. Not only does our study demonstrate a significant increase in the accuracy of hyperbolic embedding compared to Euclidean embedding, but it also shows that the latent dimension of an optimal hyperbolic embedding is by more than an order of magnitude smaller than the latent dimension of an optimal Euclidean embedding. We see this as additional evidence that hyperbolic geometry, rather than Euclidean, underlines the biological system.

Revisiting Strong Assumptions in Psychological Constructs Measurement: Further Studying Deviations from Unidimensionality and Bias in Latent Variable Measurement Models

In the realities of measurement in social and behavioral sciences, in addition to the characteristic(s) of the respondents targeted by the measurement, other influences (other characteristics of the respondents and the items) can be reflected by the responses to the items in a measure. In the current study, different levels of deviations from strict unidimensionality in measures and the accuracy of parameter estimates of widely used unidimensional latent variable measurement models were further investigated. Of interest were unidimensionality violations in measures intended/designed as unidimensional (when the items primarily reflect a dominant latent dimension, as intended in a unidimensional measure, but also reflect, to a smaller degree, some additional influences). In the simulated conditions of interest, varying degrees of systematic error (bias) in the unidimensional model item and person parameters estimates were demonstrated (e.g., factor loadings overestimation and measurement error underestimation). The strength of the relevant relations and the size of bias were examined. If the size of these systematic distortions is uncommunicated, various negative consequences can ensue for substantive research and applied measurement (in relation to the reliability, validity, and fairness of research/measurement outcomes), when the model estimates are used. The utility of the approach employed in the study was discussed.

A Latent Measure of Political Protest

In recent years, scholars have developed a number of new databases with which to measure protest. Although these databases have distinct coding rules, all attempt to capture incidents of social conflict. We argue, however, that due to a variety of sources of measurement error, subjective coding decisions, and operational specifications, no single indicator of protest adequately measures how much protest exists in a given place at a given time. As a result, empirical studies that employ these measures yield inferences with limited generalizability. To increase the generalizability of the empirical findings, we suggest using an Item Response Theory (IRT) approach to estimate a latent dimension of protest using nine different protest measures that vary in their operational specifications as well as their temporal and spatial coverage. The estimates of the IRT models are used in two ways. First, to demonstrate how existing measures differ, the IRT’s item estimates are used to compare the nine measures of protest based on their degree of difficulty (the quantity of latent protest required to observe a ‘1’ in the data) and their ability to discriminate (the speed with which changes in the latent quantity of protest affect the probability of observing a ‘1’ in the data). Second, the estimated quantity of protest is applied to both monthly and yearly models of authoritarian breakdown. The results demonstrate that the latent protest variable increases the out-of-sample classification of authoritarian breakdown events; and improves in-sample prediction relative to existing global protest variables. Our study illustrates the potential value of modeling a latent dimension of protest rather than solely relying on observed indicators.

A Critical Evaluation of the Reflective Functioning Questionnaire

The Reflective Functioning Questionnaire (RFQ-8) is a short self-report measure of reflective functioning (i.e., the ability to understand mental states of the self and others) that is presumed to capture individual differences in hypo- and hypermentalizing. Despite its broad acceptance by the field and its regular use in primary investigations of the construct, we argue that the validity of the measure is still not well established. The current research elaborates on why the proposed scoring procedure may be methodologically problematic, the item content might not sufficiently cover the full breadth of the mentalizing construct, and it is unclear whether the measure captures mentalizing processes in particular or rather general psychological impairment. In a clinical sample (N = 861) and a sample of young adults (N = 566), we explore these critical considerations and demonstrate that the RFQ-8 may assess a single latent dimension related to hypomentalizing, but provides little unique variance above and beyond broad dimensions of personality pathology and is unlikely to capture maladaptive forms of hypermentalizing. The findings cast doubt on the validity of the RFQ-8 as a measure of reflective functioning. Future research should increase validation efforts concerning the RFQ-8 or develop new measures of reflective functioning.