Indira Sen

Indira Sen, Dennis Assenmacher, Mattia Samory et al.

NLP models are used in a variety of critical social computing tasks, such as detecting sexist, racist, or otherwise hateful content. Therefore, it is imperative that these models are robust to spurious features. Past work has attempted to tackle such spurious features using training data augmentation, including Counterfactually Augmented Data (CADs). CADs introduce minimal changes to existing training data points and flip their labels; training on them may reduce model dependency on spurious features. However, manually generating CADs can be time-consuming and expensive. Hence in this work, we assess if this task can be automated using generative NLP models. We automatically generate CADs using Polyjuice, ChatGPT, and Flan-T5, and evaluate their usefulness in improving model robustness compared to manually-generated CADs. By testing both model performance on multiple out-of-domain test sets and individual data point efficacy, our results show that while manual CADs are still the most effective, CADs generated by ChatGPT come a close second. One key reason for the lower performance of automated methods is that the changes they introduce are often insufficient to flip the original label.

Indira Sen, Mattia Samory, Claudia Wagner et al.

Counterfactually Augmented Data (CAD) aims to improve out-of-domain generalizability, an indicator of model robustness. The improvement is credited with promoting core features of the construct over spurious artifacts that happen to correlate with it. Yet, over-relying on core features may lead to unintended model bias. Especially, construct-driven CAD -- perturbations of core features -- may induce models to ignore the context in which core features are used. Here, we test models for sexism and hate speech detection on challenging data: non-hateful and non-sexist usage of identity and gendered terms. In these hard cases, models trained on CAD, especially construct-driven CAD, show higher false-positive rates than models trained on the original, unperturbed data. Using a diverse set of CAD -- construct-driven and construct-agnostic -- reduces such unintended bias.

Marlene Lutz, Indira Sen, Georg Ahnert et al.

Persona prompting is increasingly used in large language models (LLMs) to simulate views of various sociodemographic groups. However, how a persona prompt is formulated can significantly affect outcomes, raising concerns about the fidelity of such simulations. Using five open-source LLMs, we systematically examine how different persona prompt strategies, specifically role adoption formats and demographic priming strategies, influence LLM simulations across 15 intersectional demographic groups in both open- and closed-ended tasks. Our findings show that LLMs struggle to simulate marginalized groups but that the choice of demographic priming and role adoption strategy significantly impacts their portrayal. Specifically, we find that prompting in an interview-style format and name-based priming can help reduce stereotyping and improve alignment. Surprisingly, smaller models like OLMo-2-7B outperform larger ones such as Llama-3.3-70B. Our findings offer actionable guidance for designing sociodemographic persona prompts in LLM-based simulation studies.

Dimosthenis Antypas, Indira Sen, Carla Perez-Almendros et al.

The detection of sensitive content in large datasets is crucial for ensuring that shared and analysed data is free from harmful material. However, current moderation tools, such as external APIs, suffer from limitations in customisation, accuracy across diverse sensitive categories, and privacy concerns. Additionally, existing datasets and open-source models focus predominantly on toxic language, leaving gaps in detecting other sensitive categories such as substance abuse or self-harm. In this paper, we put forward a unified dataset tailored for social media content moderation across six sensitive categories: conflictual language, profanity, sexually explicit material, drug-related content, self-harm, and spam. By collecting and annotating data with consistent retrieval strategies and guidelines, we address the shortcomings of previous focalised research. Our analysis demonstrates that fine-tuning large language models (LLMs) on this novel dataset yields significant improvements in detection performance compared to open off-the-shelf models such as LLaMA, and even proprietary OpenAI models, which underperform by 10-15% overall. This limitation is even more pronounced on popular moderation APIs, which cannot be easily tailored to specific sensitive content categories, among others.

Mats Faulborn, Indira Sen, Max Pellert et al.

Prompt-based language models like GPT4 and LLaMa have been used for a wide variety of use cases such as simulating agents, searching for information, or for content analysis. For all of these applications and others, political biases in these models can affect their performance. Several researchers have attempted to study political bias in language models using evaluation suites based on surveys, such as the Political Compass Test (PCT), often finding a particular leaning favored by these models. However, there is some variation in the exact prompting techniques, leading to diverging findings, and most research relies on constrained-answer settings to extract model responses. Moreover, the Political Compass Test is not a scientifically valid survey instrument. In this work, we contribute a political bias measured informed by political science theory, building on survey design principles to test a wide variety of input prompts, while taking into account prompt sensitivity. We then prompt 11 different open and commercial models, differentiating between instruction-tuned and non-instruction-tuned models, and automatically classify their political stances from 88,110 responses. Leveraging this dataset, we compute political bias profiles across different prompt variations and find that while PCT exaggerates bias in certain models like GPT3.5, measures of political bias are often unstable, but generally more left-leaning for instruction-tuned models. Code and data are available on: https://github.com/MaFa211/theory_grounded_pol_bias

Zehui Yu, Indira Sen, Dennis Assenmacher et al.

Machine learning (ML)-based content moderation tools are essential to keep online spaces free from hateful communication. Yet, ML tools can only be as capable as the quality of the data they are trained on allows them. While there is increasing evidence that they underperform in detecting hateful communications directed towards specific identities and may discriminate against them, we know surprisingly little about the provenance of such bias. To fill this gap, we present a systematic review of the datasets for the automated detection of hateful communication introduced over the past decade, and unpack the quality of the datasets in terms of the identities that they embody: those of the targets of hateful communication that the data curators focused on, as well as those unintentionally included in the datasets. We find, overall, a skewed representation of selected target identities and mismatches between the targets that research conceptualizes and ultimately includes in datasets. Yet, by contextualizing these findings in the language and location of origin of the datasets, we highlight a positive trend towards the broadening and diversification of this research space.

Shayan Alipour, Indira Sen, Mattia Samory et al.

Large language models (LLMs) are known to exhibit demographic biases, yet few studies systematically evaluate these biases across multiple datasets or account for confounding factors. In this work, we examine LLM alignment with human annotations in five offensive language datasets, comprising approximately 220K annotations. Our findings reveal that while demographic traits, particularly race, influence alignment, these effects are inconsistent across datasets and often entangled with other factors. Confounders -- such as document difficulty, annotator sensitivity, and within-group agreement -- account for more variation in alignment patterns than demographic traits alone. Specifically, alignment increases with higher annotator sensitivity and group agreement, while greater document difficulty corresponds to reduced alignment. Our results underscore the importance of multi-dataset analyses and confounder-aware methodologies in developing robust measures of demographic bias in LLMs.

Myrthe Reuver, Indira Sen, Matteo Melis et al.

This paper investigates hybrid intelligence and collaboration between researchers of sexism and Large Language Models (LLMs), with a four-component pipeline. First, nine sexism researchers answer questions about their knowledge of sexism and of LLMs. They then participate in two interactive experiments involving an LLM (GPT3.5). The first experiment has experts assessing the model's knowledge about sexism and suitability for use in research. The second experiment tasks them with creating three different definitions of sexism: an expert-written definition, an LLM-written one, and a co-created definition. Lastly, zero-shot classification experiments use the three definitions from each expert in a prompt template for sexism detection, evaluating GPT4o on 2.500 texts sampled from five sexism benchmarks. We then analyze the resulting 67.500 classification decisions. The LLM interactions lead to longer and more complex definitions of sexism. Expert-written definitions on average perform poorly compared to LLM-generated definitions. However, some experts do improve classification performance with their co-created definitions of sexism, also experts who are inexperienced in using LLMs.

Bolei Ma, Yong Cao, Indira Sen et al.

Large Language Models (LLMs) are increasingly used to simulate public opinion and other social phenomena. Most current studies constrain these simulations to multiple-choice or short-answer formats for ease of scoring and comparison, but such closed designs overlook the inherently generative nature of LLMs. In this position paper, we argue that open-endedness, using free-form text that captures topics, viewpoints, and reasoning processes "in" LLMs, is essential for realistic social simulation. Drawing on decades of survey-methodology research and recent advances in NLP, we argue why this open-endedness is valuable in LLM social simulations, showing how it can improve measurement and design, support exploration of unanticipated views, and reduce researcher-imposed directive bias. It also captures expressiveness and individuality, aids in pretesting, and ultimately enhances methodological utility. We call for novel practices and evaluation frameworks that leverage rather than constrain the open-ended generative diversity of LLMs, creating synergies between NLP and social science.

Jana Jung, Marlene Lutz, Indira Sen et al.

Psychometric tests are increasingly used to assess psychological constructs in large language models (LLMs). However, it remains unclear whether these tests -- originally developed for humans -- yield meaningful results when applied to LLMs. In this study, we systematically evaluate the reliability and validity of human psychometric tests for three constructs: sexism, racism, and morality. We find moderate reliability across multiple item and prompt variations. Validity is evaluated through both convergent (i.e., testing theory-based inter-test correlations) and ecological approaches (i.e., testing the alignment between tests scores and behavior in real-world downstream tasks). Crucially, we find that psychometric test scores do not align, and in some cases even negatively correlate with, model behavior in downstream tasks, indicating low ecological validity. Our results highlight that systematic evaluations of psychometric tests is essential before interpreting their scores. They also suggest that psychometric tests designed for humans cannot be applied directly to LLMs without adaptation.

Max Pellert, Clemens M. Lechner, Indira Sen et al.

This study introduces "Survey and Questionnaire Item Embeddings Differentials" (SQuID), a novel methodological approach that enables neural network embeddings to effectively recover latent dimensions from psychometric survey items. We demonstrate that embeddings derived from large language models, when processed with SQuID, can recover the structure of human values obtained from human rater judgments on the Revised Portrait Value Questionnaire (PVQ-RR). Our experimental validation compares multiple embedding models across a number of evaluation metrics. Unlike previous approaches, SQuID successfully addresses the challenge of obtaining negative correlations between dimensions without requiring domain-specific fine-tuning. Quantitative analysis reveals that our embedding-based approach explains 55% of variance in dimension-dimension similarities compared to human data. Multidimensional scaling configurations from both types of data show fair factor congruence coefficients and largely follow the underlying theory. These results demonstrate that semantic embeddings can effectively replicate psychometric structures previously established through extensive human surveys. The approach offers substantial advantages in cost, scalability and flexibility while maintaining comparable quality to traditional methods. Our findings have significant implications for psychometrics and social science research, providing a complementary methodology that could expand the scope of human behavior and experience represented in measurement tools.

Lukas Birkenmaier, Matthias Roth, Indira Sen

Computational text classification is a challenging task, especially for multi-dimensional social constructs. Recently, there has been increasing discussion that synthetic training data could enhance classification by offering examples of how these constructs are represented in texts. In this paper, we systematically examine the potential of theory-driven synthetic training data for improving the measurement of social constructs. In particular, we explore how researchers can transfer established knowledge from measurement instruments in the social sciences, such as survey scales or annotation codebooks, into theory-driven generation of synthetic data. Using two studies on measuring sexism and political topics, we assess the added value of synthetic training data for fine-tuning text classification models. Although the results of the sexism study were less promising, our findings demonstrate that synthetic data can be highly effective in reducing the need for labeled data in political topic classification. With only a minimal drop in performance, synthetic data allows for substituting large amounts of labeled data. Furthermore, theory-driven synthetic data performed markedly better than data generated without conceptual information in mind.

Mykola Trokhymovych, Indira Sen, Martin Gerlach

With over 60M articles, Wikipedia has become the largest platform for open and freely accessible knowledge. While it has more than 15B monthly visits, its content is believed to be inaccessible to many readers due to the lack of readability of its text. However, previous investigations of the readability of Wikipedia have been restricted to English only, and there are currently no systems supporting the automatic readability assessment of the 300+ languages in Wikipedia. To bridge this gap, we develop a multilingual model to score the readability of Wikipedia articles. To train and evaluate this model, we create a novel multilingual dataset spanning 14 languages, by matching articles from Wikipedia to simplified Wikipedia and online children encyclopedias. We show that our model performs well in a zero-shot scenario, yielding a ranking accuracy of more than 80% across 14 languages and improving upon previous benchmarks. These results demonstrate the applicability of the model at scale for languages in which there is no ground-truth data available for model fine-tuning. Furthermore, we provide the first overview on the state of readability in Wikipedia beyond English.

Mattia Samory, Indira Sen, Julian Kohne et al.

Research has focused on automated methods to effectively detect sexism online. Although overt sexism seems easy to spot, its subtle forms and manifold expressions are not. In this paper, we outline the different dimensions of sexism by grounding them in their implementation in psychological scales. From the scales, we derive a codebook for sexism in social media, which we use to annotate existing and novel datasets, surfacing their limitations in breadth and validity with respect to the construct of sexism. Next, we leverage the annotated datasets to generate adversarial examples, and test the reliability of sexism detection methods. Results indicate that current machine learning models pick up on a very narrow set of linguistic markers of sexism and do not generalize well to out-of-domain examples. Yet, including diverse data and adversarial examples at training time results in models that generalize better and that are more robust to artifacts of data collection. By providing a scale-based codebook and insights regarding the shortcomings of the state-of-the-art, we hope to contribute to the development of better and broader models for sexism detection, including reflections on theory-driven approaches to data collection.

Indira Sen, Fabian Floeck, Katrin Weller et al.

Peoples' activities and opinions recorded as digital traces online, especially on social media and other web-based platforms, offer increasingly informative pictures of the public. They promise to allow inferences about populations beyond the users of the platforms on which the traces are recorded, representing real potential for the Social Sciences and a complement to survey-based research. But the use of digital traces brings its own complexities and new error sources to the research enterprise. Recently, researchers have begun to discuss the errors that can occur when digital traces are used to learn about humans and social phenomena. This article synthesizes this discussion and proposes a systematic way to categorize potential errors, inspired by the Total Survey Error (TSE) Framework developed for survey methodology. We introduce a conceptual framework to diagnose, understand, and document errors that may occur in studies based on such digital traces. While there are clear parallels to the well-known error sources in the TSE framework, the new "Total Error Framework for Digital Traces of Human Behavior on Online Platforms" (TED-On) identifies several types of error that are specific to the use of digital traces. By providing a standard vocabulary to describe these errors, the proposed framework is intended to advance communication and research concerning the use of digital traces in scientific social research.