Statistics

by Sohail Khan. Language models are becoming integral to daily life as trusted sources of advice. While their utility has expanded from simple tasks like text summarization to more complex interactions, the empathetic quality of their responses is crucial. This article explores methods to assess the emotional appropriateness of these models, using metrics such as BLEU, ROUGE, and Sentence Transformers. By analyzing models like LLaMA in mental health dialogues, we learn that while they suffer through traditional word-based metrics, LLaMA's performance in capturing empathy through semantic similarity is promising. In addition, we must advocate for continuous monitoring to ensure these models support their users' mental well-being effectively.

by Taesoo Song. Many American cities continue to face severe rental burdens. However, we rarely examine rental affordability through the lens of quantitative data. In this blog post, I demonstrate how to download and visualize rental affordability data for the San Francisco Bay Area using R packages like `tidycensus` and `sf`. This exercise shows that mapping census data can be a straightforward and powerful way to understand the spatial patterns of housing dynamics and can offer valuable insights for research, policy, and advocacy.

by Nanqin Ying. This study analyzes over 50 million judicial decisions from China’s Supreme People’s Court to examine disparities in legal representation and their impact on sentencing across provinces. Focusing on 290 000 drug-related cases, it employs large language models to differentiate between private attorneys and public defenders and assess their sentencing outcomes. The methodology combines advanced text processing with statistical analysis, using clustering to categorize cases by province and representation, and regression models to isolate the effect of legal representation from factors like drug quantity and regional policies. Findings reveal significant regional disparities in legal access driven by economic conditions, highlighting the need for reforms in China’s legal aid system to ensure equitable representation for marginalized groups and promote transparent judicial data for systemic improvements.

by Yue Lin. What are the key challenges and opportunities of applying experiments in the International Political Economy (IPE) research? In this blog, I reviewed an enduring methodological battle between statistics and experiments, and pointed out that the difficulties of randomization and locating credible counterfactuals have served as main hurdles for IPE scholars to widely adopt experimental tools. However, I further demonstrated some new progress in applying survey, field, and lab experiments in the recent IPE scholarship. I concluded that it is crucial for future researchers to think innovatively about how to combine different research methods to make causal claims in IPE studies.