paper-with-me

홈 › Papers

Quantity versus quality in publication activity: knowledge production at the regional level

2023-11-15 · Timur Gareev, Irina Peker

This study contributes to the ongoing debate regarding the balance between quality and quantity in research productivity and publication activity. Using empirical regional knowledge production functions, we establish a significant correlation between R&D spending and research output, specifically publication productivity, while controlling for patenting activity and socioeconomic factors. Our focus is on the dilemma of research quantity versus quality, which is analysed in the context of regional thematic specialization using spatial lags. When designing policies and making forecasts, it is important to consider the quality of research measured by established indicators. In this study, we examine the dual effect of research quality on publication activity. We identify two groups of quality factors: those related to the quality of journals and those related to the impact of publications. On average, these factors have different influences on quantitative measures. The quality of journals shows a negative relationship with quantity, indicating that as journal quality increases, the number of publications decreases. On the other hand, the impact of publications can be approximated by an inverse parabolic shape, with a positive decreasing slope within a common range of values. This duality in the relationship between quality factors and quantitative measures may explain some of the significant variations in conclusions found in the literature. We compare several models that explore factors influencing publication activity using a balanced panel dataset of Russian regions from 2009 to 2021. Additionally, we propose a novel approach using thematic scientometric parameters as a special type of proximity measure between regions in thematic space. Incorporating spatial spillovers in thematic space allows us to account for potential cross-sectional dependence in regional data.

📄 PDF Abstract BibTeX arXiv:2311.08830

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

Focus 설명 없음

Similar Papers 제목 키워드 기반

Decoding Knowledge Claims: The Evaluation of Scientific Publication Contributions through Semantic Analysis

2024-07-26 · Luca D'Aniello, Nicolas Robinson-Garcia, Massimo Aria, Corrado Cuccurullo

The surge in scientific publications challenges the use of publication counts as a measure of scientific progress, requiring alternative metrics that emphasize the quality and novelty of scientific contributions rather t…

text similarity

The Impact of Research Funding on Knowledge Creation and Dissemination: A study of SNSF Research Grants

2020-11-23 · Rachel Heyard, Hanna Hottenrott

This study investigates the impact of competitive project-funding on researchers' publication outputs. Using detailed information on applicants at the Swiss National Science Foundation (SNSF) and their proposals' evaluat…

Articles

Is Training Data Quality or Quantity More Impactful to Small Language Model Performance?

2024-11-24 · Aryan Sajith, Krishna Chaitanya Rao Kathala

This study investigates the relative impact of training data quality versus quantity on the performance of small language models (SLMs), utilizing the TinyStories dataset for empirical analysis. Analysis of dataset varia…

Language ModelingLanguage ModellingSmall Language Model

A Case Study of User Communication Styles with Customer Service Agents versus Intelligent Virtual Agents

2020-07-01 · SIGDIAL (ACL) 2020 7 · Timothy Hewitt, Ian Beaver

We investigate differences in user communication with live chat agents versus a commercial Intelligent Virtual Agent (IVA). This case study compares the two types of interactions in the same domain for the same company f…

ChatbotDiversity

Within-Model vs Between-Prompt Variability in Large Language Models for Creative Tasks

2026-01-29 · Jennifer Haase, Jana Gonnermann-Müller, Paul H. P. Hanel, Nicolas Leins 외 arxiv

How much of LLM output variance is explained by prompts versus model choice versus stochasticity through sampling? We answer this by evaluating 12 LLMs on 10 creativity prompts with 100 samples each (N = 12,000). For out…