Scientific Productivity and IP Data: Uzbekistan, Kyrgyzstan, Tajikistan, 2002–2023

Data desciption.
JAST, IMPORTIP, EXPORTIP, GERD are the Journal Aryicles in Basic sciences and engineering (unit), ImportIP and ExportIP are import and export of Intellectual properties (in currend US$ and GERD is Government Expenditures in Research and Development (current US$).
Abstract.This study investigates the impact of intellectual property imports and research and development (R&D) expenditures on scientific productivity in the Central Asian countries of Uzbekistan, Kyrgyzstan, and Tajikistan. Initially, country-specific time series models were developed to explore individual dynamics; however, diagnostic tests revealed notable methodological limitations. Specifically, the Ramsey RESET test indicated specification issues in the model for Uzbekistan, while the Augmented Dickey-Fuller (ADF) test suggested non-stationarity in the series for Kyrgyzstan and Tajikistan. Additionally, data gaps for Tajikistan were identified for certain years. To address these limitations and leverage the cross-sectional and temporal dimensions of the dataset, a fixed-effects panel regression model was employed using 22 years of data spanning 2002–2023.
Findings.
Empirical results reveal a statistically significant inverted U-shaped relationship between intellectual property imports and scientific productivity, implying a non-linear quadratic effect. Furthermore, R&D spending positively and statistically robustly influences scientific output.

数据集说明。
JAST、IMPORTIP、EXPORTIP、GERD分别代表基础科学与工程领域期刊论文量（单位：篇）、知识产权进口额、知识产权出口额，其中IMPORTIP与EXPORTIP的计价单位为当前美元，GERD为政府研发支出（Government Expenditures in Research and Development），以当前美元计价。
摘要。本研究以中亚国家乌兹别克斯坦、吉尔吉斯斯坦、塔吉克斯坦为研究对象，探讨知识产权进口与研发（R&D）支出对科学产出的影响。研究初期构建了国别专属时间序列模型以探究各国的动态特征，但诊断检验结果显示该方法存在显著局限：拉姆齐RESET检验表明乌兹别克斯坦的模型存在设定偏误，增广迪基-富勒（Augmented Dickey-Fuller, ADF）检验则显示吉尔吉斯斯坦与塔吉克斯坦的序列存在非平稳性；此外，塔吉克斯坦部分年份存在数据缺失问题。为解决上述局限并充分利用数据集的截面与时间维度特征，本研究采用固定效应面板回归模型，使用2002—2023年共计22年的面板数据开展实证分析。
研究结果。
实证结果表明，知识产权进口额与科学产出之间存在显著的倒U型非线性二次关系，即二者间呈现非线性二次效应；同时，研发支出对科学产出具有显著且稳健的正向影响。