Survey data on inter-firm linkages and innovation activities of Chinese manufacturing SMEs

The data presented in this article is based on a questionnaire survey regarding the inter-firm linkages and innovation activities of Chinese manufacturing SMEs. A valid sample of 420 SMEs was collected, covering six manufacturing industries. The data involves the inter-firm linkages, innovation performance, R&D intensity, IT adoption, and other demographic indicators of the sampling firms. In this data article, we also briefly present the subject area, background of data source, survey process, scale of data, value of data, and main methods used to analyze the data.

relevant items constructed according to prior innovation literature. Statistical methods based upon large scale sample data were adopted to analyze both full-sample data set (targeting SMEs) sub-sample data set (targeting medium-sized firms). A paper that investigated the effects of inter-firm linkages on SME's innovation performance was published in Technological Forecasting and Social Change [1].
Descriptive statistics of the date file can be found in Tables 1e4. Table 1 shows that most of the SMEs in the dataset were private firms (79.3%). Table 2 shows that most of the SMEs in the dataset did export on international market (62.9%).

Distribution by the firm export orientation
Specifications Table   Subject area  Business, Management and Accounting  More specific subject area  Management of Technology and Innovation  Type of data  Table  How  Value of the Data -The survey could serve as a database contributing knowledge on China's manufacturing SMEs. As SMEs' data is difficult to access in general channels like statistical yearbooks of China, the survey provided a template for Chinese SMEs data collection, contributing to the research on SMEs' orientation for external innovation linkages, internal innovation management, and performance. -The survey and relevant data analysis process could serve as a reference facilitating comparative studies on SMEs' innovations among other geographical countries' contexts. Prior literature has provided enriched evidences on innovation activities from enterprises in Europe via survey, such as the Community Innovation Survey (CIS) of EU. However, data related to Chinese enterprises is scarce, let alone the SMEs. Therefore, the date could be regarded as such references to complement the Chinese evidences for global innovation research. -The two measures on absorptive capacity, i.e. R&D intensity measure and IT adoption measure, are useful for future absorptive capacity research, and the measure of inter-firm linkages can be used in future open innovation research. Table 3 shows that six types of manufacturing industries were covered by the sampling firms, involving Bio Instruments and Equipment Industry (6.0%), Electronics Equipment Industry (11.9%), Hardware & Software Industry (6.4%), Light Industry (36.9%), Machinery Industry (29.3%), and Textiles Industry (9.5%). Table 4 shows the distribution of the firm size, indicating that most of the firms were medium-sized firms (76.0%), while small part of them were small-sized firms (24.0%).

Ethical statements
The research was approved by the Research Ethics Committee of Tsinghua University and was carried out in accordance with the code of ethics governing questionnaire surveys.

Data collection
The data was collected via a questionnaire survey in Zhejiang Province, China.
The date was filtered based on the requirements of the statistical analysis method (i.e. structural equation modelling). We only kept the variables that are relevant to the analysis in the data file and removed other irrelevant variables. The filtering was carried out manually by using SPSS.
Specific variables in the data file were derived in the following ways: Firm ownership was calculated by the item (Private), firm scale was calculated by the item (Person) with the item (lnperson) be transferred by natural logarithm algorithm, firms' export orientation was measured by the item (export).
IP was calculated by set (p1-p4), representing the innovation performance of firms. LPO presented in the full model was calculated by set (a1-a4), representing the linkages with competitors, suppliers, lead users and customers, and complementors; LSI was calculated by set (a5-a8), representing the linkages with universities, research institutes, governance agencies, and finance and law service agencies. AC in one model was calculated by the standardized value of R&D expenditures (represented by ZRD), and in the other mode directly calculated by the item (IT), indicating the firm engaging information technology activities or not.
The item (industry) represented the industry background of samples. Upon the measurements above, general statistical analysis approaches were used via software SPSS, which involve descriptive analysis, factor analysis, reliability and validity tests, and regression analysis. In addition, Structural Equation Model was adopted for data analysis via statistical software AMOs.