Agricultural Small-medium Enterprises
Description
data is used by 371 agricultural small-medium enterprises in China.The data collection was conducted in China from November 19th to December 23rd in 2021, collaborated with the local economic and information commission. In the beginning, we distributed survey questionnaires to each agricultural enterprise and then collected them after they had finished.
Files
Steps to reproduce
we use the direct method of calibration, which chooses the values 0.95, 0.50, 0.05 to transform the data into the [0,1] range (Fiss, 2011). Nevertheless, since the data of FP in this study is skewed to the left, we use 80%,50%, and 20% as the thresholds for is “fully in”, “fully out”, and somewhere in between (“crossover”) (Pappas et al; Pappas and Woodside, 2021). In detail, for the continuous value including FP, absorb employment, drive farmers’ number, increase the amount of farmers' income, number of trademarks, we processed these data by SPSS to compute the values of 95%, 50%, 5% as the three thresholds. Then using fsQCA software, the score of variables’ transformation is automatically computed on three membership which is “fully in”, “fully out”, and somewhere in between (“crossover”) by calibration (Pappas and Woodside, 2021). For the dichotomous values, including brand management department, brand award, quality certification, quality management department, and a full-time person responsible for quality management, we use 1 or 0 to represent whether this enterprise has met this condition or not (shown in Table 1). According to the process for research practice by fsQCA 3.0 software (Pappas and Woodside, 2021), we should confirm necessary conditions first, and then construct the truth table algorithm to identify appropriate cases associated with outcomes based on the configurations of the causal conditions. Last, we examined the paths for the condition configurations to the high FP. To filter out configurations with stronger subset relationships and reduce the number of configurations, we chose the consistency threshold to 0.90 (Park et al., 2020), greater than the 0.8 minimum recommended (Leppänen et al., 2019), and the PRI consistency to 0.71, greater than 0.5 significant inconsistency value (Greckhamer et al., 2018). We set the minimum acceptable frequency of cases with at least 1 case. In general, when a dimension is measured with multiple items, we need to compute one value per dimension (Pappas and Woodside, 2021). However, in this study, the individual effect of each condition needs to be considered (C. DiStefano et al., 2009), because for agricultural small-medium enterprises, each of the conditions could be operable on management strategy.