3.1 Data Sources
In 2023, the researchers conducted a field visit to communities in Langao, Hanyin, and Xixiang counties and distributed questionnaires to 300 farmers who were of legal working age (16-60 years old) using probability sampling. After excluding 30 questionnaires with missing data and incorrect responses, a total of 285 valid questionnaires were collected, resulting in a return rate of 95%, which was generally in line with the experimental expectations. Based on the basic information gathered, the age range of the respondents was between 42 and 54 years old, and all of them were involved in crop farming. Among the respondents, 20 belonged to the low-income group in China. Of the total respondents, 119 were male (41.8%) and 136 were female (58.2%). The educational level of the respondents was generally low, with 46 having completed only elementary school, 132 having completed junior high school, 78 having completed middle school, and only 29 having completed high school or above. There were no significant differences observed in demographic variables among the respondents.
3.2 Experimental Design
A questionnaire experiment is a research method that involves using a carefully designed questionnaire to collect and present information to participants, who then provide responses based on the provided options or open-ended questions. Questionnaire experiments are widely used in the social sciences due to their intuitive and straightforward nature. In this study, two separate and sequential questionnaire experiments were designed to test all the hypotheses mentioned earlier. These experiments were conducted to gather empirical data and examine the relationships and effects proposed in the hypotheses. The questionnaire experiments provide a structured and controlled environment for collecting data and analyzing the impact of information framing and regulatory orientation on farmers' willingness to engage in waste classification.
3.2.1 Experiment 1: Measurement of mediation orientation
To facilitate the triage of manipulation conditions, a preliminary study was conducted to assess the mediation orientation of the participants. Given the limited capacity of farmers to comprehend and analyze complex information, we opted to employ the RFQ questionnaire [15] to gauge respondents' mediation orientations. This questionnaire possesses a straightforward structure and an implicit measurement intention, making it the most widely utilized mediation orientation scale globally [16].
The results of the preliminary study were finalized after assessing and categorizing all respondents based on their score ranges, as depicted in Figure 1. The graph reveals that out of the total respondents, 98 exhibited a moderating orientation termed "facilitating orientation" (SCORE≥0), with the majority falling within the range of 0.5 to 1.5. On the other hand, 187 respondents demonstrated a "preventive orientation" (SCORE<0), with the majority scoring between -1.0 and -2.0. These findings suggest that farmers in the region tend to prioritize potential loss scenarios and negative outcomes, indicating a social profile akin to that of a more conservative and risk-averse society. It is worth noting that the CITI values of the indicators exceeded 0.7 in the question groups related to the dimensions of promotion orientation and prevention orientation. Furthermore, the alpha value, after item deletion, was lower than the overall reliability value, indicating the questionnaire's stability in measuring the intended constructs.
The results of the preliminary study are shown in Figure 2.
Figure 2 about here
3.2.2 Experiment 2: News promotion based on different information frames
To mitigate the potential influence of the Hawthorne effect on the experimental outcomes, Experiment 2 employed a between-groups design for a controlled experiment. Following the principle of balanced distribution and considering the findings from Experiment 1, the researcher utilized stratified random sampling to evenly assign participants to either the benefit-based information framing group (Group A) or the profit and loss-based information framing group (Group B). Ultimately, 142 valid questionnaires were collected from Group A, while Group B yielded 143 valid questionnaires. Statistical analysis revealed no significant difference between the two groups.
Respondents in Group A would read a piece of news on the China News Network promoting garbage categorization, which is based on the textual presentation of the Beneficial Information Framework:
" The Earth serves as our shared abode, and attaining sustainable development is of utmost importance. In our community, we are actively promoting resourcefulness in waste management and providing various incentives. By sorting waste appropriately, it can be converted into recyclable materials based on market prices. Once the recyclables you deposit are weighed and valued, you can directly redeem them for points. These points can then be exchanged for rewards, benefiting social workers who accumulate a certain number of points. It is worth noting that rewards are limited, but the more effort you put in, the greater the potential for rewards. Let us surpass our neighbors and make a positive impact together!"
Respondents in Group B will read a news paragraph on China News Network nudging waste classification, which is a textual presentation based on the profit-and-loss type of information framework:
" The Earth is our shared home, and it is crucial that we strive for sustainable development. In our community, we are actively encouraging resourcefulness in waste management and providing a variety of incentives. When waste is sorted correctly, it is converted into recyclable materials based on their market value. However, failing to properly classify recyclable waste results in a significant reduction in dividends and the inability to earn points. Once a certain number of points is accumulated, they can be exchanged for rewards for social workers. It's important to note that rewards are limited, and the amount received is directly proportional to the effort put in. So, let's make sure we don't fall behind our neighbors and maximize our contributions!"
Following the review of the promotional text, participants were administered two question items. The first item pertained to the manipulation of the independent variable through information framing. Participants were asked to determine the primary position advocated in the preceding news article. Responses were measured on a scale ranging from -3 to 3, where -3 indicated the belief that engaging in waste sorting would yield personal benefits, 0 denoted neutrality, and 3 reflected the perception that not participating in waste sorting would result in personal losses. The second question item focused on assessing participants' willingness to participate, serving as the dependent variable. This item followed the established questionnaire format employed in previous studies conducted with Chinese farmers [17, 18]. Participants were asked, "After reading the aforementioned policy, are you interested in engaging in waste separation?" Responses were measured on a five-point Richter scale, with 1 indicating "not interested," 2 representing "not too interested," 3 denoting "uncertain," 4 indicating "interested," and 5 signifying "very interested." To ensure the accuracy and validity of the results, the researchers employed a five-point Richter scale for the dependent variable.