The report to the 20^th National Congress of the Communist Party of China proposed the strategic principle of “leveraging culture to shape tourism and tourism to promote culture, advancing the deep integration of cultural and tourism development” ( Xi, 2022 ). Guided by this policy, the Hangzhou municipal government has capitalized on the poetic cultural heritage of the Qiantang River, Fuchun River, and Xin’an River basins to develop the “Three Rivers and Two Banks” waterway into a premier tourism route, transforming cultural resources into experiential consumption opportunities.

Previous research on poetry-themed cultural tourism has primarily focused on cultural value assessment, resource excavation, and regional development strategies. For instance, Jiang et al. (2022) conducted a quantitative evaluation of the cultural tourism resources along the Eastern Zhejiang Tang Poetry Trail using the Analytic Hierarchy Process (AHP) and Fishbein-Rosenberg Model, subsequently proposing tailored development strategies for four cities based on the findings . Cheng (2023) leveraging the rich poetic resources of five poetry routes in the Yangtze River Delta, outlined three integrated cultural-tourism pathways to revive the cultural legacy of Tang poetry. This approach aims to fulfill tourists’ spiritual needs while advancing principles of sustainable development.

Regarding tourism product development, Xu (2023) examined the status and challenges of tourism development along the Lishui Oujiang Landscape Poetry Route using the Product-Market-Resource (PMR) Theory, proposing targeted strategies to drive innovative product development in the Lishui section. Chen (2022) developed a four-tier experiential design strategy based on scenario theory, leveraging digital cultural-creative products to integrate cultural resources of the Grand Canal Poetry Route.

In research on consumption willingness, Li et al. (2018) employed structural equation modeling (SEM) to reveal that tourists’ ecological values positively influence eco-consumption behaviors through the mediating role of consumption willingness, yet face a “knowledge-behavior gap” requiring policy interventions to strengthen behavioral translation. Separately, Li et al. (2020) developed an SEM framework demonstrating that service delivery experiences in rural lodging constitute the core factor affecting consumer loyalty, with satisfaction mediating the experience-loyalty relationship.

To determine the sample size, a pilot survey was conducted in Xiaoshan District, Hangzhou, yielding an observed variance of 0.6084 against a theoretical variance of 0.5069. The design effect (Deff) was therefore calculated as 1.2.

$\text{Deff}=\frac{Vcomplex}{VSRS}=\frac{0.6084}{0.5069}\approx 1.2$

In the pilot survey, 71 out of 78 contacted potential respondents completed the survey, yielding a preliminary response rate of 91.3%.

$\begin{array}{c}\text{Response}\text{rate}=\left(\frac{\text{Number}\text{of}\text{people}\text{who}\text{completed}\text{the}\text{survey}}{\text{Total}\text{number}\text{of}\text{people}\text{contacted}}\right)\times 100\%\\ =\frac{71}{78}\times 100\%=91.3\%\end{array}$

As the survey concerns satisfaction measurement, this can be treated as a proportion estimation problem. Targeting a 95% confidence level with a 4% margin of error, we employ the most conservative approach by setting $\hat{P}=0.5$.

With z = 1.96 (95% confidence level) and e = 0.04 (4% margin of error), we define N as the permanent resident population of Hangzhou’s nine core districts at the end of 2024: Chun’an County, Jiande City, Tonglu County, Fuyang District, Shangcheng District, Binjiang District, Xihu District, Xiaoshan District, and Qiantang District. The sample size was calculated using the finite population correction formula as follows:

$n=\frac{z^2\times p\times \left(1-p\right)}{e^2+\frac{z^2\times p\times \left(1-p\right)}{N}}\approx 600$

Employing a stratified sampling design (DEFF = 1.2), we calculated the adjusted sample size using the formula:

$n_{adjusted}=600\times 1.2=720$

Accounting for anticipated non-response factors (e.g., invalid questionnaires) with a conservatively projected 90% response rate, the final sample size was adjusted as follows:

$n_{final}=\frac{720}{0.913}\approx 789$

To ensure questionnaire validity, the sample size was strategically increased, resulting in a finalized requirement of 800 complete responses.

To ensure sample representativeness and result accuracy, we employed multiple sampling methods—including stratified sampling, random sampling, and quota sampling—for different research stages, as detailed below:

First, the stratified sampling method was employed, dividing the survey area into strata. The sample size for each stratum was determined according to the population proportion of each region based on data from the Hangzhou Municipal Bureau of Statistics, as detailed in Table 1 .

Within each stratum, two Poetry Road cultural-tourism sites were randomly selected. Following site confirmation, questionnaires were equally allocated across all sites; convenience sampling was then employed to administer surveys to visitors at respective locations. The distribution of questionnaires for attractions in each district and the actual number of questionnaires collected are shown in Table 2 , with a final questionnaire recovery rate of 97.75%.

An In-depth Analysis Using Structural Equation Modeling: Relationships Among Consumer Experience, Infrastructure, Cultural-Tourism Value, Satisfaction, and Consumption Intention.