Page 104 - Journal of Library Science in China 2020 Vol.46

FU Caiwu & WANG Wende / “Weak participation” in rural cultural benefiting project 103
and its reform strategy: A survey from 282 administrative villages in 21 provinces across the country

Table 5. Evaluation dimension of approval
Dissatisfied Satisfied
Don’t need Opposition Neutrality

Need Expectation Support

By the survey results on the approval of various CBP by rural residents (see Table 6), the
overall approval of rural residents to CBP is characterized by high demand and low satisfaction
rates. The overall demand for cultural services is high (71.12±11.10%). We can conclude that
such products ( public services ) are needed, indicating that rural residents have a high demand
for public culture, and they generally acknowledge the implementation of CBP. However, the
overall satisfaction with cultural services was low (38.44±3.43%). In other words, they were
less satisfied with the quality of the supply of rural CBP, which did not fully satisfy their public
cultural needs.

Table 6. Approval of different cultural projects ( % )
RRTC RFP RL DRA FSFP NCIRS
Opposition 4.4 11.6 13.0 8.3 8.8 16.5
Neutrality 10.7 17.8 24.5 17.1 11.8 28.9
Expectation 50.2 46.9 37.6 47.5 45.3 34.4
Support 34.7 23.7 25.0 27.1 34.1 20.2
Demand rate 84.9 70.6 62.6 74.6 79.4 54.6
Satisfaction rate 40.9 33.6 39.9 36.3 42.9 37.0

3.2 Analysis of influencing factors

Since the independent variables in the discussion include numerical, ordinal, and nominal
variables, it is impossible to analyze them by numerical regression models such as least squares
regression. So it is necessary to use multi-categorical variables regression models (e.g., Logit,
Probit) or categorical regression by optimal scale regression models in this paper. However, Logit
and Probit models may violate the parallelism test and Hessian matrix singularity problem in the
regression analysis process. In contrast, the optimal scale regression model can adequately deal
with the independent variables’ situation, including nominal variables, ordered variables, and
numerical and interval variables. Therefore, this paper adopts the optimal scale regression model to
solve the problems in the regression of multi-categorical variables.
Optimal scale regression is a categorical variable-oriented regression method proposed by the
DTSS research group at Leiden University in the Netherlands (Meulman, 1998). The principle
of least squares regression (OLS) is to minimize the objective function by regression weight b.