Research on the Analysis of Learning Influencing Factors Based on Multiple Regression in Online Ideological and Political Education in Colleges and Universities

The purpose of network ideological and political education in colleges and universities is to guide students to use the network correctly, enhance students’ awareness of national security and network security, cultivate students’ feelings of patriotism, and improve students’ sense of social responsibility and civic awareness. However, the current network ideological and political education in colleges and universities has many influencing factors [1–3].

First, there is a lack of comprehensive understanding of network ideological education. Many ideological workers in colleges and universities think that network ideological education is not practical, and they prefer the mode of face-to-face communication of ideological education in reality. But now the network has become an indispensable part of college students’ life, if we leave the network and then talk about the ideological education of college students is very one-sided [4–7]. Secondly, the effect of network ideological education is not obvious. In the face of college students for the network information misperception or network behavior is out of specification, the ideological workers in colleges and universities are often based on past work experience [8–9], take the traditional criticism and education, persuasion and didactic way to carry out the work, such work has an inevitable lag, most of which are for emergency events to take the emergency action, there is no normalization, standardized working mechanism, its education is often less satisfactory [10–12]. Finally, there is a lack of effective guidance for online public opinion. Many colleges and universities do not pay much attention to the supervision and management of online public opinion, which makes many untrue and negative information spread among college students. However, the ideological workers in colleges and universities often start the remedial mechanism only at the time, missing the best time for effective ideological education [13–16].

Literature [17] studied the evaluation of the effectiveness of online Civic Education methods in colleges and universities. The evaluation model was implemented by using PCA algorithm and MLP neural network, and the data were analyzed and dimensionality reduced by PCA algorithm, and the MLP neural network was trained, and the results pointed out that the evaluation model showed good accuracy and recognition ability. Literature [18] indicated that not only should we pay attention to the benefits of the network for Civic and Political Education, but also recognize the opportunities and challenges it brings to traditional Civic and Political Education as well as the effective measures to strengthen the network Civic and Political Education, and the promotion of the network to become an important path for Civic and Political Education services is an important issue that colleges and universities need to deal with at present. Literature [19] examines the current dilemma of network civic education in colleges and universities and the characteristics of AI education, and discusses the advantages and implementation strategies of AI to improve the effect of network civic education in colleges and universities. Literature [20] specifies that the direction of teaching students in the big data environment will tend to improve the quality of thinking and learning ability. In addition to relying on the network Civics platform and utilizing the Markov model to eliminate the differences between students, teachers can also make use of the Markov chain in the evaluation of teaching effect in order to improve students’ thinking ability. Literature [21] reveals that the development of the network brings opportunities and challenges for Civic and Political Education. An online education management system based on cloud computing technology is proposed to improve the traditional information management process and classification algorithm so as to realize the teaching task of online education.

Literature [22] built a model of factors affecting college students’ acceptance of online Civic and Political Education based on acceptance theory and UTAUT, and the results of the study showed that gender, grade, and job expectations had little effect on acceptance, while social influence, content quality and other factors had a positive effect on acceptance. Literature [23] constructed a network Civic and Political Education teaching innovation platform, utilizing the B/S model to divide it into user layer, network layer and application layer. It is proved through experiments that the platform system has a relatively high F value, which is conducive to improving the effect of network teaching. Literature [24] developed a higher education network teaching platform and the role of this platform in network Civic Education in colleges and universities based on the research on the recommendation method of network teaching courses in colleges and universities. Literature [25] integrated analytical deep learning and network technology to design an online education system for Civics and Politics, aiming to improve the development of online education for Civics and Politics in colleges and universities. Based on the teaching characteristics and development trajectory of Civics education, a teaching-examination system was created to realize online information interaction and improve the quality of online Civics teaching and the coordination of the teaching process. Literature [26] used the qualitative analysis method of rooted theory and NVivo11 analysis software to code, organize and summarize the information related to online Civic Education in universities. The results illustrate that research in this area deepens students’ understanding of cybercivics and provides a reference for colleges and universities to carry out cybercivics.

In this study, the general form of regression analysis and the related contents of multiple regression linear model are firstly elaborated, and then on the basis of methodological research, a self-administered questionnaire was compiled to investigate the views of college students studying in a university on the effect of cyber Civics education, and descriptive statistics were used to conduct a descriptive analysis of the effect of the students’ learning about cyber Civics. Then regression analysis was conducted on the relationship between the characteristics of cyber Civics learners, cyber learning behavior, cyber learning environment and cyber Civics learning effect of the first level dimension and its second level dimension factors respectively. According to the analysis results, the structural equation modeling test using AMOS was continued. In response to the results of the analysis of the influencing factors, this paper proposes strategies to improve the learning effect of online Civics education from the perspective of teachers.

2

Method

2.1

General form of regression analysis

The relationship between variables is either a functional relationship or a correlation, where a functional relationship indicates a definite one-to-one relationship between variables and a correlation indicates that there is a relationship between variables but it cannot be precisely expressed by a functional relationship, for example, the relationship between a father’s height and his children’s height. Correlation analysis mainly describes the closeness of the correlation between variables, while regression analysis not only reveals the degree of interaction between variables but also allows estimation and prediction through the establishment of regression equations.

Regression is a statistical method proposed by Frances Galton. Regression analysis is a computational method and theory used to study the association between one or more independent or predictor variables and a (continuous-valued) dependent or response variable [27]. That is, when there is a correlation between a variable X = {x₁, x₂,…, x_n} and a random variable y, we can characterize this relationship using a mathematical formula as: 1 $y = f (X) + ε$

Where, X is the independent variable, y is the dependent variable, ϵ is the random error, and f (·) is the deterministic functional relationship between the independent variable and the dependent variable. When the value of the random variable X is determined, we can obtain the probability distribution of the dependent variable y through the correlation model. The formula accurately describes both the connection and uncertainty in the correlation between variables. ε as a random error describes the relationship between the independent variable and the dependent variable as a stochastic equation, thus allowing us to study the relationship between the independent variable X and the dependent variable y with the help of stochastic mathematics.

2.2

Multiple linear regression models

In quantitative analysis, a regression model in which there are more than two independent variables is called a multiple regression model, and the regression analysis performed on this model is called multiple regression analysis. Generally speaking, the dependent and independent variables obey the following relationship: 2 $y = b_{0} + b_{1} x_{1} + b_{2} x_{2} + ... + b_{n} x_{n} + u$

Then, after k observations of dependent variable y and independent variable x₁, x₂…x_n, k sets of observation samples (y_t, x_1t, x_2t…x_nt) are obtained, of which t = 1, 2, 3…k. These k sets of observation samples will satisfy the following relationship: 3 $y_{t} = b_{0} + b_{1} x_{1 t} + b_{2} x_{2 t} + .... + b_{n} x_{n t} + u_{t}$

This relationship is the general form of a multiple linear regression model, where (y_t, x_1t, x_2t…x_nt)t = 1, 2, 3…k is the observed sample, b_j(j = 0, 1, 2…n) is the regression parameter, and u_i is the random error [28–29].

The regression coefficient b_j(j = 0, 1, 2…n) in the model represents the effect of the j th independent variable on the dependent variable when the other independent variables are constant. This regression coefficient is called partial regression coefficient.

The expectation function of y is: 4 $E (y_{t}) = b_{0} + b_{1} x_{1 t} + b_{2} x_{2 t} + ...... + b_{n} x_{n t}$

Assuming that the value of the unknown parameter can be estimated by appropriate methods, replacing it with the unknown parameter of the overall regression function, the multiple linear regression equation for the sample can be obtained: 5 ${\hat{y}}_{t} = {\hat{b}}_{0} + {\hat{b}}_{1} x_{1 t} + {\hat{b}}_{2} x_{2 t} + .... + {\hat{b}}_{n} x_{n t}$

It is an estimate of the overall regression equation, where ${\hat{b}}_{j} (j = 0, 1, 2... n)$ is an estimate of b_j(j = 0, 1, 2…n).

The sample regression equation yields an estimate of the dependent variable ${\hat{y}}_{i}$ with a residual e_t usually with some error between it and the normal measurement y_i.

The multiple linear regression equation can be expressed using a matrix as: 6 $(\begin{matrix} y_{1} \\ y_{2} \\ ... \\ y_{n} \end{matrix}) = (\begin{matrix} 1 & x_{11} & x_{21} & ... & x_{n 1} \\ 1 & x_{12} & x_{22} & ... & x_{n 2} \\ ... & ... & ... & ... & ... \\ 1 & x_{1 k} & x_{2 k} & ... & x_{n k} \end{matrix}) (\begin{matrix} b_{0} \\ b_{1} \\ ... \\ b_{k} \end{matrix}) + (\begin{matrix} u_{1} \\ u_{2} \\ ... \\ u_{k} \end{matrix})$ where $Y = (\begin{matrix} y_{1} \\ y_{2} \\ ... \\ y_{n} \end{matrix})$ is the vector of observations of the dependent variable, $X = (\begin{matrix} 1 & x_{11} & x_{21} & ... & x_{n 1} \\ 1 & x_{12} & x_{22} & ... & x_{n 2} \\ 1 & ... & ... & ... & ... \\ 1 & x_{1 k} & x_{2 k} & ... & x_{n k} \end{matrix})$ is the matrix of observations of the independent variable, $B = (\begin{matrix} b_{0} \\ b_{1} \\ ... \\ b_{k} \end{matrix})$ is the vector of regression parameters, and $U = (\begin{matrix} u_{1} \\ u_{2} \\ ... \\ u_{k} \end{matrix})$ is the vector of random errors.

The model can also be expressed as a matrix: 7 $Y = X B + U$

Similarly, the dependent variable expectation matrix can be expressed as: 8 $E (Y) = X B$

The matrix form of the regression equation for a multivariate linear sample can be expressed as: 9 $\hat{Y} = X \hat{B}$

The method of least squares was invented in the 18th century. Relevant scholars in the processing of data found that m + 1 measured such as x₀, x₁, x₂…‥x_m and some unknown parameters such as b₀, b₁, b₂…‥b_n there is a certain linear relationship, that is, x₀ + x₁b₁ + x₂b₂ + …‥ + x_mb_m = 0, due to the existence of measurement error, this formula is not strictly established, in fact, the expression on the left side is not 0, and its true value is closely related to the measurement. If the same thing for n measurements, in order to have more information to accurately estimate the parameter, n needs to be much larger than m, the existence of x_0i + x_1ib₁ + … + x_mib_m = 0(i = 1, 2, …., n), if this formula is strictly established, you can solve for the value of the parameter, but this formula is not strictly established, to estimate the parameter we need to find a suitable algorithm. That is, the following equation: 10 $\sum_{i = 1}^{n} {(x_{0 i} + x_{1 i} b_{1} + ... + x_{m i} b_{m})}^{2} = \min$

The method of least squares is inextricably linked to multiple regression, which can estimate the parameters of the system of equations easily and quickly, and has great advantages in the field of error analysis. The specific procedure of estimating the parameters of multiple linear regression using the least squares method is described as follows:

Let (y_i, x_1t, x_2t…x_nt) be the tnd observation sample, in order to minimize the sum of squared errors, according to the principle of extreme value there are the following conditions: 11 $\frac{\partial \sum (e_{i}^{2})}{\partial b_{j}} = 0 (j = 0, 1, 2..., n)$ assume (office) 12 ${\begin{matrix} \sum 2 e_{t} (- 1) = - 2 \sum [y_{t} - ({\hat{b}}_{0} + {\hat{b}}_{1} x_{1 t} + {\hat{b}}_{2} x_{2 t} + ... + {\hat{b}}_{n} x_{n t})] = 0 \\ \sum 2 e_{t} (x_{1 t}) = - 2 \sum x_{1 t} [y_{t} - ({\hat{b}}_{0} + {\hat{b}}_{1} x_{1 t} + {\hat{b}}_{2} x_{2 t} + ... + {\hat{b}}_{n} x_{n t})] = 0 \\ \sum 2 e_{t} (x_{2 t}) = - 2 \sum x_{2 t} [y_{t} - ({\hat{b}}_{0} + {\hat{b}}_{1} x_{1 t} + {\hat{b}}_{2} x_{2 t} + ... + {\hat{b}}_{n} x_{n t})] = 0 \\ \dots \dots \\ \sum 2 e_{t} (x_{n t}) = - 2 \sum x_{n t} [y_{t} - ({\hat{b}}_{0} + {\hat{b}}_{1} x_{1 t} + {\hat{b}}_{2} x_{2 t} + ... + {\hat{b}}_{n} x_{n t})] = 0 \end{matrix}$

The above n +1 equation is a regular equation, then the matrix is represented as: 13 $(\begin{matrix} \sum e_{i} \\ \sum x_{l t} e_{t} \\ .... \\ \sum x_{n t} e_{t} \end{matrix}) = (\begin{matrix} 1 & 1 & ... & 1 \\ x_{l 1} & x_{l 2} & ... & x_{\ln} \\ ... & ... & ... & ... \\ x_{k l} & x_{k 2} & ... & x_{k n} \end{matrix}) (\begin{matrix} e_{l} \\ e_{2} \\ ... \\ e_{n} \end{matrix}) = X^{'} e = 0$

Both sides of the regression model $Y = X \hat{B} + e$ for the sample are simultaneously multiplied by the matrix of sample observations X′ to obtain: 14 $X^{'} Y = X^{'} X \hat{B} + X^{'} e$

Bringing in the extreme value condition, the formal system of equations is as follows: 15 $X^{'} Y = X^{'} X \hat{B}$

By the classical assumption condition (XX)^−l exists and multiplying both sides by (XX)^−l simultaneously gives an estimate of the parameter: 16 $\hat{B} = (X X)^{- l} X Y$

3

Results and discussion

3.1

Analysis of Civic Education on the Internet

3.1.1

Questionnaire design

This survey relies on the “questionnaire star” questionnaire platform, some students of a university survey, a total of 550 questionnaires were returned, all of them are valid questionnaires. Due to the proposed factor analysis and influence relationship research, it was initially decided to design a scale-based questionnaire, and the project team members designed the questionnaire and conducted a probing survey on the basis of referring to the scales in the literature of the predecessors and consulting the experts’ opinions, and then modified the questionnaire according to the results of the probing survey and formed a formal questionnaire. The questionnaire was designed taking into account the basic situation of students, network equipment and conditions, teaching methods, students’ learning attitudes, teachers’ teaching attitudes and evaluation of teaching effects. The questionnaire consisted of 19 questions, except for the first question and the second question on basic information, the remaining 17 questions were on a five-point Lister scale.

3.1.2

Basic information on respondents

The demographic variables of the respondents are shown in Table 1. The table reflects the distribution of the respondents in this survey, where the mean value represents the concentration trend and the standard deviation represents the fluctuation. In general, the respondents possess diversity characteristics in terms of ethnicity and professional construction status, and comprehensiveness characteristics in terms of gender, age, grade, and discipline. According to the frequency analysis of each variable, it is found that the distribution basically meets the requirements of the sample survey.

Table 1.

Demographic variables of interviewees

Variable	Options	Frequency	Percentage(%)	Mean Value	Standard Deviation
School Level	General Undergraduate	550	100%	1.91	0.352
Gender	Man	285	51.8%	1.56	0.495
Gender	Female	265	48.2%	1.56	0.495
Peoples	Han	500	90.9%	1.12	0.366
Peoples	Ethnic Minorities	50	9.1%	1.12	0.366
Age	Age 18 And Below	180	32.7%	2.05	0.38
	19~22	230	41.8%
	Age 23 And Over	140	25.5%
Grade	Freshman Year	110	20%	2.68	1.025
	Sophomore	108	19.6%
	Junior	135	24.5%
	Senior Year	197	35.8%
The Subject Of His Own Profession	Education/Management/Economics/Law	260	47.3%	2.22	0.725
	Science/Engineering/Agriculture/Medicine	130	23.6%
	Literature/History/Philosophy/Politics	135	24.5%
	Art Class	25	4.5%

3.1.3

Descriptive statistical analysis of survey results

The electronic questionnaire collected and organized by “Questionnaire Star” was used to carry out descriptive statistics on the data using the SPSSAU platform, and the descriptive statistics results of the questionnaire on the effect of network Civic Education are shown in Table 2. As can be seen from the table: 1)

From the results of Q19, it can be seen that the mean value of students’ overall satisfaction with the effect of online Civics teaching is 4.45, and the overall satisfaction rate of students with the effect of online Civics teaching reaches 66.66%, but there are still 33.34% of the students who evaluate the effect of online Civics teaching as average and relatively dissatisfied. What exactly are the factors affecting students’ evaluation of the overall effect? It is necessary to continue our analysis.

2)

From the results of Q3-Q7, it can be seen that students’ feelings about the use of the online teaching platform and teachers’ satisfaction with the use of the online teaching platform basically reached more than 71%, but it should be noted that the satisfaction rate of teachers being able to make full use of the teaching tools for teaching is only 66.14%, which indicates that most teachers, although they can skillfully use the online teaching platform, do not make full use of the resources and functions of the Civics teaching platform.

3)

From the results of Q8-Q11, it can be seen that in terms of the attitude of Civics teaching, the overall satisfaction of students with teachers’ punctuality in class, adequate preparation for class, and timely follow-up after class for answering questions and correcting assignments is still very high, which is more than 86%. However, it is worth noting that the recognition of teachers having good classroom interaction with students is only 52.78%.

4)

From the results of Q12-Q15, it can be seen that the students’ evaluation of the change of teachers’ teaching mode in the process of online Civics teaching basically did not exceed 60% in the other three questions, except for the satisfaction of Q15, which reached 75%, which indicates that the students’ feelings about the change of teachers’ online teaching mode are not obvious, and there is no particularly big change compared with traditional classroom teaching.

5)

Q16-Q18 centered on students’ self-evaluation of their independent learning, and the satisfaction rate of these three questions was generally low, 66.53% satisfaction rate of homework completion, 58% satisfaction rate of after-class learning of online resources, indicating that some students in colleges and universities do not have a high degree of self-motivation to learn, and the option with the lowest satisfaction rate was active participation in classroom interactions, with a satisfaction rate of only 48.34%, which together with the The lowest satisfaction option is active participation in classroom interaction, only 48.34% satisfaction rate, which together with the result of 52.78% satisfaction rate of good classroom interaction between teachers and students reflect that in the process of online teaching and learning, the effect of online teaching and learning is greatly reduced due to the lack of face- to-face exchanges and communication between teachers and students.

Table 2.

Descriptive statistical results of the survey questionnaire

Evaluation index	Mean analysis		Proportional distribution(%)
Evaluation index	Mean	Standard deviation	Disrelish	Unsatisfactory	General	Satisfaction	Very satisfied
Q1 gender	1.66	0.45	-	-	-	-	-
Q2 grade	1.67	0.75	-	-	-	-	-
Q3 The online teaching platform is convenient to use	3.83	0.74	1.43	9.72	17.39	35.95	35.51
Q4 The online teaching platform is fully functional	3.34	0.62	2.07	10.06	15.72	36.5	35.65
Q5 The online learning network is fluent	3.47	1.36	5.11	19.87	24.41	31.73	18.88
Q6 Teachers are skilled in using intelligent teaching tools	4.33	0.95	0.9	6.8	14.61	40.83	36.86
Q7 The teacher makes full use of intelligent teaching tools	3.84	1.24	1.3	10.29	22.27	38.42	27.72
Q8 Teachers can attend classes on time	4.2	1.16	1.97	1.06	7.04	23.37	66.56
Q9 The teacher is fully prepared	4.18	1.33	0.64	1.89	12.52	32.98	51.97
Q10 Teachers and students have good classroom interaction	3.38	1.09	2.61	15.99	28.62	30.47	22.31
Q11 The teacher carefully follows the homework review	4.15	1.01	0.39	2.49	11.34	35.02	50.76
Q12 Teachers’ teaching patterns have changed a lot	3.64	0.96	0.77	15.68	28.51	34.32	20.72
Q13 The teacher designed different teaching patterns	3.83	1.08	1.19	15.18	23.21	38.59	21.83
Q14 Teachers have more teaching cases	3.58	0.65	1.2	14.52	28.43	34.38	21.47
Q15 Teachers will provide electronic learning resources	4.19	0.88	0.04	5.6	18.26	44.2	31.9
Q16 In class I actively participate in classroom interaction	3.72	1.95	2.13	22	27.53	25.89	22.45
Q17 After class, I finish my homework on a quality basis	4.02	1.14	0.42	10.53	22.52	34.44	32.09
Q18 After class, I study electronic resources seriously	3.65	1.92	3.06	15.51	23.43	29.36	28.64
Q19 Your evaluation of online teaching effect	4.45	1.19	1.58	10.62	21.14	39.04	27.62

3.2

Analysis of Learning Influencing Factors of Online Civic Education in Colleges and Universities

3.2.1

Regression Analysis of Level 1 Dimension Variables and Learning Effectiveness

In this study, firstly, we take Civics learners’ characteristics, learning behaviors and online learning environment as independent variables, and Civics learning effect as dependent variable, all of which are continuous variables, and use forced entry method to carry out multiple linear regression analysis, and its results are analyzed as follows:

The coefficients of the regression analysis are shown in Table 3, which shows that the significance P of the three independent variables of Civic Learner Characteristics, Civic Learning Behavior and Online Learning Environment is less than 0.05, so all three of them can significantly predict the Civic Learning Effect, and the regression coefficients are positive, which indicates that the three independent variables significantly and positively influence the online Civic Learning Effect of the students, i.e., the higher the scores of Civic Learner Characteristics, Learning Behavior and Online Learning Environment are, the higher the Civic Learning Effect is. The higher the scores of Civic and Political Learning Characteristics, Learning Behavior and Online Learning Environment, the better the Civic and Political Learning Effect.

Table 3.

Coefficient of linear regression analysis

	Nonnormalized coefficient		Standard coefficient			Common linear statistics
	B	Standard error	beta	t	significance	tolerance	VIF
(Constants)	0.582	0.115		5.221	0.000
Learner characteristics	0.361	0.031	0.342	10.423	0.000	0.614	1.642
Learning behavior	0.195	0.029	0.195	6.423	0.000	0.685	1.472
Learning environment	0.342	0.025	0.376	12.551	0.000	0.751	1.351

Based on the above regression results, a mathematical model is established for the learning effect of online Civics and Politics among college students:

Y=0.582+0.361*X1+0.195*X2+0.342*X3

Y denotes the learning effect of college students’ Civics and Politics, X1 denotes the learner characteristics, X2 denotes the learning behavior, X3 denotes the e-learning environment, and 0.582 is a constant.

3.2.2

Regression Analysis of Secondary Dimension Variables and Learning Effectiveness

Multiple linear regression analysis was conducted using the forced entry method with seven secondary dimension variables of attitude and self-efficacy, professional literacy, learning strategies, personal knowledge management, operational behavior, communicative behavior, and the learning environment as the independent variables, and learning effect as the dependent variable.

The coefficients of the regression analysis are shown in Table 4, which shows that the significance P of the six independent variables, namely, attitude and self-efficacy, professionalism, learning strategy, operational behavior, communication behavior, and learning environment, is less than 0.05, which means that they can significantly affect the learning effect. Whereas, the significance P for personal knowledge management is greater than 0.05, indicating that it cannot significantly influence learning outcomes. Further, the regression coefficients of the six independent variables of attitude and self-efficacy, professionalism, learning strategy, operational behavior, communicative behavior, and learning environment are all positive, indicating that they significantly and positively affect the learning effect of college students’ online Civics.

Table 4.

Coefficient of linear regression analysis

	Nonnormalized coefficient		Standard coefficient			Common linear statistics
	B	Standard error	beta	t	significance	tolerance	VIF
(Constants)	0.405	0.113		3.591	0.000
Attitude and self-efficacy	0.133	0.029	0.155	4.642	0.000	0.533	1.885
Subject quality	0.098	0.031	0.115	3.220	0.002	0.483	2.078
Personal knowledge management	-0.022	0.022	-0.025	-0.862	0.389	0.612	1.596
Learning strategy	0.115	0.036	0.146	3.982	0.000	0.475	2.112
Operating behavior	0.197	0.036	0.211	6.612	0.000	0.621	1.626
Communication behavior	0.054	0.019	0.074	2.645	0.007	0.746	1.344
Learning environment	0.321	0.024	0.349	11.782	0.000	0.725	1.385

Based on the results of the above regression analysis, a mathematical model is established for the learning effect of online Civics for college students:

Y=0.405+0.133*X1+0.098*X2+0.115*X3+0.197*X4+0.054*X5+0.321*X6

Y denotes the effect of college students’ online Civics learning, X1 denotes attitude and self-efficacy, X2 denotes disciplinary literacy, X3 denotes learning strategy, X4 denotes operational behavior, X5 denotes communicative behavior, X6 denotes learning environment, and 0.405 is a constant.

3.3

Analysis of structural equations

3.3.1

Research hypotheses

Using AMOS22 software and based on the results of the previous SPSS correlation and regression analysis, this study constructs a structural equation model of the factors influencing the learning effect of college students’ online Civics and tests the theoretical model. Based on the existing research results, the following research hypotheses are proposed:

H1: Attitude and self-efficacy have a significant positive effect on operational behavior.

H2: Attitude and self-efficacy have a significant positive effect on communication behavior.

H3: Attitude and self-efficacy have a significant positive effect on online Civics learning effect.

H4: Disciplinary literacy has a significant positive effect on operational behavior.

H5: Disciplinary literacy has a significant positive effect on communication behavior.

H6: Disciplinary literacy has a significant positive effect on the learning effect of online Civics.

H7: Learning strategy has a significant positive effect on operational behavior.

H8: Learning strategy has a significant positive effect on communication behavior.

H9: Learning strategies have a significant positive effect on the learning effect of online Civics.

H10: The online learning environment has a significant positive effect on operational behavior.

H11: The online learning environment has a significant positive effect on communication behavior.

H12: The online learning environment has a significant positive effect on online Civics learning effect respectively.

H13: Operational behavior has a significant positive effect on the learning effect of online Civics.

H14: Communication behavior has a significant positive effect on online Civics learning effect.

3.3.2

Research hypothesis testing

To test whether the relationship between the variables is significant, the model parameters need to be tested. If there is no outlier (S.E.>0) in the standardized error in the model, the path coefficients between the latent variables are positive, and the significance test of the parameters meets the standard range (C.R.>2, P<0.05), it means that the relationship between the quantities is significant. Table 5 shows the test results of model parameters and the verification of research hypotheses (*** means P<0.001, ** means P<0.01, and * means P<0.05). As can be seen from the table, the H2 path coefficient is negative, P<0.001, indicating that “attitude and self-efficacy” have a significant negative effect on “communication behavior”, so the H2 hypothesis is not valid, and it is deleted when the model is modified. The path coefficients of H3, H4 and H5 were positive, but did not meet the requirement of significance probability P<0.05, indicating that “attitude and self-efficacy” had a positive but not significant impact on “learning effect”, “discipline literacy” on “operation behavior” and “discipline literacy” on “communication behavior”, so the assumptions of H3, H4 and H5 were not valid, and they were deleted when the model was modified. The 10 hypotheses H1, H6, H7, H8, H9, H10, H11, H12, H13, and H14 were all tested by the model parameters, indicating that the hypothesis was valid. From the perspective of ideological and political learner characteristics and learning behavior, attitude and self-efficacy had a significant positive impact on operational behavior (β =0.276, P<0.001), indicating that students’ good learning attitude and self-efficacy are the prerequisites for promoting their online ideological and political learning operations, which is helpful to enable students to devote themselves to online ideological and political learning behaviors with a more proactive attitude and sufficient self-confidence. Learning strategies had a significant positive impact on operational behavior (β =0.234, P<0.001) and communicative behavior (β =0.424, P<0.001), indicating that mastering good learning strategies could effectively guide, regulate and promote students’ learning and communicative behaviors and operational behaviors.

Table 5.

Model parameter test results and research hypothesis validation

Hypothesize	Pathname	Estimate	S.E.	C.R.	P	Validation
H1	Operating behavior←Attitude and self-efficacy	0.276	0.059	4.7000	***	support
H2	Communication behavior←Attitude and self-efficacy	-0.254	0.075	-3.472	***	unheld
H3	Learning effect←Attitude and self-efficacy	0.078	0.046	1.695	0.093	unheld
H4	Operating behavior←Subject quality	0.019	0.064	0.295	0.770	unheld
H5	Communication behavior←Subject quality	0.083	0.083	1.000	0.315	unheld
H6	Learning effect← Subject quality	0.143	0.047	2.895	**	support
H7	Operating behavior←Learning strategy	0.234	0.058	4.261	***	support
H8	Communication behavior←Learning strategy	0.424	0.075	5.766	***	support
H9	Learning effect←Learning strategy	0.122	0.044	2.761	**	support
H10	Operating behavior←Learning environment	0.265	0.049	5.485	***	support
H11	Communication behavior←Learning environment	0.147	0.056	2.631	**	support
H12	Learning effect←Learning environment	0.363	0.045	8.556	***	support
H13	Learning effect←Operating behavior	0.266	0.053	5.244	***	support
H14	Learning effect←Communication behavior	0.055	0.038	2.012	*	support

4

Observations and recommendations

4.1

Enrichment and implementation of measures to build the ideology and politics of the curriculum

Firstly, schools should allow every teacher to participate in training on Civics in the curriculum, so as to provide them with ideas and methods on how to integrate Civics elements into their professional knowledge, and to avoid the situation of “not knowing what to ask”. In addition, since each teacher has different political qualities and teaches different contents, the training received by teachers of different specialties should also be different. Secondly, the effectiveness of “Curriculum Civics” education should be effectively integrated into the teachers’ assessment mechanism [30]. The results of the assessment will affect the appointment, salary, rewards and punishments of teachers. We can add the evaluation of Civics content in the evaluation of online courses, and the students’ answers and teachers’ lesson plans will be used as one of the reference bases for the assessment. Thirdly, teachers are regularly organized to exchange information on the implementation of “Civic and political thinking in the curriculum”, to share ideas and accumulate experience with each other.

4.2

Enhancement of teachers’ political literacy and professionalism

Teachers should take the initiative to learn the content of Civics, develop their own ideological and moral quality, and explore the places where the content of specialized courses can be combined with Civics content and put it into practice, so that they can undertake the task of “educating people” together with Civics teachers. In addition, from the regression results above, it can be seen that whether or not all teachers carry out “curriculum Civics” has an effect on students’ preference for the way of teaching Civics content. If all teachers carry out “curriculum Civics”, students will prefer the way of teaching Civics content, and will have a greater interest in “curriculum Civics”. If all the teachers implement “Curriculum Civics”, students also prefer the way teachers teach Civics content, and understand and pay more attention to “Curriculum Civics”. Therefore, the teachers of professional courses should participate in the “course ideology” in order to cultivate the college students into talents with all-round development of morality, intellectuality, physique, aesthetics and labor.

4.3

Civic education should be “adapted to local conditions”

First, according to the degree of difficulty and credit hours of online Civics courses and the educational level of students, the teaching hours of online Civics content should be reasonably arranged. Some courses have tighter credit hours and more and more complicated class contents, so in order to complete the teaching task, the content of Civics and Politics should be appropriately reduced. Secondly, according to the survey results, teachers can incorporate more current events in teaching undergraduates in order to stimulate students’ interest and improve the effectiveness of parenting.

5

Conclusion

The article explores and analyzes the learning influencing factors of online Civics education in colleges and universities by conducting a questionnaire survey of students in a college and using statistical analysis methods such as descriptive statistical analysis and regression analysis. The results of the study show that in the regression analysis of the first-level dimension variables and learning effect, the significance P of the three independent variables of Civic Learner Characteristics, Civic Learning Behavior and Networked Learning Environment is less than 0.05, which can be obtained that the higher the scores of Civic Learner Characteristics, Learning Behavior and Networked Learning Environment are, the better is the effect of Civic Learning. Academic literacy, learning strategy, learning environment, operational behavior and communication behavior have a direct positive and significant effect on the effect of online Civics learning. Among them, online learning environment has the greatest influence on online Civics learning effect, and communication behavior has the least influence on learning effect. Although attitude and self-efficacy do not have a direct significant effect on the learning effect, they indirectly affect the learning effect by influencing learners’ operational behavior.

Sprache:: Englisch

Zeitrahmen der Veröffentlichung:: 1 Hefte pro Jahr
Fachgebiete der Zeitschrift:: Biologie, Biologie, andere, Mathematik, Angewandte Mathematik, Mathematik, Allgemeines, Physik, Physik, andere

Zeitschrift RSS Feed

Research on the Analysis of Learning Influencing Factors Based on Multiple Regression in Online Ideological and Political Education in Colleges and Universities

Xiaoguang Liu

Online veröffentlicht: 22. Sept. 2025

Eingereicht: 28. Jan. 2025

Akzeptiert: 30. Apr. 2025

DOI: https://doi.org/10.2478/amns-2025-0960

SchlüsselwörterMultiple regression, Online civic education, Learning influences, Structural equations

© 2025 Xiaoguang Liu, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Schlüsselwörter
Multiple regression, Online civic education, Learning influences, Structural equations