The Effect of Distance Teaching on Academic Achievement in the Online Course of Introduction to Computers Science

For distance teaching, if the teacher can obtain the predicted scores that students may obtain in the final exam at this time, the students whose predicted scores do not meet the standard can be found. Teachers can strengthen the teaching of this type of students in the teaching process, which will greatly improve the overall teaching efficiency of teachers. The target of this research is the students of the course of the “Introduction to Computers Science” at Open University of Kaohsiung. This course is a distance teaching based on online teaching, and the questionnaire and final test are also completed on line. A total of 95 students filled out the questionnaire online, accounting for 77.24% of the electives. This study aims to assess the measurement of students’ pre-learning computer experience, software operation ability, learning motivation, and computer attitude, and analyze their relationship with the performance of subsequent distance teaching on the course of the “Introduction to Computers Science”, and find a multiple regression model to predict student performance. It was found that a multiple linear regression model combining 7 independent variables was statistically significantly related to the final test scores.


Introduction
The Open University of Kaohsiung (OUK) is an adult education university that provides lifelong learning. It provides a dreaming opportunity for people who drop out of school early and fail to receive higher education, so that people who want to obtain a university diploma do not need to pass any exams, and there are no admission qualification restrictions. As long as the people of Taiwan who are computer literacy and performance.
This study will use the frequency of computer use, the number of computer course ever studied, the most familiar computer skills and other factors to explore the relationship between the students' pre-learning computer experience and the learning performance of the course of the "Introduction to Computers Science" based on distance teaching.

Learning Motivation
According to human development professionals, there are two types of learning motivation, namely extrinsic motivation and intrinsic motivation. External motivation refers to the use of external things to induce students to engage in learning activities; internal motivation refers to the motivation caused by people's interest in learning itself. When a person likes to engage in a certain learning activity because he feels that this learning activity has "Interest", "Like", "Happy", "Need", "Target", and no other external reasons, this potential internal force is "intrinsic motivation". Lepper (1988) pointed out that external motivation refers to obtaining some kind of reward or avoiding some kind of punishment external to the activity itself, such as reward or teacher's approval; internal motivation is for its own sake, and regards internal motivation as what it is the enjoyment provided, the learning allowed or the sense of accomplishment inspired. Afzal, Ali, Khan, and Hamid (2010) believe that students with intrinsic motivation prefer to use strategies that require more effort and make them process information more extremely, while students with external motivation tend to pay the least effort to obtain the maximum return. Stipek, Feiler, Daniels, and Milburn (1995) believe that learning motivation is a student's motivation for achievement in learning, a psychological need for individual pursuit of success, and one of the main factors affecting academic achievement. The study found that learning motivation has a strong relationship with students' academic performance, and it is concluded that students with intrinsic motivation are academically better than students with extrinsic motivation (Afzal et al., 2010).
The students of this course are all adults. Adult motivation is reflected in setting goals, maintaining attention in certain activities, and exerting effort and perseverance to achieve goals. According to the conclusion drawn by the Institute of Educational Psychology, it is pointed out that in the learning of adults, high positive motivation can play a compensatory role, especially when a person's ability is low or knowledge reserves are insufficient (Lukianova, 2016). And what is the inner learning motivation of adults? According to M. Bryn and S. Mann, the inner motivation of adult learners is caused by the constant desire to learn, which can be characterized as an attitude: "I really want to do" Instead of "because I want to do it, so I do", or "I need" (Lukianova, 2016).
Therefore, before teaching this course of the "Introduction to Computers Science" based on distance teaching, it is necessary for adult students from all directions to thoroughly understand their learning motivations, determine their needs, interests and tendencies, because these usually have a significant impact on learning outcomes.

Learning Attitude and Performance
Students' learning attitudes have a considerable impact on academic performance, and most educational scholars and teachers agree that a positive learning attitude can improve students' academic performance. Kiekkas et al. (2015) assessed the impact of biostatistics course on the statistical attitudes of nursing students, and explored the relationship between these attitudes and their performance in course examinations. The study found that the correlation between the overall scale score and examination performance was positive and significant, but weak; and previous research on medical students has reported that the relationship between the two is moderately related (Beurze, Donders, Zielhuis, de Vegt, & Verbeek, 2013;Zhang et al., 2012). However, Bidegain and Mujika (2020) explore the relationship between scientific attitudes and PISA (The Program for International Student Assessment) scientific performance. From a broader perspective (e.g., across countries and regions), the relationship between all attitude variables and science achievement using aggregated PISA scores was found to be negative. Hignite and Echternacht (1992) believed that computer attitude cannot be used to predict the existence of corresponding levels of computer literacy. But, Klein, Knupfer and Crooks (1993) evaluated the students' attitudes towards computer learning, and checked the students' computer knowledge and computer skills performance after attending the course, and found that students with positive learning attitudes had significantly better performance assessments. Shen, Wu and Lee (2014) proposed a case study to discuss the impact of their proposed system on college students' attitudes towards computer science, and found that students' attitudes towards computer science improved, and previous studies have shown that positive attitudes toward course tend to produce better results on achievement measures. Gopu (2016) focused on to explore the introduction of student's attitude towards computer science. It was found that there was a strong correlation between students' academic performance and their personal perceptions of competence during the course (Gomes, Santos, & Mendes, 2012); and there was a significant positive correlation between students' attitudes and their achievements in programming (Baer, 2013).
Most of the above studies on the relationship between learning attitudes and achievements are aimed at general students, and there is relatively little research on the relationship between computer learning attitudes and academic performance of adults who participate in distance teaching like OUK, so they are also included in this study.

Methodology
The main purpose of this research is to evaluate the measurement of students' computer experience, software operation ability, learning motivation, and computer attitude before taking the course of the "Introduction to Computers Science" based on distance teaching, and analyze its relationship with subsequent course performance. Find a multiple regression model to predict student achievement in computer literacy course. The content of the course is that teachers use general computer content to conduct online teaching (Kuo, 2020). Based on the content of the lecture, 100 questions are given, 1 point for each question, and a final online test is taken, which is regarded as the score obtained by the students after taking the course. This research offers the following contributions to the literature: 1) Understand the relationship between demographic variables and final test scores.
2) Analyze the relationship between students' pre-learning computer experience, clerical software operation ability, learning motivation and computer attitudes and the final test.
3) Identify a set of multivariate variables that have a significant relationship with the final test scores to build a predictive model.
To verify the research questions, the following null hypothesis are assessed: 1) There are no statistically significant differences at the level of (0.05) between final test scores and demographic variables.
2) There are no statistically significant correlations at the level of (0.05) between final test scores and pre-learning computer experience.
3) There are no statistically significant correlations at the level of (0.05) between final test scores and learning motivation.
4) There are no statistically significant correlations at the level of (0.05) between final test scores and computer use experience.

5)
There are no statistically significant correlations at the level of (0.05) between final test scores and computer attitudes.
6) There are no statistically significant correlations at the level of (0.05) between final test scores and clerical software operation ability.

Design of Research Experiment and Evaluation Questionnaires
This research questionnaire mainly includes four groups: demographic variables, computer learning experience, computer use experience, personal feelings and opinions, and includes a computer attitude scale and a clerical software operation ability scale. Demographic variables include five items including gender, age, education, OUK time already enrolled, and whether you have a personal computer. There are two questions for computer learning experience and personal feelings and opinions, and three questions for computer use experience.
In addition to the two scales, there are 12 items in the questionnaire. The content of the questionnaire was modified from Lin (2007). The instrument of clerical software operation ability scale is the author's own proposition based on teaching experience. The above content is slightly modified by expert opinions. The computer attitude scale uses the scale used by Kuo (2020). The contents of the 12-items questionnaire are shown in Table 1. The clerical software operation capability scale is shown in Table   2.
These two scales were measured by a Likert scale from 1 strongly disagree to 5 strongly agree. After analysis, the Cronbach α of the computer attitude scale for this study was.949, the clerical software operation ability scale was .937. The Cronbach α of these two scales is >.9 in this study, which is an excellent reliability level according to the rules of George and Mallery (2003).
Published by SCHOLINK INC.

Participant and Data Analysis
This research is aimed at students who take the course of the "Introduction to Computers Science"  10. I can use the indent function to set the paragraph format.

Gender
Males have higher academic performance than Females, as shown in Table 3. Further analysis of the ANOVA test results in F=.469, P=.495, and the null hypothesis 1 cannot be rejected. It indicates that there is no obvious gender difference in the final grades of the students. Because there is no age limit for the selection of course for the OUK students. The age distribution of students who choose to take this course is very wide as shown in Table 4. It is found from Table 4 that the average scores of the 36-45, 46-55, and >56 age groups are 86.1, 93, and 87.7 points respectively, which are higher than the average score of the whole class, while the youngest 18-25 age group has the lowest average score of 79.1. As a result, it is clear that older students are not inferior to young students in the study of the course toward the "Introduction to Computers Science". This is different from the current belief that young people's acceptance of computer literacy is higher than that of older people.
This also shows that older students are more serious in their studies than younger students. After the ANOVA test results in F=6.661, P=.012, because the P value was <0.05, null hypothesis 1 can be rejected. It shows that there is obvious age difference in the students' final grades.  Since the OUK has no restrictions on students' qualifications for taking this course, the qualifications of students taking this course are also scattered.  which is higher than the class average. It indicates that the students of these two groups are most used to and familiar with the school's on-line teaching model and learning method. It can also indicate that students in these two groups are studying more seriously. The average score of freshmen is 78 which is the lowest and much lower than the average of the whole class. It may be caused by freshmen who are not used to online learning. This part provides teachers with more attention to freshmen to improve their learning effect. After the ANOVA test results in F=1.456, P=.231, and null hypothesis 1 cannot be rejected. It shows that there is no obvious difference in the OUK time already enrolled for students' final grades. It is found from Table 7 that most of the students at OUK have personal computers. This is because OUK is a school mainly based on distance teaching. Students usually go online for distance learning. It can be seen from Table 7 that the performance of students who do not have a dedicated computer in the course of the "Introduction to Computers Science" is not lower than the students who have a dedicated computer. This is because the content of the course and the quizzes are not actually operated by the computer, but mainly the knowledge content of computer literacy. After the ANOVA test results in F=.429, P=.514, and null hypothesis1 cannot be rejected. It shows that there is no obvious difference in dedicated computers for students' final grades. For students who answered that they have taken any computer-related course, further investigate their learning status. The problem of computer-related course that I have taken is multiple choice (a total of 8 options), the results of multiple choices are not helpful for regression analysis. Therefore, we need to establish a single score to represent the measurement of a single student in the computer-related course that have been studied before we can perform subsequent regression analysis. The survey results are shown in Table 9, where the percentage is the ratio of the number of students who selected this option to the number of students who participated in the questionnaire, and the average score is the average score of the students who selected this option. The selected percentage represents the relative importance of this option. The higher the percentage, the more representative this option is, and the average score of students who choose this option is also an important indicator, so we will multiply the average score and the percentage. The result obtained by multiplication is used as the intensity index of the computer-related course that have been taken. This intensity index is called "Pre-learned scores".
The sum of the scores of all computer-related course ever taken by a single student is the student's score on this question. We call this score "COURSE". The student's final grade and "COURSE" were subjected to Person correlation test analysis to obtain a correlation coefficient of 0.113. Regression analysis was performed to obtain R-Square=.013, F-statistics=1.2, P=.276, and null hypothesis 2 could not be rejected.   Table 10. From Table 10, it is found that the main motivation for students to learn computers is "Strengthen the ability to organize or handle life affairs" for 73.68% of the sample and "To browse and use the Internet function" to account for 54.74%. Since this project is a multiple-choice question, most students have many learning motivations, and it is found that the average score of students with multiple learning motivations is higher than the average score of the whole class, but the difference is not significant.
In order to perform regression analysis, according to the method of dealing with any computer-related course that have been taken in, we established a motivation option intensity indicator called "Motivation scores". The sum of all the scores of multiple motivations selected by a single student is the student's score on this question. We call this score "Motivation". The student's final score and "Motivation" were analyzed by Person correlation test to obtain a correlation coefficient of 0.104.
Regression analysis was performed to obtain R-Square=.008, F-statistics=.780, P=.379. Unable to reject null hypothesis 3. It shows that students' learning motivation has no significant relationship with final test scores.
Published by SCHOLINK INC.   The purpose of students using computer-related equipment and the final test results are shown in Table   13. This question is a multiple-choice question. From Table 13, 73.68% of the students answered "Watch the OUK online teaching", followed by 56.84% of the students answered "Searching information, watching news" and "Listening to music or watching videos online", 51.58% of the answers to the "E-Mail", 42.11% of the answers to the "Go to a community website" and 41.05% of the answers to "Use software such as clerical processing, briefing design and production, spreadsheet, and drawings". Only 27.37% of the students answered "play online games", and 22.11% of the students answered "Go to the chat room or discussion forum". Observed from the final test scores, the purpose of the students' use of the computer has no obvious relationship with the final test scores.

Published by SCHOLINK INC.
Refer to the previous way to deal with multi-option problems, and create a computer-used purpose option score called "Purpose scores". The sum of all the scores of multiple items selected by a single student is the student's score on this question. We call this score "Purpose". The student's final grade and "Purpose" were analyzed by Person correlation test, and the correlation coefficient of -0.138 was obtained. Regression analysis was performed to obtain R-Square=.019, F-statistics=1.799, P=.183, and null hypothesis 4 could not be rejected. Explain that the students' learning purpose has no significant relationship with the final test scores.

Personal Feelings and Opinions
For the discussion of students' personal feelings and opinions on learning computers, it mainly investigates two questions. One is what computer skills the students are most familiar with? The other is do students feel that the technological skills they possess are sufficient? We will discuss in order.

Computer Skills Most Familiar to Students
Students think that they are most familiar with the computer skills and final test scores shown in Table   14. This question is a multiple-choice question. From Table 14, 78.49% of the students answered that their most familiar computer skills were "Clerical processing", 44.09% of the students answered "Basic Introduction to Computers", and 32.26% of the students answered "Briefing design and production". In response to "Basic Introduction to Computers" and other higher-skilled student groups, such as "spreadsheets", "drawing software" and "programming", the average final test score is much higher www.scholink.org/ojs/index.php/fce Vol. 1, No. 2, 2020 than the average class score. Among them, 3 students answered "Other", their average score was 58.7 points, which was far lower than the average score of the whole class, 83.9 points. "Other" may be professional information system operation skills at work, such as counter registration and pricing staff in large hospitals, or the cashier in the store, etc., said that students familiar with keyboard and mouse operation skills similar to this type have no advantage over ordinary people in the course of the "Introduction to Computers Science". This phenomenon is the same as Brock's et al. (1999) research results. The use of computers only in the workplace does not affect computer literacy, because individuals may only learn one application, such as an electronic form or a single word processing program, and may even only know how to turn on the computer. "Computer" in simple English, they are typists who moved to detached by keyboards and never learned about the computer as we have defined computer literacy.

Frontiers of Contemporary Education
Refer to the previous way to deal with multi-choice questions, establish a computer skills score that students are most familiar with is called "Skills scores". The sum of all the scores of the most familiar computer skills selected by a single student is the student's score on this question. We call this score "Skills". The student's final grade and "Skills" were subjected to Person correlation test analysis to obtain a correlation coefficient of 0.103. Regression analysis was performed to obtain R-Square=.011, F-statistics=.995, P=.321, and null hypothesis 2 could not be rejected. This shows that there is no significant relationship between the computer skills that students are familiar with before the study and the final test scores. than the average score of the whole class. For the student groups who think that they do not have enough technological skills, we further investigate the course or training that students think they need.
The results are shown in Table 16. From Table 16, students who answered that the higher-level skills were insufficient, their average score is higher than the class score, for example: students who answered "Drawing software" and "Computer animation". In addition, 64.21% of students think they need "Briefing design and production", 54.74% think "Spreadsheet", 48.42% think "Clerical processing", it can be seen that most of OUK students think they need to learn office software. Like Kuo's (2020) research, this part can be used as a reference for decision makers in OUK course.

Clerical Software Operation Ability Scale
The student's final grade and the score of the clerical software operation ability scale were analyzed by Person correlation test to obtain a correlation coefficient of 0.047. Regression analysis was performed to obtain R-Square=.002, F-statistics=.21, P=.648, and null hypothesis 6 could not be rejected. This shows that there is no significant relationship between the measurement of clerical software operation ability and the final test scores of students before learning this course.

Prediction Model for Final Test
From the results of the discussion of the individual independent variable and the dependent variable above, it is found that in terms of demographic variables, "Age" P<0.05 has the greatest impact on the final test scores, followed by "Education" P=0.063, and other variables have no significant impact on the Computers Science course based on distance teaching. In the analysis of descriptive statistics, it is found that only "Computer attitude scale measurement" has significant impact on the final test scores on the Computers Science course based on distance teaching, and the other variables have no significant impact. However, Afzal et al. (2010) found that there is no significant relationship between individual independent variables and dependent variables. But, there is a significant relationship between the model combining six independent variables and dependent variables. Rex and Roth (1998) found that none of the three measures of computer experience can be used as an important predictor of performance. However, after combining the computer literacy scores, there is a significant relationship between the four independent variable models and learning performance. Therefore, this study hope to establish a set of prediction models that has an important relationship with the final test score is based on the "Introduction to Computers Science" course based on distance teaching from all the independent variables discussed in this research. After multiple regression analysis, a predictive model is established. The multiple regression coefficients of the prediction model are shown in Table 17. This model combined seven independent variables, including "Age", "Education", "Computer attitude measurement", "Motivation", "Course", "Clerical software operation ability scale measurement" and "Skills". Table 17 shows the results of regression analyses with the values of R-Square=.238, F-statistics=3.889, and Significance=.001. The results reveal that the model is significant (p<.05) and there is a strong relationship between independent and dependent variables.

Results, Limitations, and Future Research
This study explores the relationship between the final test scores and factors, including "demographic variables", "computer learning experience", "computer use experience", "Personal feeling and opinions", "computer attitude scale measurement" and "clerical software operation ability scale measures" in the "Introduction to Computers Science" course based on distance teaching. In terms of the relationship between individual independent variables and final test scores, there is obvious age difference. The computer attitude scale measurement and academic performance show a statistically significant correlation. This result is similar to Klein et al. (1993), but different from Hignite and Echternacht (1992). In addition, there is no direct statistically significant association between the other independent variables and the final test result.
In the questionnaire questions of this research, there are many questions with multiple options. There is no way to perform regression analysis on the data of such multiple options. This research uses weighting to quantify each option in order to obtain a single quantification value of the students in this multiple options question data. A final test score prediction model with 7 independent variables is established in the "Introduction to Computers Science" course based on distance teaching. This is just like previous studies (Afzal et al., 2010;Rex & Roth, 1998), combining some independent variables that have no significant relationship between individual independent variables and dependent variables to establish a significant relationship between the independent variable model and academic performance relationship.
This study has limitations. Only 95 students participated in the questionnaire. The amount of data is slightly insufficient. If we can collect more classes or more semesters of accumulated data, in short, increase the amount of research data, the results obtained will be more practical. Another limitation is that for the data of multiple-choice questions, it is not possible to directly perform regression analysis.
In this study, the weighted method is used to generate the option intensity scores of multiple-choice questions to perform regression analysis. This part increases the load of data consolidation and analysis, but it is also the study's important contribution.
From the above results, we know that there are complex and important relationships between factors such as pre-learning computer experience and computer motivation and the learning performance in the "Introduction to Computers Science" course based on distance teaching. Nowadays, in the era of big data, artificial intelligence technology is booming, and it has a strong ability to analyze and calculate complex multiple independent variables and dependent variables. If we can collect enough data, we can consider using machine learning, such as support vector machines, artificial neural networks or deep learning, are used to build predictive models.

Conclusion
Due to the diverse sources of OUK students, teachers have insufficient understanding of students' knowledge background, computer experience before class, and learning motivation. In the course of "Introduction to Computer Science" based on distance learning, it is necessary to evaluate students' computer experience and learning motivation before the class. According to the results of this research, students can fill out the 7 independent variable questionnaires used in this research before class. Then, the teacher can predict the student's final academic performance at the beginning of the class based on the prediction model. This can help teachers obtain early warning effects and then provide counseling strategies for students who may not get enough grades at the end of the semester in the "Introduction to Computer Science" course based on distance learning.
In addition to establishing a predictive model that is significantly related to final learning performance, this study also found that although 33% of the students have studied "Clerical processing", 14% have studied "Briefing design and production" and 5% have studied "spreadsheet". And 78% of students said that their most familiar computer skills were "Clerical processing", 32% were for "Briefing design and production" and 6% were for "spreadsheet". But 48% of students still think that the computer course they need most are "Clerical processing", 64% are for "Briefing design and production" and 55% are for "spreadsheet". It can be seen that a large part of the students at OUK are unfamiliar with general office software and feel the need. This result is the same as the Kuo's (2020) study. Similarly, in response to the government of the Republic of China (Taiwan) is promoting the Open Document Format (ODF) through its ODF-CNS15251 policy. And to meet the needs of students, it is recommended that OUK launch Libre Office free software-related course, such as Writer, Impress, and Calc software. Allow OUK students to better meet the expectations and goals of the government and society, and make themselves more competitive in the workplace and adapt to society.