Current study's research questions are: 1) What are some key challenges that university staff face when adopting E-learning? 2) What are some socio-demographic and occupational factors that affect University staff's acceptance of E-learning?
To identify the most serious challenges that university staff face, I calculated mean rating of each challenge variable across subjects. The average rating of each challenge variable represents the overall seriousness rated by subjects. Challenge with higher average rating is considered to be more serious.
To measure the overall acceptance of E-learning, I computed the average of Y1, Y2, Y3, and Y4 variables, which measure different aspects of acceptance of E-learning. The overall acceptance is represented by the acceptance variable.
There's no missing value in the selected variables. However, I recoded some categorical variables by changing the number levels into string levels. For example, in the original Gender variable, 1 represents male and 2 represents female. It is hard for readers to get what 1 and 2 mean in graphs. As a result, I change all 1 into male and 2 into female for Gender variable.
To explore the relationship between socio-demographic and occupational factors and the acceptance of E-learning, I firstly used different methods for categorical and continuous variables. For categorical variables (Gender, Residence, Maritalstatus, Haveyouhighinternetspeedathome, HaveyoutaughtanonlinecoursebeforeCovid19), I displayed a table of mean and standard deviation of acceptance for each level so that the group difference is clear to see. In addition, I displayed a boxplot for each categorical variable. The boxplot provides a summary of the first, second, and third quartiles, the minimum, the maximum and outliers. The minimum and the maximum are found at the end of the whisker, the solid line. The lower boarder of the box represents the first quartile (Q1), the middle thick line represents the second quartile (Q2), and the upper boarder of the box represents the third quartile (Q3). An outlier is data point that is significantly distant from the rest of the data and is located outside the whiskers of the box plot. An outlier can also be found numerically, 1.5 times the interquartile range (Q3-Q1) above the upper quartile and below the lower quartile.
For the countinuous variable, yearsofCollegeTeachingExperience, I display a scatter plot to visualize the relationship between it and acceptance of E-learning. The line in the scatter plot shows the general relationship trend. I also calculated the correlation coefficient between yearsofCollegeTeachingExperience and acceptance. The correlation coefficient shows the strength and direction of the linear relationship between two variables. The value of r ranges between −1 and 1.
Lastly, I used multiple linear regression model to statistically examine the relationship between all socio-demographic and occupational factors (IVs) and acceptance of E-learning (DV). The multiple linear regression model allows us to capture the relationship between a specific explanatory variable and the outcome variable when controlling for other variables. In other words, multiple regression analysis allows us to evaluate the strength of the relationship between each explanatory variable and outcome variable while eliminating the influence of other explanatory variables. In the model, the intercept is the predicted outcome value for a unit with a score of 0 on all the predictor variables. The slope is the change in the predicted value of the outcome associated with a one unit increase in the predictor while holding constant the values of the other predictors in the model. The regression analysis is significant at 5% error margin (0.05 alpha level of significance). The p-value for each term tests the null hypothesis that the coefficient is equal to zero. A low p-value (< 0.05) indicates that we can reject the null hypothesis, meaning there's a significant relationship between the independent variable and dependent variable.
I also checked whether the multiple linear regression model satisfies the assumptions for statistical inference: linearity (the dependent variable is a linear combination of the independent variables and residuals), constant variance (the error has a constant variance for all values of independent variable), normality (error term is normally distributed), and independence (the errors for each unit are assumed to be independent). I also calculated Variance inflation factors to detect if there's any collinearity issue. If value of VIF is greater than 10, there may exist a severe problem of multicollinearity.
In conclusion, the key challenges that university staff faced when using E-learning are difficult applying distance learning for practical sessions and courses, technical problems, insufficient unstable internet connectivity, lack of suitable online environment at home, and lack of incentives and compensation for Internet outside the university. According to multiple regression analysis, factors that can significantly affect university staff's acceptance of E-learning are gender, marital status, whether there's a high speed internet at home, and years of college teaching experience.
Related the findings to literature review, gender and years of teaching experiences are found to be significantly related to acceptance of E-learning in current and previous studies. Major barriers that university staff face found by current and previous studies are unstable internet connectivity and technical problems.
Since E-learning has gradually become a dominant section of education worldwide, it is important to address the key barriers and challenges that users face to make E-learning easier and to enhance using efficiency. E-learning developers and university can prioritize the most serious barriers that staff face and solve the remaining challenges gradually. E-learning offers many advantages such as flexible time and location. As a result, we need to encourage more university staff to accept E-learning so that they can utilize E-learning tools to diverse students' learning options and to enhance students' learning experiences. E-learning can also facilitate university staff's teaching. Current study has identified several factors that affect the acceptance of E-learning. By understanding these factors and significant variables found by other studies, educational institutions can figure out efficient ways to enhance staff's acceptance of E-learning.
However, current study has some limitations. First of all, the sample size only includes 346 staff from only one university in Egypt. The results cannot be generalized to all university staff population around the world. Future studies can collect more diverse sample. Secondly, considering potential socio-demographic and occupational factors related to acceptance of E-learning, current study investigated only six factors. Future studies can incorporate more associated factors such as age, income level, teaching subjects, and so on. Thirdly, since the multiple regression model in this study doesn't satisfy homoscedasticity and normality assumptions, the accuracy of regression analysis results may be weakened. Future studies can transform certain variables or use more valid measurements to address the assumption issues.