Descriptives

From table 3 above, the average value of the design and development variable is 7.12 which implies that the average scores of the participants are around 7. This is supported by the median which has a value of 7. The close proximity of the mean and the median may imply the lack of outliers in the scores. However, this may not be the case due to the scale of the scores which is small i.e. it is between one and ten. At the 95% level of confidence, the mean is between 7.06 and 7.17 which are the upper and lower limits of the mean respectively. The small difference between the upper bond and lower bound of the mean at a 95% level of confidence implies that the scores converge at a very high degree. Convergence in the scores is further supported by an interquartile range which has a value of 2. The least score is 2 while the highest score is 10. The size in the data is relatively large as indicated by the proximity of the maximum score and the range which has a value of 8. The skewness has a negative value of 0.118. Negative skewness indicates the skewness of the scores to the left. There is also evidence of a small outlier from the small value of the kurtosis. The data has a positive kurtosis value of 0.294 which indicates that the peaking of the standard normal distribution in the scores is relatively moderate.

The figure below shows the histogram for the distribution in the data.

The general shape of the histogram indicates the normal distribution in the data. The skewness to the left is observed form the histogram. The peaking in the histogram is relatively moderate for the scores. The deviation to the left is also observed from the histogram. These observations from the histogram are in line with the values from the descriptive statistics.

The figure below shows the box plot for the design and development variable.

The shape of the boxplot above indicates that there is actually a presence of outliers in the data, which is in line with findings by Peter J. Rousseeuw &Bert C. van Zomeren [16], which have small values which are indicated by the values below the boxplot. As a result of the presence of these outliers, the scores for design and development indicate a left bias distribution.

The normality of the scores was further tested by observation of the normal q-q plot shown in the figure below.

The QQ map shows data points and an oblique fitted line. The data points do not fit the oblique line well. This refutes the suggestion that the data follows a normal distribution.

Descriptive of Experts’ rating of Design and Development

The above is the exploratory data analysis of the student scoring data set, and the further exploratory data analysis of this variable in the expert scoring data set as shown by the table below.

The mean score of the design and development varies according to the experts is 6.96 while the median is valued at 7.0. The median in the experts and students’ scores of the same variable are equal but the mean is higher. Through the comparison of the skewness and kurtosis coefficients, consistency between the expert scoring data and the student scoring data is consistent which is consistent with previous studies by Richard A. Groeneveld and Glen Meeden [17]. The small difference between the intervals of the mean is still maintained together with the left skewness in both data sets.

Mean difference test of Design and Development and analysis results

It is crucial to check on the mean difference of the two variables in the two data sets to check on any significant difference between the mean of the score according to the experts and according to the students.

The variance test indicates that the variance of the scores between the two groups can be assumed to be equal because both data sets come from the same experimental group. The t-statistic value for the test for the difference between the means has a value of 1.838 and a p-value of 0.066. The t-test statistic value is less than the critical value from the t tables indicating leading to the acceptance of the null hypothesis at the 95% level of confidence. The null hypothesis for this test states that there is no significant difference between the experts’ scores and the students’ scores.

Gender differences of Design and Development in student scoring data set

Descriptive Analysis of Gender in student scoring data set

Analysis of the scores according to gender is important in the analysis of this data sets mainly because gender acts as a grouping variable. The figure below shows the descriptive statistics for gender.

From the descriptive statistics, the mean for males is 7.27 while that of females is 6.96. The average performance of women is slightly lower than that of men. The men are also expected to perform higher than the women as indicated by the lower limit of the mean is higher than the lower limit of women.

The figure below shows the box plot for the comparison of male and female performance.

From the box plots, it is observable that the boxplot form male scores is higher than the box plot for female scores indicating a higher number of the small outliers in males than in female scores.

Normality test of Gender in student data set

The figures below show the QQ plots for both male and female scores

From both figures, the fitted lines do not represent the data accurately although the male’s fitted line is a better performer. The inaccuracy of the fitted line indicates that the performance between both males and females is not normally distributed.

Mean difference test of gender in student data set and analysis results

The data set for the females and males if from one population hence the variance between the two groups is equal. The t value is 5.599 and the p-value is 0. The p-value of this test is lower than the 95% alpha level of confidence hence the null hypothesis for this test is rejected in favor of the alternative hypothesis which states that there is a significant difference between the two means. The scores in this variable are influenced by gender.

Gender differences of Design and Development in expert scoring data set

Descriptive analysis of gender in experts data set

It is crucial to determine the impact of gender on the expert rating. The table below shows the descriptive statistics of the scores graded by experts.

According to the descriptive statistics table, the average scores for the males and females respectively are 6.97 and 6.95 respectively. There is a small difference between the means. However, it is important to determine if the small difference between the means is significant at 95%.

The figure below shows the significance of the difference between the performance of men and women.

The null hypothesis for the test of the difference between means states that there is a significant difference between the means. The t value for this test is 1.36 while the p-value is 0.892. The p-value is greater than 0.05 indication the failure in rejecting the null hypothesis. For the experts score the gender of participants has no effect on the scores. This is different from previous findings from the student scoring data results.

The difference between students’ and experts’ rating of Delivery Voice

Descriptive analysis of Delivery Voice

Finally, delivery voice variables can be studied carefully. According to Carissa Portone-Maira et al.[19], The overall state of the data from a macro perspective is shown in table 15 and table 16 (in the appendix). The gap between the average scores of students and experts is not large.

Figure 1 student scores data

The histogram indicates a partial distribution which is oriented to the left. The students’ data on this variable shows a very strong peak distribution.

The range of the data as described by the box plot is relatively large. Many outliers are in the students’ data on voice notes as indicated by the values below the box plot.

From the QQ plot below it is easily discernable that the scores do not follow a normal distribution as indicated by the poor fitting of the line on the data set (R. A. Brace [20]).

The figure below shows the scores of voice form the expert data.

The scores show a left partial distribution but with a lesser degree compared to the voice scores in the student data. The peak distribution is relative to the normal distribution unlike that of the students’ data which is not relative.

Om the boxplot below the variable does not show any signs of possible outliers.

The QQ line plot shows a fitted line which does not represent the data. This indicates the presence of outliers that are very small and have minimal significance.