Introduction
This report considers whether gender, quiz result, lecture and SGT attendance affect university students’ exam results. The effectiveness of learning platforms must be evaluated so students reach their full potential. To achieve this, a sample of previous BSB123 students’ results and attendance will be summarised, analysed and interpreted.
Data
2.1 Research
Stanford University found that on average, males outperform females on mathematics tests (Niederle & Vesterlund, 2010), which would be expected in the statistics course BSB123. The majority of the BSB123 sample (56.8%) was female, which may affect exam results. Additionally, quiz results are positively correlated with exam results (Wambuguh & Yonn-Brown, 2013) and students …show more content…
t-test of Male and Female Exam Results
Equal variances were assumed as male and female variances are similar. The null hypothesis, that the difference between male and female results was 0, was not rejected (0.50 is greater than 0.05 at 5% significance) for the alternative hypothesis, that the difference between male and female results would be greater than 0. Therefore, there is insufficient evidence that males outperform females on exams.
The effect of lecture attendance on exam results was also examined using a two-sample t-test.
Table 5. t-test of Exam Results with High and Low Lecture Attendance
At 5% significance, the highlighted one-tail p-value of 0.002 is smaller than 0.05. Hence, the null hypothesis that those who have high lecture attendance achieve the same exam results as those who do not is rejected for the alternative hypothesis that high lecture attendance increases exam results. Hence, using this t-test, there is sufficient evidence that higher lecture attendance increases exam score.
Model
4.1 Interpretation
The regression output in table 3 can be expressed as an equation to model exam results.
y_(exam result)=2.39267+1.48505〖 x〗_1+0.98494 x_2+2.81056 x_3+0.54784〖 x〗_4 where: x_1= quiz …show more content…
x_1 has a coefficient of 1.48505. Holding all else constant, every 1-mark increase in quiz results should increase exam results by 1.5. The p-value of 1x10-16 means that quiz results are significant to exam results at 5% significance. This is logical and consistent with research as periodic testing of unit content should improve students’ understanding.
x_2 has a coefficient of 0.98494. If 6-10 lectures are attended, exam result would increase by 1. The p-value of 0.41745 however, is less than alpha at a significance of 5% and 10%. Hence, high or low lecture attendance seems insignificant. Yet research and a two-sample t-test of exam scores (table 5), found sufficient evidence that higher lecture attendance increases exam score. Despite this, in the model, lecture attendance is not a good predictor of exam result, as lectures are recorded, on-campus attendance is unnecessary and students do not practice questions. Furthermore, the 1-mark increase is small. Hence, lecture attendance is insignificant in this
This report considers whether gender, quiz result, lecture and SGT attendance affect university students’ exam results. The effectiveness of learning platforms must be evaluated so students reach their full potential. To achieve this, a sample of previous BSB123 students’ results and attendance will be summarised, analysed and interpreted.
Data
2.1 Research
Stanford University found that on average, males outperform females on mathematics tests (Niederle & Vesterlund, 2010), which would be expected in the statistics course BSB123. The majority of the BSB123 sample (56.8%) was female, which may affect exam results. Additionally, quiz results are positively correlated with exam results (Wambuguh & Yonn-Brown, 2013) and students …show more content…
t-test of Male and Female Exam Results
Equal variances were assumed as male and female variances are similar. The null hypothesis, that the difference between male and female results was 0, was not rejected (0.50 is greater than 0.05 at 5% significance) for the alternative hypothesis, that the difference between male and female results would be greater than 0. Therefore, there is insufficient evidence that males outperform females on exams.
The effect of lecture attendance on exam results was also examined using a two-sample t-test.
Table 5. t-test of Exam Results with High and Low Lecture Attendance
At 5% significance, the highlighted one-tail p-value of 0.002 is smaller than 0.05. Hence, the null hypothesis that those who have high lecture attendance achieve the same exam results as those who do not is rejected for the alternative hypothesis that high lecture attendance increases exam results. Hence, using this t-test, there is sufficient evidence that higher lecture attendance increases exam score.
Model
4.1 Interpretation
The regression output in table 3 can be expressed as an equation to model exam results.
y_(exam result)=2.39267+1.48505〖 x〗_1+0.98494 x_2+2.81056 x_3+0.54784〖 x〗_4 where: x_1= quiz …show more content…
x_1 has a coefficient of 1.48505. Holding all else constant, every 1-mark increase in quiz results should increase exam results by 1.5. The p-value of 1x10-16 means that quiz results are significant to exam results at 5% significance. This is logical and consistent with research as periodic testing of unit content should improve students’ understanding.
x_2 has a coefficient of 0.98494. If 6-10 lectures are attended, exam result would increase by 1. The p-value of 0.41745 however, is less than alpha at a significance of 5% and 10%. Hence, high or low lecture attendance seems insignificant. Yet research and a two-sample t-test of exam scores (table 5), found sufficient evidence that higher lecture attendance increases exam score. Despite this, in the model, lecture attendance is not a good predictor of exam result, as lectures are recorded, on-campus attendance is unnecessary and students do not practice questions. Furthermore, the 1-mark increase is small. Hence, lecture attendance is insignificant in this