University of Alberta
STAT 252
Question 1 (a) The study is focussed on teaching strategies for high school math students, thus the population of interest is all high school math students. The subjects were all 120 math 30 students at a particular high school. The response variable is the students midterm score on a standardized test. The factor is the teaching strategies that
...[Show More]
Question 1 (a) The study is focussed on teaching strategies for high school math students, thus the population of interest is all high school math students. The subjects were all 120 math 30 students at a particular high school. The response variable is the students midterm score on a standardized test. The factor is the teaching strategies that are composed of different combinations involving homework, quizzes and computer tutorials. There is also the control treatment, which includes only lectures. (b) This was a randomized experiment, as all students were randomly allocation to the different teaching strategies. Thus, causal conclusion can be made regarding the effects of the different treatments. The experiment was conducted on only high school students from one particular high school. There should be caution when making any generalizations to students in other high schools unless it can be assumed that those in this high school are a representative sample of the population of all high school students. Question 2 (a) The means and standard deviations for the six treatment groups are displayed below: Report SCORE 60.93 20 8.229 78.43 20 6.927 82.53 20 6.899 62.09 20 6.398 69.82 20 9.573 68.91 20 9.309 70.45 120 11.125 STRAT C H HT Q QT T Total Mean N Std. Deviation It appears that the treatment with homework and computer tutorials is most effective with the treatment with only homework just behind. The control group with no additional components to the lecture is the least effective, as is probably expected. Note that the treatment with only quizzes is not much better than the control group, indicating that quizzes have a very small effect. This is also apparent by comparing the ‘QT’ group to the ‘T’ group. There aren’t any significant differences among the standard deviations, which vary from 6.398 to 9.573. No standard deviation is twice any other, thus the assumption of equal variability seems to be ok. (b) The side-by-side boxplot are displayed below: N = 20 20 20 20 20 20 Midterm Scores
[Show Less]