Question

You believe that requiring students to attend a series of presentations on study skills will improve their grades. You are concerned that boys and girls might react differently to such a series, and that students in grades 9 and 10 might react differently than those in grades 11 and 12. You set up an experiment involving 100 students in ninth and tenth grade and 100 students in eleventh and twelfth grade. (Assume there are equal numbers of girls and boys in each grade.) You intend to measure improvement based on pre- and post-treatment grade-point averages.

Describe the design of an experiment to help you determine if a set of presentations on study skills is effective in improving grades.

Answer:

Suppose you have 50 students and need to assign them to two groups of equal size. Describe a randomization procedure that would achieve this.

Answer:

Suppose you have 50 students and need to assign them to two independent groups. Describe a randomization procedure that would achieve this.

Answer:

Describe an appropriate method of analysis for the data you collect from this experiment.

Answer:

Answer 1

Solution:

1. Randomized block Design is a suitable design for the above problem. Where the blocks will be the 9 10 11 and 12 grade students and the treatments are the series of presentations on the study skills that the students attend.

2. One method is the simple randomization, where students are assigned randomly in the two groups (say for example 25 students in each groups). that is each student can be numbered from 1 to 50 and chits can be made. Then chits can be taken then it can be allocated to group A and the next to group B, next to A and so on.

Another method can be Systematic random sampling where every rth component (here student) can be assigned to either of the groups. that is each student can be numbered from 1 to 50 and every 5th student can be assigned to group A (say) and the next 5th student to group B and next one to A and so on, until each student has been assigned to any one of the group.

3. Randomization to two independent groups can be based on [a]. All male students in one groups and female students in the other. [b]. or those students who have got a grade of 70% in grade can be put in one group and those less than in another groups, so these groups can be called as two independent groups.

4. Method of Analysis:

To test the performance of the above two groups one can do Wilcoxon Signed Rank Test ( paired data ).

Where the Null Hypothesis (H) : Whether the two groups differ significantly in their performance. and the alternative hypothesis (K) : H is false. And accordingly if the calculated value is greater than or equal to the tabulated value then we reject the null hypothesis. o.w we fail to reject the null hypothesis.