## Question 814:

1## Answer:

No answer provided yet.If I understand your particular need, it appears that you want to compare the outcomes of two groups of students on some measure, for example grades or attendance. One set of students study in 7 daily periods of 45 minutes per period. The second group of students can choose their classes and have 90 minute periods.A 2- sample z-test would be used if you knew the population standard deviations for both groups of students in the two types of periods. It is rarely the case that one knows the actual standard deviation of the population and in fact, the z-test is rarely used in statistics. The only time I know it's used is with standardized tests when everyone has been tested. If you don't know the population standard deviation you estimate it using the sample standard deviation.

What would likely serve you better is the 2-sample t-test. The 2-sample t-test will tell you if the difference between two means is greater than what you would expect from chance alone. You can use a sample of any size (above say 2) and use the sample standard deviations. Once you have data on some measure (e.g. attendance or grades) for both groups of students, simply copy and paste the data into the 2-sample t calculator. It will tell you whether the difference is statistically significant given the sample size, two standard deviations and the means.

You would want to setup some type of data collection system by which :

- Identify your dependent variable (grades, attendance record, some survey data or other measure).
- Sample a reasonable amount of students from each of the two types of period groups (aim for between 30-60 in each group) to be able to detect a medium size difference. Keep in mind that any difference between these groups might be small and you'd need many more students to detect this. However, small effects sizes aren't very noticeable--it all depends on what you're measuring and your research goals.
- Once you've completed sampling from the groups, conduct the 2-sample t-test on the raw data to determine if there is a difference than what you'd expect from chance fluctuations.