Discussion Weaknesses
The study has inherent flaws which pertain to the validity of our study. First, data was collected by a voluntary survey, which was sent out to a randomly selected students from our high school population. It was sent through our lunch study halls, where the announcements are supposed to be broadcast daily. One of the weaknesses that we saw immediately was that some students did not have a lunch study hall. However, there were other confounding variables that we foresaw: Lunch study halls did not always have the TV they were broadcast on plugged in or turned on, surveys were returned with unusable answers, and some surveys were not returned leading to a nonresponse bias. In sending out surveys, we saw that some students had responses that had responses that didn’t make sense.
We had an issue of our sample size, because we only had 24 surveys respond correctly out of 100 that were sent out. It also did not meet the requirements for the assumptions of our sample mean t-test, where we would have needed a sample size of 30, because we didn't know that our population was normally distributed. Our sample size was extremely small, in part because many study halls had problems playing the announcements, and so we did not realize that there were not enough surveys returned to be able to gauge an accurate response from. Doing this research is very hard, and it is not easy to conduct. We have some suggestions for anyone who would like to conduct this study in our Suggestions discussion.
Discussion: Extrapolation
In our study, we should be able to extrapolate the data only to the entire population of students at our own North Olmsted High School who have lunch study halls that can see the video announcements. We cannot extrapolate beyond our high school only having data for NOHS. However, it is not too far of a stretch to extrapolate to other high schools that have video announcements, but this assumption is false because other schools were not included in our sample. They have not been included in this study because the population of interest was lunch study hall students that watch the announcments.
Discussion: Suggestions
There are many things that we can look into doing in the future in relationship to this study. However, we aslo have to keep in mind that research is very hard, and that not all of this can be accomplished. We did, however, come up with these suggestions:
1. Increase the number of initial surveys sent out. If the same study were to be repeated, more surveys should initially be sent out to account for the large proportion of unusable responses returned. This would maximize the chance that we would have a large enough sample size to conduct a suitable test on the data.
2. Change the type of test. Collect data on the actual length of the announcments, and compare that data to the other data that we gathered. Comparing these data sets would allow us a better understanding of what the viewers are reporting based upon the lenght of the announcments.