Weaknesses

There were some weaknesses and confounding variables in our study.  One of the weaknesses was the presence of non-response bias.  Out of the 100 surveys we sent out, only 52 surveys were returned.  This is a clear indication of non-response.  A number of students that we tried to contact were unavailable- they had moved, went to Polaris full time, etc.  The biggest weakness in our study was that we asked for numbers that are almost never exactly memorized by students.  If we had given them a month’s notice, our study would be much more accurate.  A confounding variable in our study would have simply been an expectation that people may have.  Students do not want to seem unpopular so they put down numbers higher than the true values.  This may result in response bias. 

Extrapolation

The data we found can be extrapolated to schools in the area.  This is because there seems to be no particular reason for students at North Olmsted High School to have different social habits than students at other schools.  Furthermore, this data could be extrapolated to similar age students around the country.  However, we would not feel comfortable extrapolating the data to adults or any other age group, since Facebook and texting habits most likely differ.

Suggestions

There are a number of ways in which our study could be improved and advanced.  One possible suggestion is to create groups of students based on grade level to provide more specific results.  The best way to improve the validity of the test is to provide the students with the survey in advance, and give them more than a day to complete it.  This would allow them to give us the exact number of Facebook friends they have and make a better estimate of the number of texts they receive per month.  Finally, I would suggest an increased amount of privacy for students when filling out surveys.  When students are filling out surveys amongst their friends they are more likely to lie than if they were to fill out surveys alone in a room.