a more professional study I would of liked to use a more random design. In my sampling frame is stuck basically to a place where I knew many different groups come in and children frequently visit. There was no list to really find out information regarding this and it wasnt like I would just mail out my surveys. So I can say I used a little bit of stratified and then simple random technique. I picked one group, which was people who came to the facility, and then out of those I just randomly picked some people to fill out the simple questionnaire. An example of the survey is on the last page but I would like to go over the significance of the questions. I wanted to be careful in my questions so that the person felt comfortable in telling me whether they smoked. So in the first question I kept it to the parents. I asked if anyone smokes in his or her household. Then asked how frequently does that person smoke? Next I wanted to slowly make the change to parent to child. Then I ask do any sibling smoke and how frequently. Then I asked if they have ever been offered to smoke and by whom. Next was did they accept the offer, and finally how frequently they have smoked if ever. Finally if they did accept what was the most influencing factors. I felt that I made the participants feel more comfortable because I avoided the question do you smoke. They never feel as if admitting to smoking. Then I dealt more into what where they key factors in making their decisions and not if they did it or not. In this survey I wanted the quasi-random design with interrupted time series. Though it was hard to do this I wanted to show more so that the age influenced more then the parents. In a real setting I would have followed the same child over a period of time and every few years check in to see what the habits are and the reasoning behind it. I dont believe the study would have problem with maturation and the selection would definitely be ...
