It is expected that given the sensitive nature of the subject, there will be large numbers of selected participants who will decide not to participate, will drop out, or whose parents will not sign the consent form. All participants selected will be taken from the middle school and high school age teens, so aged from 12 through to 17 years old. Half of the sample, so around 400 participants, will be selected through simplified random sampling from school registers of the appropriate years from schools in the district. To ensure that an adequate number of the targeted groups of interest are selected, the other 400 participants will be selected by stratified random sampling. This will involve selecting 100 participants according to each of the following categories, with participants equally taken from each strata: Race: Caucasian, African-American, Hispanic, Asian

Socioeconomic status: Income under $25,000 per annum, over $25,000 and below $50,000, over $50,000 and under $100,000, over $100,000

Family structure: Married parents, divorced parents

Pregnant teens and non-pregnant teens

Data Collection Tools

The main tool for collection of data on the student's attitudes will be my own self-report survey, which will be developed according to the specific questions of interest. In order to compare the answers given with the opinions of family and caregivers, the Family Assessment Measure (FAM-III) will also be administered. This is a self-report measure which assesses the strengths and weaknesses within the family. It would be given to all pre-teens, teens and adult family members within the household. Comparison of the different responses within the family would then give the counselor an overview of how the different family members view their levels of interaction.

Data Collection Procedures

Permission will be sought from the legal guardian of the student prior to commencing data collection. The student will be surveyed within their school's library during school hours. This will allow for data collection to be a non-disruptive process within that child's school life, and may also encourage higher participation than if conducted outside of school hours. Performing the survey in the school library will also allow for up to 15 students to be called in to take the survey simultaneously. After selection, each student will have the choice of whether they wish to participate in the survey or not. This means that if they feel uncomfortable when they come to the library to complete the survey, they may still withdraw. All answers which are given will remain private and confidential, so students will be reassured that their parents and others will not have access to their answers. Therefore coding of papers would be required to ensure that measurement tools may be matched, but without revealing who the specific participants were. It would be expected that I would need to also offer after-school sessions for family members to come to the school to complete the FAM-III measures, or that I would need to visit the student's house. It may be possible that using a computer-assisted data entry system may help to overcome the logistical issues which may arise from this task. This has been used in previous studies which have collected information from both parents and teens on sexual behavior and communication (Aspy et al., 2006), but this would require further investigation.

Data Analysis

Although the main focus of the study is on qualitative data, systematic analysis will still be applied in deriving conclusions. This will involve coding of the data which is collected to accurately assess trends in responses and agreements between family members on the level of communication occurring...

From these trends it should be possible to draw conclusions as to how important parent-teen communication is as a risk factor or preventative factor in teen pregnancy. The quantitative data collected will be analyzed by standard statistical tests such as correlation coefficients and ANOVA analysis to establish the factors which may be associated with teen pregnancy. Scoring qualitative responses from participants may also allow for such statistical methods to be applied to deduce whether there are associations between parent-teen communication and these other demographic variables.