Study area and Design
A cross sectional study was conducted between February 1 to March 1, 2009 in Fasiledess preparatory school, which is found in Gondar town with estimated residents of 250,000 populations. Gondar town is located 730 km away to the North West from the capital city, Addis Ababa. The school prepares students for National collage entrance examination.
Study population and sampling procedure
There were a total of 658 preparatory students (grade 11 and grade 12) in Fasiledess preparatory school during the study period distributed into 11 sections (6 sections (A-F) for grade 11 and 5 sections (A-E) for grade 12). The number of students between sections were varies ranging between 50 and 70. Assuming homogeneity in academic status among students in the same grade, a total of 240 students were recruited for the study. Stratified proportionate sampling was used to select study participants. Shortly, students' list was obtained from student registration books and stratified into grades. Five sections were selected randomly (3 sections from grade 11 and 2 from grade 12) from stratum and of which proportionate numbers of students were selected using systematic random sampling to obtain the total sample for the study.
Data collection
The study instrument was a self-administered questionnaire which comprised of four parts. Part- A related to student's sociodemographic background, Part- B on knowledge regarding HIV and other Sexual transmitted infections (STIs), Part- C on students' attitude scale towards HIV/AIDS and Part -D on high risk behavior or practice related to HIV& other STIs transmission.
The knowledge, attitude and practice questionnaire was modified from the instrument used by a survey on HIV/AIDS knowledge, attitude & practice (KAP) reported by the Department of Education, Free State South Africa (2006) which was adopted from the WHO AIDS Questionnaire (WHO 1990).
Knowledge was assessed using a 31-item questionnaire which includes knowledge on STDs, ways of risk reduction for HIV transmission, predisposing risk behavior and practice to HIV and other STIs, symptoms of STDs other than HIV/AIDS, treatment of AIDS, prevention methods of HIV and other STIs and the contribution of other STDs acquisition for the spread of HIV infection.
Attitude was assessed using a 10-item questionnaire which includes attitude: towards screening of HIV and other STIs, towards AIDS patients, towards gender equality as related to sexuality and towards visiting clinics for cases of STDs other than HIV/AIDS. The questions on high risk behaviors had 11 items related to unprotected sex and needle sharing.
The English questionnaire was translated into simple Amharic (local language) and back translated into English. Pre- test of questionnaire was done on twenty students and the result was used to improve the phrasing of questions in the questionnaire. Questionnaire validation tests showed that the Alpha Cronbach was 0.87 for knowledge, 0.71 for attitude and 0.72 for risk behaviors.
Scoring
For knowledge, each right response was given a score of 1 while a wrong or unsure response was scored 0. Total knowledge scores can range between 0-31. Knowledge scores from 0 to 15 were considered as poor knowledge while knowledge scores more than 15 was considered as having good knowledge regarding HIV/AIDS. Attitude towards HIV/AIDS patients was assessed using a 10-item questionnaire where attitude scores between 0 to 5 were considered as negative attitude, and scores 6 to 10 were considered as positive attitude. High risk behavior or practice was assessed using an 11-item questionnaire where report of at least one negative behavior related to HIV transmission is considered as having high risk behavior.
Data management and analysis
During data collection process, the data were checked for completeness and any incomplete or misfiled questions were sent back for correction. Data were double entered and analyzed using SPSS-15 statistical software (SPSS Inc. Chicago, 2007). Descriptive statistics were used to give a clear picture of background variables like age, sex and other variables in well structured questionnaire. The frequency distribution of both dependent and independent variables were worked out. The association between variables was measured and tested using chi square and odds ratio. P-value < 0.05 was considered as significant in all cases.
Ethical consideration
Ethical clearance was obtained from University of Gondar collage of Medicine and Health Science Ethical Review Board. Prior to data collection, all study participants were given information on the study and assured that all data is confidential and will only be analyzed as aggregates. All respondents signed the informed consent form before participation.