T-tests For Independent And Related Samples Essay Example for Free

T-tests For Independent And Related Samples Essay Research Problem   Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚   With the increased violence that occurs in our society, it is a reality that most children on social welfare or those who have been placed in foster care have had negative experiences. Some children need to be counseled and psychologically helped in order to regain their sense of self and to help them cope with the difficulties that lie ahead of them. The most evident problems that these children have are in terms of their education and schooling. Despite the efforts of social welfare to help them get the education they need, most children drop out of high school. This study would like to determine the difference in the graduation rates of children in foster care who received counseling and those who did not receive counseling. Research Question   Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚  Ã‚   Is there any difference in the graduation rates of children in foster care who had received counseling compared to those who did not? Hypothesis: Null hypothesis: There is no significant difference in the graduation rates of children in foster care whether they received counseling or not. Alternative hypothesis: There is a significant difference in the graduation rates of children in foster care when they received counseling compared to those who did not. Independent variable: Counseling sessions Dependent variable: Graduation rate (high school) Statistical Design: For this study, the appropriate statistical design is the related samples design (within groups) since the goal is to determine whether counseling sessions significantly increased or decreased the graduation rates of children in foster care which satisfies the assumption of the related sample t-test measures. Effect size: The effect size between the two means would indicate the degree or level of the difference between two groups, it could be high, moderate or low effect which would also indicate just how significant the difference between the two means are from each other. A probability value would not be able to clearly point out just how big the difference is. Table 1 t-test Result t df p.05 Graduation rate (with counseling vs without counseling) 2.69 73 .009 N = 30; ES = .60 Table 1 indicates that children in foster care who had received counseling were found to significantly have higher rates of graduation than those who have not received counseling. The t value (2.69) was found to be statistically significant (p.05), thus we reject the null hypothesis and accept the alternative hypothesis that children in foster care are more likely to graduate if they had received counseling. It was also found that the effect size for this result was .60 which is according to Cohen (1988) is of medium degree. This would mean that the difference in the graduation rates for the two groups may be statistically significant but the difference is actually not so big that other factors may actually come into play with why children who receive counseling have better graduation rates. Statistics N = 30 Mean 1 = 14.75; SD = 6.03 Mean 2 = 10.74; SD = 6.69 t = 2.69 df = 73 p = .009