To date, there have been over 100,000 times a teen has opted to attend meetings for support and outreach.
While the attitudinal study showed positive results regarding the teens’ success in addressing personal challenges, two factors arose while evaluating the data. First, many were asked to answer questions involving specific chemical use, which were unrelated to their personal addiction. For example, Question 2 asked teens to circle the number of times per day they used drugs or a mind altering substance. Fifty out of the 81 students, who completed the program, circled a score of “0” for both the pretest and posttest. This suggested that those students did not use drugs on a daily basis during any point of the program. While it was gratifying to know that such a high percentage of students were not abusing drugs, it altered the research calculations. Students who circled “0” for both tests showed no improvement with that addiction. Therefore, the researcher performed two different calculations involving each of the ten research questions. The first test statistic included all who completed the program, including those who submitted a “0” for each of the pretest and posttest questions. A second calculation was performed in the same manner. However, those who submitted a “0” for each of the questions had those respective values removed. This resulted in a lower sample size for each question within the survey and a smaller degree of freedom for each t-statistic. While such calculations typically weaken statistical results for categorizing data as being statistically significant, they did not change the conclusions for any of our research questions. Further, the distribution of values for Question 2 still resulted in an approximately normal distribution, satisfying the conditions for a t-test. The calculations made for this condition were stated previously.
A second concern relates to the importance of having independent values from the subjects. In some instances, two or three data values came from the same student due to the repeated 6-session intervals in which the students participated in the program. One may assume that a data value from one test subject may influence other values from the same subject. However, this didn’t prove to be influential for Question 2. Had the additional values from Teen AA participants been eliminated from the sample of data, the remaining data would still produce an approximately normal distribution and provide a p-value smaller than the alpha cut-off of .05; thus, providing an outcome that’s statistically significant.