Given Hapunda

97 Advantages of Time Series Evaluation Designs Time series deigns have the advantage of detecting changes or trends in data (within the comparison and experimental groups) before and after the intervention. To that end, these designs are stronger than single pre-test/post-test designs as they are able to detect any changes or trends that could be caused by problems of maturation, testing, and instrumentation (Marlow, 2011; Rubin & Babbie, 2013). For example, the effects of maturation on the girls’ knowledge of contraceptives could be detected in the difference between pre-test scores later on (Traeen, 2003). Disadvantages of Time Series Evaluation Designs The main problem associated with time series designs is the threat to internal and external validity. On one hand, history is the main problem associated with this design’s internal validity since the comparison group is absent. For example, exposure to other sources of information on contraceptives in the past would confound the effect of training program on contraceptives. On the other hand, history- treatment-interaction is a possible threat to external validity as prior knowledge of contraceptives would interact with the training program. So, mediation or intervention that would work in some conditions may not in others. Additionally, repetitive exposure to the same tests would also be a problem (Marlow, 2011). Pre-test Pre-test Pre-test Programme/intervention Post-test Post-test Post-test Figure 4: Time Series Designs Focus Box 9: Time Series Design A time series design was used to study a dynamic population of homeless shelters in the city of New York from January 1996 till December 2003. Numbers were recorded at the end of each month of the average daily census, monthly entries, as well as monthly exists. The results revealed a complex pattern of issues and that many factors were related to the number of people entering and leaving the shelters. For instance, placing adults from the shelter into subsidised permanent houses reduced the shelter population but at the same time increased the number of people coming in (O’Flaherty & Wu, 2008).