Time Series Design

Read Complete Research Material

Unit 7 Discussion 1 & 2

Unit 7 Discussion 1 & 2

Discussion 1

Interrupted Time Series Design

Time series refers to a group of readings of a variable observed and noted down after certain time intervals (Ferron & Rendina-Gobioff, 2005). For instance, observations might consist of moods of students in a classroom that is noted down and recorded every day for an entire consecutive semester; this would allow identification of changes in moods over a period of time, if there are any. An interrupted time series design differs from the traditional time series design in the fact that during the consistent observational period, the independent variable is suddenly changed (either artificially or naturally), which leads to an ensuing and automatic change in the dependent variable. For instance, if (let's say) the moods of students is the dependent variable, and the color of classrooms is the independent variable; observations of moods are being made from the start of September till Halloween, and then suddenly after Halloween the color is changed to yellow (in an effort to instill energy and enthusiasm or maybe to experiment to check that). Observations are not halted - they continue on till the end of the semester. The main aim during data analysis would be to look for changes in moods (whether or not it alleviated) right after the room was painted yellow. The diagram below shows the results that might validate the proposed hypothesis.

It should be noted that this design is similar to the pretest-posttest design but with various pretests and various posttests. The benefit of this method is that it offers larger support to the idea that the fluctuation in the dependent variable was due to the change in our said independent variable, instead of being a series of random, non-frequent deviations. For instance, if students' moods were changing ...