Time series analysis is a statistical research approach appropriate for an important class of longitudinal research designs. Time series designs typically involve single subjects or research units that are measured repeatedly at regular intervals over a large number of observations. Time series analysis is a great example of a longitudinal design. A time series analysis can help us to identify the pattern of change in a behavior over time or evaluate the effects of either a planned or unplanned intervention. Intervention Model
There is yet to be one time series model that is claimed to be the best. One design model that is often used in a school setting is an intervention model. This model involves the analysis of the effects of an intervention that is applied to an individual subject or unit (Crosbie, 1993). This model is commonly referred to as an “interrupted time series analysis”. Repeated measurements are taken before and after the intervention in order to provide a sufficient number of data points to conduct a statistical analysis to evaluate the effects of the intervention. Such investigations can be very useful in trying to understand the process and result of the intervention. One drawback of this model is the generalization of the effects on an intervention for a large population since investigations are usually conducted on single subjects (Crosbie, 1993). Multiple Baseline Model
Another commonly used model of time series analysis is a multiple baseline model. This model involves the measurement of multiple people both before and after a planned intervention. This model has several advantages over the intervention model which only measures a single case. In a multiple baseline model, the start of an intervention is staggered. This assures that the changes in behavior are due to the intervention rather than an outside factor. By gathering information from multiple subjects, generalizations for a larger population are likely and accurate. One disadvantage of conduction a multiple baseline model is that implementation can be time consuming and may require substantial resources (Cuvo, 1979). Advantages of Time Series Designs
Time series analysis allows for a broadening of the range of questions that can be asked in a study beyond a simple investigation of whether the intervention has had an effect (McCleary & Hay, 1980). Time series analysis has some important advantages over other research methods in that it provides the opportunity to investigate the pattern of intervention effects across time. There are several methods that can be used for researchers to gain the information necessary to analyze patterns in behaviors. These methods include keeping a daily diary, technological assisted measurement, and telemetrics. Disadvantages of Time Series Designs
Where using time series analysis can be beneficial to evaluate the affects of an intervention, it can also have some disadvantages. One challenge that can emerge is case of missing data. Missing data is an unavoidable problem and can present a number of unique problems. Life events will result in missing data and researchers will have to figure compensation into their study. Another challenge in time series analysis is the reliability on the use of a computer program. This could lead to computational issues. However, time series analysis should now be viewed as representing one of a number of potential methods of data analysis available to all researchers rather than a new and difficult procedure. Example of Potential Research Study
Time series analysis is a great research method for schools and counselors to implement. One example of a way that this methodology can be used in a classroom is when one is investigating the effects of an intervention for “talking out” in a second grade classroom. Analysis would begin with observing the number of cases of disruptive “talking out” behavior. Then, an intervention, such as praise or tangible rewards, is put into place. Lastly, data would be collected after the intervention is put into place to investigate the effects of the intervention.
Sheperis, C.J., Daniels, M.H., & Young, J.S. (2009). Counseling Research: Quantitative, Qualitative and Mixed Methods. Upper Saddle River, New Jersey: Pearson. McCleary, R., & Hay, R.A., Jr. (1980). Applied time series analysis for the social sciences. Beverly Hills, CA: Sage. Crosbie, J. (1993). Interrupted time-series analysis with brief single-subject data. Journal of Consulting and Clinical Psychology, 61, 966-974. Cuvo, A.J. (1979). Multiple-baseline design in instructional research: pitfalls of measurement and procedural advantages. American Journal of Mental Deficiency, 84 (3), 219-228.