Outline for Group Design

Formulate a proposal for a group design which is either an experiment, pre-experiment or quasi-experiment.

A) Name and diagram the type of experiment or quasi-experiment.

B) Describe the sample of subjects who comprise your experiment. Your description should be a summary of the number of subjects, their ages, sits., and other relevant demographic descriptors. What problem/need do the subjects share in common? Do they share any other characteristics which could be considered control variables? Describe your method of arriving at this sample. In other words, how did the subjects come to you or how did you get them to participate in this experiment? How many groups will be exposed to the experimental stimulus(i); will you employ a control group? Have you used randomization or some form of matching?

C) Name the dependent variable and describe when and how it will be measured. Use more than one method of measuring the dependent variable to increase the reliability and validity of your findings. Try to use pre-existing measurement instruments. Cite research literature that shows that the measurement instruments you plan to use have been proven to be reliable and valid. If you cannot find written support for your measurement instruments, explain why you believe that they are reliable and valid.

D) Describe the independent variable-- what treatment(s) or training will be provided by the clinical researcher or someone from his/her team? What theoretical or empirical evidence do you have which shows why this intervention should work?
How long will the treatment be provided? Be specific about the content and format to which each subject in the experimental condition would be exposed. How will you make certain that all subjects in the experimental condition receive the exact same treatment? Describe the training and background possessed by the worker who is carrying out the treatment.

E) Provide a well-written, one-tail experimental hypothesis indicating what you believe will be the effect of the experimental treatment. What test of statistical significance (t-test, f-test, or chi-square) will you use to see if changes are due to chance sampling errors or statistically significant?

F) Explain which threats to internal validity were controlled by the design and which were not.

G) Explain which threats to external validity are present in your design and which have been controlled. What suggestions can you make about how other researchers can use systematic or direct replication to enable greater generalization of results?
 
 
 
 

A. Choose a client from your field placement or develop an imaginary client-situation. Describe the behavior(s) and/or attitude problem(s) which this person has in clear and measurable terms. The problem must be one that can be measured incrementally ( a continuous variable)

B. The Baseline Phase

Devise a way of measuring the client's behavior and/or attitude during a baseline period. You can mention the initial frequency of the behavior, the intensity of behavior (or feeling or attitude) and/or the duration. In the case of an attitude, you can describe the position or value the client holds which results in problems for him/her or significant others. Make sure to define the problem in nominal and operational terms. What methods of measurement will be taken? What instruments, if any, will be used? Who will do the observing/measuring? How reliable and valid are the instruments and the measurers? Over what period of time will baseline measures be taken? How frequently will the dependent measure be taken? Will multiple (several indicators) measures be taken during the baseline period? What level of measurement are your dependent measures? Make sure to give a convincing explanation of your procedure for attaining baseline data.

C & D. The Intervention Phase

Your intervention should be formulated as a response(s) to your ideas about the dynamics of the problem and the client's strengths and needs. Think aloud and jot down your ideas about what contributes to the client's problem and what treatment, training, or intervention might help the client. Develop and give a rationale for a minimum of two strategies which logically could reduce the client's problematic behavior/attitude. Cite social science research literature that supports your treatment choices. Make sure to explain the strategy. The strategy(ies) is the independent variable. Explain the order and length of treatment (for each strategy). Name and provide letters which describe the single subject design you have formulated. Describe the measurement procedure during the treatment phase.

E. Type of time series design

Select a type of single subject design. Choose a design that shows two interventions and is as strong in internal validity as possible. Describe whether you chose a withdrawal, multiple baseline, and/or multiple treatment design to test the effects of the independent variables Include a diagram; for example: a-b-a-c.
 

Periodic measurement of the dependent variable(s)

At what time intervals will measurements of the dependent variable(s) be taken? Will the same measurements be taken at each measurement interval? Provide a letter diagram describing your clinical research design.
 

F. Threats to Internal Validity

What threats to internal validity were eliminated (or controlled for) and which were not eliminated by the design you formulated?
 

G. Discuss the Issue of External Validity and Single-Subject Designs

What threats to external validity were eliminated and which threats to external validity were not eliminated by your design?
 
 
 
 
 
 

Mahler, SWK 317--Evaluation Sheet for Single-Subject Design
NAME(s): __________________________

Points will be earned, grading your work in each of the following areas (WELL OK POORLY)