ABSC notes 3 Why don t behavior analysts use large group research designs Reason 1 Because they are interested in the behavior of individuals Reason 2 to use a large group design you need a large number of individuals with the same diagnosis problem Most behavior analysts counselors clinical psychologists etc are treating a handful of people at a time The single subject design allows you to scientifically evaluate your findings with only one client Reason 3 people don t want to be assigned to the control group In a single subject design everyone gets the treatment Reason 4 there may be ethical concerns about assigning a large group of individuals to a treatment that may have negative unanticipated effects The single subject design allows you to scientifically evaluate treatment before rolling them out in large group designs Reason 5 inferential statistics are a useful set of tools but study after study has shown that psychologists are not very good statisticians OTHER REASONS FOR USING THE SINGLE SUBJECT DESIGNS Single subject designs encourage the researcher to present their individual data graphically and let the consumer decide if it was successful Requiring consumers patients parents etc to understand statistics is ridiculous if the consumer cannot tell that the situation has been improved then the treatment is of no clinical use COMPARISON DESIGN We refer to this as an A B design A baseline period of observation in which the intervention is NOT in place B treatment period of observation in which the intervention IS in place COMPARISON DESIGN never used scientifically because we cannot know who what actually changed the behavior CANNOT account for time coincidences Something that happens around the same time as the intervention You could ten inaccurately assume that the intervention changed behavior when in fact it was just a time coincidence REVERSAL DESIGN Like a comparison with a reversal You can think of this as an A B A design This is a scientific approach Establishes cause effect i e a functional relationship MULTIPLE BASELINE DESIGNS Some treatments produce irreversible effects these are really the best treatments Sometimes it is unethical or impractical to remove the treatment Sometimes we want to examine an intervention across multiple people behaviors for one person or settings for one person IN ALL THESE CASES the multiple baseline design is the best option GENERAL STRATEGIES step 1 get some baseline data how much is enough Rule 1 do not introduce the treatment if the behavior is improving Rule 2 if things are getting worse go ahead and introduce the treatment Rule 3 minimize bounce variability before starting treatment Step 2 get some treatment data how much is enough Rule 1 learning takes time Too much I better than not enough Rule 2 You must convince us that the behavior change can be maintained Rule 3 Continue until data are stable Step 3 look for a convincing evidence of a treatment effect Rule 1 Are the baseline or treatment data trending toward the other Rule 2 Is there a convincing difference Rule 3 Casual experimental design LESSON QUIZ 5 A stable data set has two characteristics First it doesn t have much variability and second it has no TREND The period of behavior observation in which the intervention is not implemented is called the BASELINE James was listening to the radio when the broadcaster said that spankings were good for children He started spanking his son immediately A week later he started spanking his daughter A week after this he started spanking his dog What kind of a single subject design is James using A MULTIPLE BASELINE design BTW spanking is NOT good for children this is a joke In a single subject design you do not AVERAGE the data across participants Tom told his son Bobby to go out and rake leaves Bobby went outside but raked pretty slowly one bag in 2 hours Tom told Bobby that he would pay his 1 per bag In the next two hours Bobby raked 6 bags Tom used which single subject design A comparison design The following is an example of what design A counselor started confronting his client Annie s irrational thoughts following a few weeks of non directive therapy A few weeks later after his confrontation he started using this confrontational therapy mode with Tim another client who had been in non directive therapy for several weeks MULTIPLE BASELINE The MULTIPLE BASELINE design is the only single subject design that can rule out alternative explanations when the removal of treatment is not ethical Dr Baker assessed how content and healthy the residents of the retirement home were for 4 months Then he had students bring cats to the home every day for a period of 6 months During the latter period he continued to survey people to see how content and healthy they were Which single subject design is Dr Baker using A COMPARISON design If you collected these data on the frequency of a problem behavior Baseline 23 38 68 48 83 76Treatment 74 53 36 38 16 3 Would you be interested in conducting a reversal yes or would you conclude that the data do not really warrant a reversal because there is little evidence that the treatment produced the behavior change no YEEESSSSSS When using the behavioral strategy A create a behavioral definition B use a method of direct observation C check the RELIABILITY and social validity of your observations and D design a single subject experiment A school observed how often students used hand sanitizer after they attended the health meeting The data taken from the observations are from the TREATMENT condition The advantage of the reversal design is that it can rule out ALTERNATIVE explanations of the observed behavior change If each participant is exposed to both the baseline and treatment conditions this is a SINGLE SUBJECT design When you compare a single treatment condition with a single no treatment condition you are using a COMPARISON design Outcome recording requires observers to DIRECTLY observe the presence of a physical result Dr Smith showed his behavioral definition of senility to the directors of three area nursing homes They all agreed that the definition was appropriate Dr Smith has now established the SOCIAL VALIDITY of his behavioral definition My father suffers greatly from hair loss It makes him anxious and he is always touching his head which causes even more hair to fall out My mom bought him a bracelet that gave him a little zap every time he touched his head After about 3 weeks he never rubbed