Chapter 14: Single Subject Research Designs

Question 1

Single Subject Design

Answer 1

experimental research designs that can be used with only ONE PARTICIPANT in the entire research study.

Question 2

Can single subject designs establish cause-effects?

Answer 2

yes. variables within the subject must be controlled and manipulated.

Question 3

T/F: single subject designs are experimental research designs

Answer 3

true

Question 4

goal of single subject design

Answer 4

to establish a cause and effect relationship between the two variables, just like any other experimental design.

Question 5

how do single subject experimental designs include aspects of non experimental case studies and time series designs?

Answer 5

like a case study: single subject focuses on one participant and provides detailed notes

like a time series quasi experimental within subject design: single subject typically involves a series of observations made over time, usually before and after treatment.

Question 6

results from a single subject design needs to be interpreted via ____

Answer 6

visual inspection

Question 7

T/F: you can conduct hypothesis tests and assess variability in single subject design, just like any other experimental design

Answer 7

false. there is only one score because there is only one participant. you cannot assess variability of one score.

Question 8

it is essential that the obtained data is _______ so that an observer can see the treatment effect by ____

Answer 8

it is essential that the obtained data is UNQUESTIONABLY CLEAR so that an observer can see the treatment effect by VISUAL INSPECTION

Question 9

What is a phase

Answer 9

a series of observations made under the same conditions.

Question 10

two “main” types of phases

Answer 10

1) (A) baseline phase: observations made in the absense of a treatment.
2) (B) treatment phase. Observations made during a treatment.

B can turn into C or D or E depending on the number of treatment conditions you expose the participant to.

Question 11

Explain what is happening here:

A-B-B1-A-BC–C

Answer 11

experiment made baseline phase observations (A), a treatment observations (B), a modification of the first treatment (B1) (maybe a higher dose?), back to baseline with no treatment (A), treatment 1 and treatment 2 together (BC) (two types of drugs maybe)? and then just treatment C.

Question 12

Purpose of phase

Answer 12

to establish a clear picture about what is going on under the specific conditions of the phase. the goal of phases is to show how pattern of behaviour changes from one phase to the next.

Question 13

2 methods in defining the pattern of a phase

Answer 13

1) by its level: how constant are the markers of the pahse?2) trend: consistent increase or decrease in behavior across a series of a particular phase

Question 14

three methods of dealing with unstable data within a phase.

Answer 14

remember: you cannot administer treatment if the baseline phase data is unstable because you cannot establish any sort of pattern within the phase. Methods of dealing with unstable data:
1) wait: hope that the data wills stabilize and reveal a clar pattern
2) average out 2 or more observations to reduce the variability between data points within the same phase
3) look for patterns of inconsistency. On MWF’s does the pattern increase? may need to measure only on Tues and Thurs.

Question 15

how many observations should be made per phase minimum?

Answer 15

at least 3 observations per phase. usually at least 6 are used though.

Question 16

when would the minimum observations made per phase be violated?

Answer 16

1) if the condition is dangerous for the participant and you need to start treatment immediately

ex/ you found a treatment for homocidal tendencies. you wouldn’t wait for a psychopath to kill three more people (make three observations) before giving the murderer the homocidal tendency diminisher medication

2) if the baseline trend is showing a trend of improvement without medication
- if you give medication while the person is already improving, you cannot tell if any further improvement is due to the medication or due to natural causes.

3) may need to withdraw treatment phase if severe deterioration in behavior results.

ex/ you started the treatment phase of a new drug. the drug caused them to have a seizure everytime they coughed. you cannot wait for 3-6 observations because it will compromise the person’s health. you need to return to baseline phase.

Question 17

when should you make more than three minimum observations before continuing to the next phase?

Answer 17

when baseline data has high variability. Get more observations to try and establish a clear pattern/level/trend/stability

Question 18

problem with visual inspection technique

Answer 18

subjectivity. Person may look at data to see trends they want

Question 19

general things to look for when visually analyzing data

Answer 19

1) a change in average level. large differences in means between the two phases may indicate a difference between baseline and treatment conditions
2) immediate change in level. comparing the LAST DATA POINT in one phase to the FIRST DATA POINT in the second phase.
3) change in trend. maybe the first phase goes up but the second phase immediately makes the trend go down.

4) any latency in change?
maybe there is a difference in data after an extended period of time on the second phase compared to the first phase.

Question 20

___ change is more convincing than ____ change.

Answer 20

IMMEDIATE change is more convincing than SLOW-ONSET change.

Question 21

goal of ABAB reversal design

Answer 21

to demonstrate that the treatment causes change in behaviour by showing that pattern in each treatment phase is different from pattern in the baseline phases.

Question 22

Practical and ethical issues in ABAB design

Answer 22

1) practical issues
- participant behavior may not return to baseline after treatment is removed and you thus cannot test another round of “AB.”
- good from a clinical perspective. the person is cured
- bad from an experimental perspective. you cannot replicate the results on the same person

2) ethical issues
- you cannot withdraw a treatment that is working for someone in order to make them go back to baseline phase.

Question 23

problem with A-B-B1-C variation of ABAB design? What should you do instead? assume that C at the end of the experiment caused a change.

Answer 23

the problem with A-B-B1-C variation is that if C caused a change, you don’t know whether this is:

1) coincidence. perhaps extraneous variables may have caused a change in C
2) change in behavior may be a slow-onset effect of B or B1.
3) treatment C was truly effective.

You can see if treatment C was actually effective by bringing the subject back to baseline after the first administration of C:
A-B-B1-C- A-C.

• if C after baseline exhibits the same effect as the first C after B1, you know C is causing a change and that is it not a delayed effect of B.

Question 24

what is a multiple baseline design? when is this good to use?

Answer 24

multiple baseline design requires only ONE phase change and ESTABLISHES CREDIBILITY by replicating the phase change for a second participant or second behavior. this type of design is good to use when treatments have lasting or permanent effects in which the original subject cannot be brought back to baseline.

Question 25

Multiple baseline across subjects

Answer 25

2 or more participants

Question 26

multiple baseline across behaviors

Answer 26

2 or more behaviours correspond to a single participant

Question 27

multiple baseline across situations

Answer 27

when the same behaviour correspond to 2 or more situations of the SAME participant.

Question 28

baseline phase begins for 2 participants ____ but ____ for one participant after the treatment phase has been started for the other participant.

Answer 28

baseline phase begins for 2 participants SIMULTANEOUSLY but CONTINUES for one participant after the treatment phase has been started for the other participant.

Question 29

criteria for causal conclusion from a multiple-baseline design

Answer 29

1) shows a clear and immediate change in pattern of behaviour when switching from baseline to treatment phase
2) includes at least 2 demonstrations that behaviour changes when the treatment is introduces.

Question 30

pros of multiple baseline design

Answer 30

• eliminates the need for reversal or return to baseline in the same patient. good for ethical concerns if the treatment is working

Question 31

cons

Answer 31

• if you are using a multiple baseline design across behaviours, it may be difficult to identify similar but independent behaviours within the same participant that you can test the drug on.
• treatment applied to one behavior may generalize and produce changes in a second behavior
• there is a conflict between clinical and experimental goals
• clinical: you want one treatment to improve a variety of behaviors
  research: if treatment effects both behaviors, the credibility of the treatment effect is undermined.
• different participants may respond differently to treatments in a multiple baseline design across subjects.

Question 32

pros of a single subject design in general

Answer 32

1) can establish causal effect with one person
2) can be tied with clinical practices
3) is flexible. can try multiple treatments without compromising integrity: you just need to return it to baseline first.
4) good practical significance, but not a lot of statistical significance.

Question 33

cons of a single subject design in general

Answer 33

1) lack of generalizability
- treatment may be different for other individuals.
2) subjective measurement
- hard to run statistics because you do not get group data.
3) time consuming
4) measurement reactivity/obtrusive.
- participant reactivity; single subject may say something to get the researcher out of their face
- threat to internal validity.