Single-subject designs (also known as single-case designs) give an alternative to group
The focus is on N = 1, a single subject. However, this can be referred to as
„small-N design‟ and be used for 1-9 subjects.
Useful for research on interventions in behaviour analysis and clinical practice
Small-N design: in-depth study of a single or relatively few subjects under tightly
controlled experimental conditions in which the independent variable(s) is repeatedly
manipulated over successive trials or conditions and in which the dependent variable(s)
is repeatedly measured
Small-N and single-subject designs usually have four components:
1. Repeated measurement of the dependent variable
2. Baseline phase
3. Treatment phase(s) with all subjects exposed to each phase
4. Graphic display, perhaps supplemented by statistical analysis
In ideal situation, measurements are taken before intervention then continuing through
Sometimes, measurements cannot be taken before the intervention. You may be able to
collect data from client records or by asking about previous experiences.
Client records only have information that is available and even that may have
Behaviours and actions are easier to remember than moods or feelings
Recollections should be limited for the previous month or else they may become
Def.: period in which the intervention to be evaluated is not offered to the client
This is abbreviated by the letter A
During baseline phase, repeated measurements of the dependent variable are
taken or reconstructed.
These measures reflect the status of the client on the dependent variable before
Repeated baseline measures allow for some of the threats to internal validity
In baseline stage, measurements are taken until a pattern emerges Stable line: little variability in the scores. This score is desirable because changes can
be easily detected and there will be little problems in testing, instrumentation, statistical
regression, and maturation in the data.
Trend: when the scores may be either increasing or decreasing during the baseline
Curvlinear: rate of change is accelerating over time with regular increases and
A pattern is good when you can accurately predict the next data point
When measurement during baseline phase is taken from existing data or recollection,
the threat to internal validity is greater.
Repeated measures during the baseline phase help rule out threats to validity
Validity threats should appear in the baseline
Will not control for an extraneous event (history) that occurs between the last
baseline measurement and the first intervention measurement
When baseline is a stable line, threats may be ruled out. This is more difficult when the
baseline is a trend, especially if it is moving in the desired direction.
Def.: Represents the time period during which the intervention is implemented
Signified by letter B
The patterns of the data from the treatment phase are compared with the data of the
baseline phase to see if a change has occurred.
Y-axis represents the scores of the dependent variable
X-axis represents a unit of time
Measuring Targets of Intervention
The dependent variable in a single-subject design is the concern or issue that is the
focus of the intervention. The target for change may be one specific problem or different
aspects of the problem.
Once the target of the intervention has been identified, the method of operalization for
the outcome must be decided. Operalization occurs before the beginning of the study.
Measures of behaviours, status, or functioning are often characterized in four ways: Frequency: counting the number of times a behaviour occurs or the number of
times people experience different feelings within a particular time period
Duration: how long behavior lasts
Interval: time between episodes
Magnitude: intensity of behavioral event
When measuring behaviour, researchers must make sure that it is not too difficult or
time consuming to measure it.
As well, they must be careful of reactivity. The process of measurement might change a
The choice of measurements must also be sensitive enough to detect changes.
Too global, it may be impossible to detect incremental/small changes
Whatever is measured must occur often enough that measuring it will have any
Analyzing Small-N and Single-Subject Designs
Visual examination of the graph
Assessing practical (clinical) significance is of primary importance
Practical (or clinical significance): has the intervention made any meaningful difference
in the well-being of the subject? There are several principles for determining this:
Setting criteria: establishing with the client or community the criteria for success.
If the intervention reaches that point, it is meaningful
Cut-off scores: whether the intervention has reduced the problem the problem to
a level below a clinical cut-off score.
Costs and benefits
Def.: process of looking at a graph of the data points to determine whether the
intervention has altered the subject‟s preintervention pattern of scores.
Three concepts to guide this:
Level: magnitude of the target variable; typically used when the observations fall
along relatively stable lines
Trend: direction in the pattern of the data points
Variability: how different or divergent the scores are within a baseline or
intervention phase o Widely different scores in the baseline and intervention stages make
assessment more difficult.
o One way to deal with this is to draw range lines.
Types of Small-N and Single-Subject Designs
This is the basic single-subject design. Includes a baseline phase with repeated
measurements and an intervention phase continuing the same meas