Single Subject Designs (SSDs)

Comprehensive practice flashcards covering Single Subject Designs (SSDs), including design types (AB, ABA Reversal, Multiple Baseline, Changing Criterion), key characteristics, empirical studies, threats to internal and external validity, visual analysis, and effect size calculation methods.

What three foundational criteria must all research design satisfy?

1. Allow the effect of the IV to be apparent in the DV; 2. Control for extraneous variables and other possible explanations for changes in the DV; 3. Present and analyze the data in a format that makes sense for interpretation.

Why do Single Subject Designs require repeated, reliable measurement of the dependent variable?

Repeated, reliable measurement allows researchers to observe the pattern of the DV over time and control for the normal variation expected within a short time interval.

What is the function of the Baseline (A Phase) in a Single Subject Design?

The baseline phase represents a period of no treatment (or traditional treatment in its natural state) with repeated measures of the DV, serving as a comparison condition to detect extraneous variables and functioning as a quasi-counterfactual.

What guideline is given regarding the length of the Treatment (B Phase) relative to baseline?

The treatment phase length should be approximately at least as long as the baseline phase for comparison purposes.

Why does a simple AB design offer limited control over threats to internal validity?

It provides no control for extraneous variables, meaning changes observed in the dependent variable could be caused by numerous competing factors.

In a reversal (withdrawal) design (such as ABA), what purpose does withdrawing the intervention serve?

The withdrawal of the intervention strengthens the demonstration of a functional experimental relationship between the IV and the DV.

What are the primary problems and limitations associated with reversal (withdrawal) designs?

They do not completely control for extraneous variables, are susceptible to irreversibility (carryover effects, maintenance, or irreversible behaviors), present ethical and legal issues when withdrawing beneficial treatments, and can suffer from sequence effects.

What critical overlap is required in a concurrent multiple baseline design to avoid becoming simply a series of stacked AB designs?

There must be overlap in terms of settings, researchers used, or participants used to detect potential extraneous variables.

What crucial requirement applies to dependent variables in a multiple baseline design?

The dependent variables must be independent of each other so that changing one behavior, setting, or participant does not automatically alter the untreated baselines.

What design configuration is illustrated in this diagram?

A Multiple Baseline Design across outcomes, settings, or subjects, showing staggered baseline phases ($A$, $A'$, $A''$) sequentially transitioning into intervention phases ($B$, $B'$, $B''$).

What are the three types of Multiple Baseline Designs and their distinguishing conditions?

1. Across participants: same settings and behavior, but different participants; 2. Across behaviors: same participant and setting, but different behaviors; 3. Across settings: same participants and behavior, but different settings.

How does a Changing Criterion Design demonstrate experimental control?

By holding the behavior, participant, and setting constant and showing that changes in the dependent variable occur only when each successive criterion level changes.

What three considerations must be addressed when designing phases in a Changing Criterion Design?

1. Length of phase: long enough to establish behavioral stability; 2. Size of incremental change: sufficient variation to demonstrate experimental control; 3. Number of phases: a higher number of phases provides stronger experimental control.

What is the primary disadvantage of a Changing Criterion Design when evaluating experimental control?

If the behavior regularly exceeds the set criterion level, experimental control is not clearly demonstrated.

How can researchers minimize threats of instrumentation and attrition in SSDs?

Instrumentation threats are addressed by waiting for stability of behavior and maintaining similar measures across UTOS. Attrition is minimized by increasing baseline duration until behavior stabilizes, employing more reliable measures, and offering incentives.

In SSD visual analysis, what is the distinction between 'level' and 'trend'?

'Level' refers to the magnitude or value of the DV, whereas 'trend' refers to the direction of the level of the DV over time (increasing, decreasing, stable, or variable).

What is serial dependency in Single Subject Designs, and what statistical error does it introduce?

Serial dependency occurs when successive observations are correlated such that measurement error at one point predicts error at the next. This violates the assumption of independence and inflates Type I error rates.

How was the weighted effect size ($d = 9.59$) calculated across the three multiple baseline sets shown in this figure?

Each set's $d$ value was weighted by its total number of observations in $A_1$ and $A_2$: $\frac{(8.92 \times 18) + (10.11 \times 15) + (9.82 \times 19)}{18 + 15 + 19} = 9.59$.