Topic 2 Multiple baseline design

Multiple baseline design (MBL)

alternative to the withdrawal design used in behavioral research. Family of research designs that consist of 3 main iterations

MBL Across participants

across bxs

across setteings

MBL Across participants

• most common of the 3 MBL

• employed when there is more than 1 client with similar goal

• intersubject replication

• limitation: carry over effect (interdependence of baselines) where change of one’s partcipant’’s bx might influence another partcipants’s bx even before the intervention is directly applied to them

MBL across behaviors

• less common of the 3 MBL

• used when a researcher is targeting more than one bx for a single client using the same treatment

• Intrasubject replication

• limitation: likelihood for behavioral covariation. This occurs when the targeted bx are functionally similar and belong to the same response class.

MBL across settings

• used when you are targeting the same bx for a client across different settings or situations

• common for bx to be targeted in multiple environments

• intrasubject replication

• limitation: carry over effect (interdependence of baselines) where change of one’s partcipant’’s bx might influence another partcipants’s bx even before the intervention is directly applied to them

staggered baselines

crucial for demonstrating functional relationship in MBLs because the tx effect is replicated across 2 or more baselines at different times, it provides evidence that the intervention, rather than a extraneous factor, is responsible for the observed changes in bx.

The more panels or replications with staggered baselines that show a similar effect in the introduction of the intervention, the stronger the demonstration of a functional relationship.

within-series comparisons

involves looking at the difference in DV (behavior) within a single baseline when a phase change occurs, specifically when the intervention (IV) is introduced

between-series comparisons

involves looking at the difference between a phase change in one series and the maintence of the baseline in another series

verification window

refers to the period during which data is collected to assess whether the observed changes in behavior are consistent and indicative of the treatment's effectiveness, ensuring that the intervention's impact can be confirmed.

found in MBL across Bxs

concurrent vs non current

• Concurrent MBL

  • Behavior measured at the same point in time across participants

  • More control over threats to internal validity

  • Baselines start at the same time and intervention phases are staggered across tiers

• Nonconcurrent MBL

  • Data does not start at the same time as other participants

  • Less control over threats to internal validity 

  • Graphed similar to concurrent, but has a unique x-axis

• Graph distinction made in publication & graph

behavioral covariation

refers to the situation where the Bxs that are functionally similar (belong to the same response class) undergo treatment and as a result, a change in one bx occurs when another bx is being treated

seen in MBL across BXs

Response class

refers to the Bxs that are functional similar meaning that these Bxs serve the same purpose or produce the same outcome for an individual

found in MBL across BXs

Multiple probe design

• uses quick probes for baseline asseement rather than continuous ,repeated measurement

• probe is defines as a single, discrete measurement of the target bx

• used by applied scientists

• can be used for assess maintence and generalization of the treatment effects

  adaptation with intermittent probes instead of continuous baseline measurement, useful for reactivity and infrequent behaviors

Probe

defined as a single, discrete measurement of the target bx

social validity

information, usually obtained from the client or the client’s family, friends, or coworkers, the serves to demonstrate the effectiveness of the the treatment. It is obtained usually through gathered questionnaires.

Interdependence of the BLs

aka carryover effect

limitation of MBLs.

Occurs when a change in one of the baselines carries over another baseline even though the intervention has not been applied to that other baseline.

Inconsistent effects of TXs

limitation to the MBLs

situations where the intervention alters some BXs or baselines when introduced, but not others.

prolonged baselines

limitation in MBLS

MBLs designs sometimes requires lengthy or prolonged periods of baseline data collection before the intervention is introduced across the different baselines

• may not be feasible or ethical to withhold Tx for extended period, especially if target Bx is harmful or there is no clear need for intervention

• may result in performance improving before the Tx has been applied due to factors other than the intervention, making it harder to isolate the effect of the treatment

• undesirable Bxs may emerge during a length baseline period which could complicate the intervention process

Benefits

MBLs are a powerful alternative to withdrawal designs

They can demonstrate a functional relationship3 . This is because the treatment (Independent Variable - IV) is replicated across two or more baselines

The more panels/replications, the stronger the demonstration of a functional relationship

MBLs can be used in a "Best Tx" or "Tx Validation" phase

MBLs are considered beneficial to practitioners

Unlike withdrawal designs, MBLs do not require the removal of a seemingly effective treatment (this is implied by them being an alternative to withdrawal designs

The multiple-probe design, an adaptation of the MBL, offers additional benefit

It is helpful for reactivity

It is helpful when the behavior doesn’t occur in the baseline (BL)

Treatment (Tx) may be implemented with less delay

Repeated assessment (testing effect) is less likely

Limitations

potential for carryover effects, also known as the interdependence of baselines (BLs)8 . This occurs when a change in one baseline carries over to another BL even though the intervention has not yet been applied

This carryover effect is described by the authors as a "social modeling effect" for MBL across participants, also known as the diffusion of treatment. While this might be clinically desirable, it is scientifically ambiguous9 .

MBLs can suffer from inconsistent effects of the treatment (tx), where some behaviors are altered when the intervention is introduced and others are not

Establishing lengthy/prolonged baselines can be a limitation

Prolonged baselines may not be feasible or ethical in some situations

They may result in performance improving before treatment has been applied

Undesirable behaviors may emerge during prolonged baselines

Carryover effects make MBLs a limitation for researchers

When using MBL across behaviors, it is likely to see behavioral covariation10 . This means that when behaviors that are functionally similar (belong to the same response class) undergo treatment, one behavior might change even when another behavior is treated