Method - 3

Major research design: Quasi-experimental (nonequivalent groups design)

resembles an RCT but without random assignment; uses naturally occurring or pre-existing groups (e.g., neighborhoods, school districts, community coalitions)

Quasi-experimental (nonequivalent groups design) doesn’t mean invalid or fake but is used when

random assignment is impractical, unethical, or culturally inappropriate (necessary in real-world and community research)

Quasi-experimental (nonequivalent groups design) has higher risk of

selection bias — groups may differ in important ways before the study begins

Ex of Quasi-experimental (nonequivalent groups design)

a school wants to know whether a new teaching method improves students' test scores.

The researcher uses two existing classes:

• Class A → uses the new teaching method

• Class B → uses the regular teaching method

  • Students were not randomly assigned to the classes (pre-existing)

Quasi-experimental (nonequivalent groups design)’ s Campbell & Stanley Notation

Repeated Measure Design (within-subject design)

A design in which the same participants are measured multiple times over time - can use randomized assignment or not

Major research design: Hybrid Design

combine strengths of multiple frameworks or extend designs over time

Hybrid Design address

internal validity threats

Hybrid design: Solomon Four Group Design

determines whether a pretest itself affects how people perform on the post-test (since taking a pretest can influence behavior) - called Testing Threat

Testing Threat is basically when

an internal validity risk in research where participants change their performance on a post-test simply because they already took a pre-test

Solomon design lets researchers see whether the treatment effect

differs depending on whether participants were pretested or not

Solomon Four Group Design: 2 groups receive a pretest,

2 receive treatment

Ex of Solomon Four Group Design

• Group 1: pretest → intervention → posttest

• Group 2: pretest → no intervention → posttest

• Group 3: no pretest → intervention → posttest

• Group 4: no pretest → no intervention → posttest only

The extra groups aren’t there simply because “more group are better” but

they allow researchers to examine whether a testing threat exists

Solomon Four Group Design’ s Campbell & Stanley Notation

Hybrid design: Switching Replications Design

2 groups switch between treatment and control conditions at different times, allowing researchers to see if the treatment effect is replicated

Switching Replications Design is one of the

strongest design

Switching Replications Design advantages

replication, strong experimental control, and reducing ethical concerns of permanently denying an intervention/treatment

Switching Replications Design’ s Campbell & Stanley Notation

Major research design: Single Subject (N = 1) Designs

Focus on an individual's behavior rather than a group, use repeated measurements to determine intervention effectiveness

Simple Single Subject (N = 1) Designs

a baseline phase (A) followed by an intervention phase (B); data are graphed to see if meaningful change occurs after intervention

Ex of Simple Single Subject Design

Marciano tracks water intake at baseline (A), then uses alarms as an intervention (B); increased intake suggests the intervention works

Reversal Single Subject Design

A design that alternates between baseline (A) and intervention (B) phases to show that the behavior changes when the treatment is turned on and off

Ex of Revesal Single Subject Design

Measuring a student’s out-of-seat behavior:

• Phase A (baseline): count out-of-seat frequency with no intervention

• Phase B (intervention): introduce a sticker for staying seated → behavior drops

• Phase A (return to baseline): remove the sticker → behavior increases again

• Phase B (reintroduce): reapply the sticker → behavior drops again, proving the intervention worked

Multiple Baseline Single Subject Design

A design that introduces an intervention at different time points across different behaviors, settings, or participants (avoiding the need to remove a successful intervention)

Ex of Multiple Baseline Single Subject Design

Across Behavior: Improving a Child’s Classroom Habits

1) Raise hand: collect baseline, then introduce verbal praise only for hand-raising; other behaviors remain on baseline (sharing toys and staying in seat)

2) Sharing toys: once hand-raising improves, introduce verbal praise for sharing toys while staying in seat; remain on baseline

3) Staying in seat: finally, introduce praise for staying in seat

Result: each behavior changes when behavior is applied to it

Multielement Single Subject Design

A design that rapidly alternates between 2 or more different interventions within the same session or on alternating days to compare their relative effectiveness

Ex of Multielement Single Subject Design

A researcher compares Treatment A and Treatment B for reducing a child’s tantrums by alternating between the two treatments across different sessions to see which works better

Changing criterion Single Subject Design

A design in which a baseline is followed by a series of treatment phases that get progressively stricter (gradual behavior change toward a larger goal)

• Behavior is shaped gradually towards progressively changing goals

Ex of Changing criterion Single Subject Design

Help client reduce daily cigarette smoking

• Baseline: client smokes an avg of 16 cigs a day

• Phase 1: reinforcer is provided only if client smokes 13 or fewer cigs

• Phase 2: once stable, the goal shifts to 10 or fewer cigs

• Phase 3: the goal shifts down to 7 or fewer cigs

Best research design depends on

many factors (ex: research questions, setting, ethical constraints)