Research Methods -2

Research Designs

• Descriptive methods
• Experimental methods
Descriptive Research

• Involves simply describing the behavior and the

situation within which it occurs.
• This research does not involve the manipulation of
any variables.

➢Naturalistic Observation
➢Case Studies
 Naturalistic Observation
• Naturalistic observation involves the systematic
observation and recording of behavior in its natural
environment.
• Researcher should make their observations in a
consistent and uniform manner.
 Naturalistic Observation
• Jane Goodall’s systematic study
of chimpanzee behavior in the
wild is an example for
naturalistic observation.
• By observing them in the wild,
she discovered that
chimpanzees eat meat, use
simple tools, and engage in
group conflicts. G r şmek
• Her work provided valuable
insights into their natural
behaviors.
• The naturalistic approach is
excellent for gaining rich,
detailed information.
Naturalistic Observation
• Naturalistic observation also
helps us understand learning.
• Scientists gave monkeys sweet
potatoes covered in sand.
• Monkeys liked the potatoes but
disliked the dirt.
Naturalistic Observation

• An 18-month-old female
monkey found a solution
• She washed the potatoes in a
stream
• She shared this new behavior
with her mother and playmates
• Other young monkeys learned
by watching and copying
• By 1958, most monkeys were
washing sweet potatoes
• This showed social learning and
cultural transmission in animals.
Naturalistic Observation

• This approach is commonly used in ethology, a branch of zoology that

focuses on the study of inherited behavior patterns in animals.
• Such patterns have presumably evolved to help the animal cope with
certain aspects of its natural environment.
Naturalistic Observation

Strengths Limitations
• Provides detailed, rich • Difficult to identify key
information. variables.
• Captures behavior in natural • No intervention or clarification
contexts. possible.
• Hard to determine underlying
causes of behavior.
Case Studies
• The case study approach involves the intensive examination
of one or a few individuals.
• These can be done in natural settings or more structured
settings.
Case Studies

• Case studies are particularly common in medical research.

• For example, studying individuals with certain types of neurological
damage helps us understand which areas of the brain control specific
functions.
• This can provide valuable insights into brain function and recovery.
Case Studies

• Broca’s Area (frontal lobe) → responsible for speech production.

• Wernicke’s Area (temporal lobe) → responsible for language
comprehension.
E

• Damage to Broca’s area → causes difficulty in forming words

• Damage to Wernicke’s area → causes difficulty in understanding
language
Case Studies

• Examining the lives of highly gifted

individuals.
• For example, examining the gifted
individual such as Albert Einstein
provide us information how
exceptional skills can be acquired.
Case Studies

Benefits
• Provides deep, detailed information about specific cases.
• Can reveal unique insights into exceptional skills and brain
functions.
Limitations
• The findings from case studies might not always apply to the
general population because they focus on very specific
cases.
• Difficult to specify which variables influence which behavior.
Experimental Research

• To find out why behaviors occur, we need to conduct experiments.

• Experiments are designed to find cause-and-effect relationships.
• This means we manipulate one or more factors (called independent
variables) and see how these changes affect another factor (called
the dependent variable), which is the behavior we’re interested in.
Experimental Research

Behavioral researchers use two main types of expeirmental designs:

• Control Group Designs: These involve comparing a group that
receives a treatment with a group that does not. This helps us see if
the treatment causes changes in behavior across different groups.
• Single-Subject Designs: These focus on the effect of a treatment on
one individual. This design is useful for understanding how an
individual responds to a treatment.
Control Group Designs

• Individuals are randomly assigned to either an

experimental group or a control group
• Individuals assigned to the experimental group are
exposed to a certain manipulation or treatment
• Those assigned to the control group are not
exposed to it
Control Group Designs

• Example: The experimental group gets free food every minute for 30
minutes, while the control group doesn’t get any food but just
explores the chamber.
• The rats in each group receive one session per day for 10 consecutive
days.
• After 10 days, we test them by seeing how quickly they learn to press
a lever to get food.
• On day 11, lever is placed into the chamber and rats have to learn to
press the lever to obtain food.
Control Group Designs

• The question of interest is whether the rats that previously received

free food will learn to press the lever more readily or less readily
than the rats that did not receive free food.
Control Group Designs

• The independent variable is the presence or absence of free

food during the initial phase of the experiment.
• The dependent variable is the time it takes for the rats to
learn to press the lever.
Control Group Designs

• Research has shown that animals that receive free

food subsequently have more difficulty learning
how to respond for food.
Control Group Designs

• Control group designs can get more complex.

• For instance, we could add another variable, like age. We might test
old rats, middle-aged rats, and young rats, creating a 2 × 3 factorial
design.
• This design helps us see if the effect of free food on learning depends
on the rats’ age.

• There are two independent variables (food and age)

 Example Factorial Design

Interaction effects give us a much finer understanding of the variables.

Comparative Design

• A type of control group design using different species as an

independent variable.
• Purpose: Tests evolutionary hypotheses about differences in learning
traits between species.
Comparative Design

• Comparative designs can be simple or factorial, and they can involve

more than one independent or dependent variable.
• The main distinction (other than the use of more than one species)
between comparative designs and standard control group designs is
that in a comparative design you do not have a pure control group
that receives no treatment.
Comparative Design

• Comparative designs can be simple or factorial.

• Example:
• Do rats deal with small, enclosed environments better
than dogs?
• Examine: how quickly dogs and rats learn to find a target
in a complex maze versus a large open area
Comparative Design
Hypothesis: Rats and dogs have evolved differently in their ability to
navigate environments.
• Independent Variables: Species (rat vs. dog) and Environment (maze
vs. open area).
• Dependent Variable: Speed of learning to find a target.

Design: 2 × 2 factorial design with four conditions:

• Rats in a maze
• Rats in an open area
• Dogs in a maze
• Dogs in an open area
Benefits of Control Group Designs
• You can assess the general effects of certain
variables.
• You can determine cause-and-effects.
• You can be fairly confident that differences between
groups in performance are the result of differences
in the independent variable.
• Strict control over the environment rules out
alternative explanations.
• Various characteristics of the subjects in each group
are likely to be evenly distributed.
Limitations of Control Group Designs
• They require a large number of subjects.
• They are not well suited for investigating the effect
of a certain treatment on a particular individual.
• They focus on the average performance of all
subjects in each group.
• Results are often analyzed and interpreted only at
the end of the experiment.
• Species can differ in more ways than just their
learning capacity or style (comparative designs).
mobese
Single-Subject Designs

• Require only one or a few subjects to conduct an

entire experiment.
• The problem of generalizability is typically solved by
running the study with more than one subject.
• If we find the same pattern of results for all of the
subjects submitted to these procedures, the
findings are likely to have good generality.
Single-Subject Designs

• Simple-Comparison Design
• Reversal Design
• Multiple Baseline Design
• Changing-Criterion Design
Simple-Comparison Design (AB Design)
c garette

• Behavior in a baseline condition is compared to

behavior in a treatment condition.
• The baseline is the normal frequency of the
behavior that occurs before some intervention.
• The baseline period is often called the A phase.
• The treatment period is called the B phase.
Example: Self-Punishment for Smoking

• Cory's goal: Reduce smoking.

• Self-punishment: Gives 25 cents to a buddy for
every cigarette smoked.
Example: Self-Punishment for Smoking

Baseline Phase (A phase)

• Measure typical smoking frequency.
• Record data daily (e.g., check marks on a card).
• Duration:as_
Several days is required to ensure accuracy.
Reactivity During Baseline
• Monitoring behavior can cause initial changes (reactivity).
• Extend baseline until behavior stabilizes.
Example: Self-Punishment for Smoking

Treatment Phase (B phase)

• Implement self-punishment procedure.
• Compare smoking frequency to baseline data
• If the treatment is effective, the frequency of smoking
during the treatment period should be consistently lower
than it was during the baseline period.
Cory’s Simple-Comparison Design
Results
Cory’s Simple-Comparison Design
Results
• There is one disadvantage of
this design.
• It does not control for the
possibility that some other
event that occurred at the same
time as the treatment could
have caused the change in the
behavior.maybethere s another causes
that soccurred at the
• In other words, it does not
clearly demonstrate a functional
safe
relationship between the
independent variable (self-
punishment) and the dependent
variable (smoking).
Limitations of AB Design

No Control for Other Events

•Behavior change might result from external factors, not
treatment.
•Example: Cory reduces smoking due to a cold, not self-
punishment.

Functional Relationship
•AB design doesn’t prove the treatment caused the change.
•It only provides suggestive evidence.
When to Use AB Design

as
•Useful when time/resources are limited.
•Appropriate for testing if there’s some improvement.
Reversal Design
• Single-subject design.
• Repeated alternations between baseline (A phase)
and treatment (B phase).
• Depending on the number of reversals carried out, it
is called ABA or ABAB design.
Reversal Design
• If the behavior changes each time the treatment is
instituted and later withdrawn, then a functional
relationship has been demonstrated.
• It is ideal for determining the effectiveness of a
behavioral intervention for one person.
Reversal Design
• Example:
• Cory would repeat the baseline period and then the
treatment period.

A Phase: Measure smoking habits (baseline).

B Phase: Self-punishment for smoking (treatment).
Alternate between baseline and treatment phases.
Cory’s Reversal Design Results
Cory’s Reversal Design Results

• Graphic clearly shows that

self punishment is
effective to decrease the
number of cigerattes
smoked.
• It is extremely unlikely
that some other event,
such as illness, occurred
precisely with each
application of the
treatment to produce such
systematic changes in
behavior.
 Reversal Design

Evidence of Effectiveness

• Smoking decreases during treatment phases (B).

• Smoking increases during baseline phases (A).
• Systematic changes = strong evidence for treatment effect.
l
 7
Why Is Reversal Design Better?

Demonstrates Functional Relationship

• Clear link between treatment and behavior change.
Reduces External Factor Influence
• Unlikely other events (e.g., illness) align with treatment phases
consistently.
• The systematic pattern gives strong evidence that the treatment
works.
Strengths of the Reversal Design

• Unlike the control group design, it allows an entire experiment to be

conducted with a single subject.
• As such, the reversal design is often ideal for determining the
effectiveness of a behavioral intervention for one person.
Generalization of Reversal Design Results

• Can results from a single subject be generalized to others?

• Run the study with more than one subject.
• Replication: Each new subject replicates the experiment.
• If the same results are observed across subjects, the findings
can be generalized.
More Than One Treatment

• A reversal design can be used to assess the effectiveness of more

than one treatment.
• Two treatments design.
Cory’s Results With Two Treatments
 Limitations of Reversal Designs
• The behavior must revert to its original baseline frequency when the
treatment is withdrawn.

If smoking doesn’t return to pre-treatment levels after treatment stops, we can't be

sure the treatment caused the decrease.
Limitations of Reversal Designs

• Long-lasting Effects: Reversal designs are not suitable for treatments

that aim for permanent change.
Limitations of Reversal Designs

• Ethical Concerns: It may be unethical to remove a treatment after

seeing positive results.
Multiple Baseline Design1
kuruldu
A treatment is instituted at successive points in time for two
or more persons, settings, or behaviors.
• Multiple-Baseline-Across-Persons
• Multiple-Baseline-Across-Settings
• Multiple-Baseline-Across-Behaviors
Across Persons

• Treatment applied to individuals at different times.

• Example: Three people trying self-punishment for smoking.
• If improvement in behavior coincides with treatment
implementation, a functional relationship is demonstrated.
Multiple-Baseline-Across-Persons
Eros

• measure the behavior of each person at

baseline
• measure the behavior of each person at
treatment
Across Settings

• Treatment applied to different settings (e.g., work, home, coffee

shop).
• Example: Cory self-punishing smoking at different places over time.
• If smoking reduces only in settings where treatment is implemented,
treatment is likely the cause of change
Multiple-Baseline-Across-Settings

• measure the behavior in each

setting at baseline
• measure the behavior in each
setting at treatment
Across Behaviors

• Treatment applied to different behaviors (e.g., smoking, swearing,

nail biting).
• Example: Treatment applied to each behavior at different times.
• If improvement happens only when treatment is applied, a functional
relationship is demonstrated.
Multiple-Baseline-Across-Behaviors

• measure each behavior at baseline

w̅
• measure each behavior at treatment
Benefits of Multiple Baseline Designs

• An alternative to reversal design

• No need to withdraw treatment to test
effectiveness
• Useful when behavior change is permanent
• There is no ethical concern
Limitations of Multiple Baseline Designs

• Requires more than one person, setting, or

behavior.
• Risk of generalization before treatment is fully
applied
• Hard to confirm if treatment caused improvement
Example

• Cory tries to quit smoking at work

• The change might spread to home and coffee shop before
treatment is applied there
• Makes it difficult to know if treatment caused the
improvement
Changing-Criterion Design

• A type of single-subject design

• Used when behavior change is gradual, not immediate
• Behavior must match a changing criterion set over time
• Helps measure progress in small steps
Cory’s Smoking Reduction Plan

• Baseline phase – records current smoking behavior

• Sets a small reduction goal (e.g., 2 fewer cigarettes)
• If he meets the goal for 3 days, he reduces again
• Uses self-punishment (tearing a dollar bill) for exceeding the limit
• Continues until smoking is eliminated
Strengths & Limitations

Strengths
Good for behaviors needing gradual adjustment
Useful for increasing or decreasing behaviors (e.g., studying
more, gaming less)

Limitations
If behavior does not closely follow the criterion, it’s hard to
tell if treatment works