Research Methods – Naturalistic Observation & Case Studies
Overview of the Scientific Approach
• “Scientific” = using a structured, methodological, evidence-based way to understand the world.
• Five broad research strategies were previewed in the lecture:
– Naturalistic Observation
– Case Study
– Survey
– Correlation
– Experiment
• Today’s discussion focused on the first two.
Naturalistic Observation
• Definition: Systematically watching people, animals, or events in their natural habitat with no interference or manipulation by the researcher.
• Core idea = “natural” + “observe.”
– Ex: Bird-watching in a park.
– Ex: A psychologist watching children on a playground.
– Ex: Sitting in a classroom and noting a professor’s teaching behavior.
Example Demonstration – “Who Runs the Red Light?”
• Thought experiment: Park a car at the corner of Hawaii Ave. × Ocean Pkwy. (8 a.m. – 9 a.m.) and record every driver who runs a red light.
– Code drivers by sex (male / female) and age (older / younger).
– Hypothetical finding: “Twice as many men as women ran the light.”
Why This Single Observation Is Weak Science
• Limited sampling: Only one corner, one hour, one day.
• Confounds:
– Weather (rain or snow could affect risk-taking).
– Police presence (a cop car could suppress violations).
– Traffic composition (if as many male drivers passed as female, the raw count is misleading).
• Over-interpretation: Concluding “men are more likely to take risks” from one behavior at one site assumes causality the method cannot provide.
How to Strengthen a Naturalistic Study
• Increase breadth of observation: multiple intersections, multiple cities, all seasons, varied times.
• Deploy independent observers:
– Researcher A at Wyand × Ocean;
– Researcher B at J St. × Ocean;
– Researcher C at Wyand × Flatbush;
– Additional observers in Queens (Jewel Ave.), Manhattan (34th St.), New Jersey (Cedar Ln.), Long Island (Peninsula Blvd.).
• Goal = larger, more representative data set ⇒ more accurate descriptive conclusions.
• Still descriptive only: even perfect naturalistic data cannot establish cause-and-effect because the researcher never manipulates anything.
Strengths & Limitations of Naturalistic Observation
• Strengths
– High ecological validity (behavior is real-world, not lab-induced).
– Ethical minimalism: no interference means lower risk (though privacy concerns remain).
• Limitations
– No control → many confounds, no causality.
– Observer bias & selective perception.
– Practical challenges: time-intensive, need multiple observers, potential for being noticed.
Case Study
• Definition: Intensive, in-depth examination of one person (or small group) across time and contexts.
– Participants know they are being studied → not stalking.
– Purpose = develop a comprehensive portrait of behavior, cognition, and context.
Classroom Illustration
• Researcher shadows Jeremy from wake-up to bedtime for an entire semester.
• Simultaneously, the researcher watches Anderson only during class (9–10:35, four days/week).
• Result: Far richer understanding of Jeremy—daily routines, social interactions, mood changes—because data are gathered across diverse environments.
Why Use a Case Study When Humans Exist?
• Allows exploration of unique, rare, or extreme phenomena (e.g., brain injury patients, prodigies).
• Generates detailed hypotheses that later can be tested with larger samples.
• Offers qualitative depth unattainable in broad surveys.
Pros & Cons
• Pros
– Depth of information (behaviors, thoughts, feelings, context).
– Useful for theory building and documenting rare cases.
• Cons
– Low generalizability (cannot assume findings apply to the population).
– Researcher-participant relationship may influence behavior (reactivity).
– Time and resource intensive.
Ethical & Practical Footnotes Mentioned in Class
• Naturalistic work should minimize privacy invasion; consent may be required in semi-private spaces.
• Case studies demand informed consent and respect for participant boundaries.
• Classroom incident: A student accidentally screen-shared a cooking recipe—reminder of inadvertent data leaks in online settings.
Preview of Remaining Methods (Survey, Correlation, Experiment)
• Survey = self-report questionnaires/interviews to assess attitudes or behaviors.
• Correlation = statistical measure () quantifying relationship strength/direction between variables.
• Experiment = controlled manipulation of an independent variable (IV) to observe causal effects on a dependent variable (DV), employing random assignment and controls.
The lecture stopped before diving deeply into these three methods; they will be elaborated in future sessions.