The Research Enterprise in

Psychology (Chapter 2)
Ms. Umul Baneen Rizwan
Lecturer
Department of Psychology, FCCU
PSYC 100
LOOKING FOR LAWS: THE
SCIENTIFIC APPROACH
 Science in Psychology: Big Picture
• Psychology assumes behavior follows discoverable laws
and patterns.
• Main goals of science:
– Measurement & description – define/quantify what we
study
– Understanding & prediction – explain and forecast
relationships
– Application & control – use knowledge to solve real
problems
• Key tool: empirical (systematic) observation, not just
common sense.
FROM IDEAS TO DATA
 Theories, Hypotheses, and
Variables
• Theory: organized system that explains observations
and links facts.
• Hypothesis:
– A precise, testable prediction about relationships between
variables
– Derived from a theory; tested one slice at a time
• Variables:
– Any measurable conditions, events, characteristics, or
behaviors
– Must be defined clearly (operational definitions) so others
can replicate
STEPS IN A SCIENTIFIC
INVESTIGATION
 The 5-Step Flow
• 1) Formulate a testable hypothesis
– Make it specific and measurable (requires operational definitions)
• 2) Select a research method & design the study
– Choose among experiments, surveys, naturalistic observation, case
studies, etc.
– Plan sampling, materials, procedures
• 3) Collect the data
– Use systematic procedures: observation, questionnaires, interviews,
tests, physiological measures, archival records
• 4) Analyze data & draw conclusions
– Compute statistics to evaluate the hypothesis
• 5) Report findings
– Publish/present so others can critique, replicate, and extend
HOW PSYCHOLOGISTS COLLECT
DATA
Common Data Collection Methods
• Direct observation
– Watch and record behavior as it occurs
• Questionnaires & Interviews
– Ask people directly about thoughts, feelings, or behaviors
• Psychological tests
– Standardized measures of personality, intelligence, etc.
• Physiological recordings
– Heart rate, hormone levels, brain activity, etc.
• Archival records
– Use existing data sets (records, databases)
LOOKING FOR CAUSES:
EXPERIMENTAL RESEARCH
 Core Logic of an Experiment
• Purpose:
– Show that changes in X (independent variable) cause changes in Y
(dependent variable)
• Key pieces:
– Independent Variable (IV): manipulated factor
– Dependent Variable (DV): measured outcome
– Experimental group vs. Control group: identical in every way
except the IV
• Internal validity essentials:
– Random assignment to conditions
– Hold constant or balance out extraneous variables
– Avoid confounds so IV is the only plausible cause
 Classic Illustration: Does Misery
Love Company?
• Schachter (1959):
– High-anxiety group vs. low-anxiety group
(manipulated expectations)
– Measured desire to wait alone or with others
(need for affiliation)
– Result pattern supported hypothesis: anxiety
increased desire for company
• Takeaway: with control and manipulation,
experiments can reveal cause-effect links.
 Variations in Experimental Design
• Between-subjects vs. Within-subjects:
– Between: different people in each condition
– Within: same people experience multiple conditions
(reduces noise, but order effects)
• Factorial designs:
– Two or more IVs to test combined effects and
interactions
• Multiple DVs:
– Broader picture of effects (e.g., behavior + self-report
+ physiology)
 Pros & Cons of Experiments
• Advantages:
– Strongest method for inferring causality
– Tight control reduces alternative explanations
• Disadvantages:
– Some questions can’t be manipulated ethically or
practically
– Artificial settings may limit generalization (external
validity)
LOOKING FOR LINKS:
DESCRIPTIVE/CORRELATIONAL
RESEARCH
 What These Methods Do (and
Don’t) Do
• Purpose:
– Describe behavior and discover associations
between variables
• Limit:
– No manipulation → cannot demonstrate cause-
and-effect
 Naturalistic Observation
• Definition:
– Observe behavior without intervening, in natural
settings
• Strengths:
– Rich, real-world detail; useful when little is known
• Challenges:
– Reactivity (subjects change because they’re
observed)
– Hard to convert notes to precise numbers
 Case Studies
• Definition:
– In-depth investigation of a single subject or small
group
• Strengths:
– Detailed insight; good for rare conditions or new
areas
• Challenges:
– May not generalize; vulnerable to bias; can’t show
causation
 Surveys
• Definition:
– Structured questionnaires or interviews to assess
attitudes/behaviors
• Strengths:
– Efficient for large samples; can reveal patterns and
links
• Challenges:
– Sampling quality matters; relies on accurate self-
report
 Correlation ≠ Causation
• Correlation coefficient (r):
– Shows direction (±) and strength (0 to 1) of
association
– Large |r| means strong link, but cannot confirm X
caused Y
• Possible explanations: X→Y, Y→X, third
variable Z → both.
LOOKING FOR CONCLUSIONS:
STATISTICS
 Descriptive Statistics –
Summarizing Data
• Central tendency:
– Mean (average) – sensitive to extremes
– Median (middle) – robust to outliers
– Mode (most frequent)
• Variability:
– Standard deviation: typical distance from the mean
– Higher SD = more spread
• Correlation:
– Numeric index of association between variables
 Inferential Statistics – Deciding
What’s Real
• Goal:
– Use sample data to make judgments about a
larger population
• Common ideas:
– Statistical significance and probability
– Confidence that effects aren’t due to chance alone
• Interpret with caution: size of effect and
sample matter.
LOOKING FOR FLAWS: EVALUATING
RESEARCH
 Threats to Good Evidence
• Sampling bias:
– Sample doesn’t represent population → limits generalization
• Placebo effects:
– Expectations produce change; use placebo controls when relevant
• Distortions in self-report:
– Social desirability, memory errors, response sets
• Experimenter bias:
– Researchers’ expectations subtly influence outcomes
• Solution example:
– Double-blind procedure: neither participants nor key researchers know who’s in
which group
• Replication is the repetition of a study to see whether the earlier results are duplicated.
• Meta-analysis combines the statistical results of many studies of the same question, yield-
ing an estimate of the size and consistency of a variable’s effects.
• A sample is the collection of subjects selected for observation in an empirical study. In
contrast, the population is the much larger collection of ani- mals or people (from which the
sample is drawn) that researchers want to generalize about
LOOKING AT ETHICS: DO THE ENDS
JUSTIFY THE MEANS?
Deception with Human Participants
• Why used:
– Sometimes necessary to study natural behavior
and avoid demand characteristics
• Concerns:
– Stress, mistrust, or harm to dignity if poorly
handled
• Safeguards:
– Research ethics review, informed consent as
appropriate, debriefing to explain deception
 Animal Research
• Debate:
– Critics argue cruelty and questionable
generalizability to humans
– Supporters note benefits to science/medicine and
strict welfare standards
• Core principles:
– Justify scientific value, minimize harm, provide
humane care
 General Ethical Principles
• Respect for persons: informed consent,
voluntary participation, right to withdraw
• Beneficence: minimize risk, maximize benefits
• Justice: fair selection and treatment of
participants
• Integrity and responsibility in all phases of
research
PERSONAL & CRITICAL THINKING
APPLICATIONS
 Finding and Reading Journal
Articles
• What journals are:
– Peer-reviewed, technical reports that advance
knowledge
• How to find:
– Library databases, reference lists, keywords
• How to read:
– Skim abstract → methods → results/figures →
discussion; evaluate quality and relevance
 Beware Anecdotal Evidence
• Single stories (“I know someone who…”) can
be vivid but misleading
• Prefer systematic data and representative
samples over personal testimonies
KEY TAKEAWAYS & EXAM TIPS
 What to Remember
• Science = systematic observation + precise definition
+ transparent reporting
• Experiments test causation;
descriptive/correlational reveal patterns/links
• Statistics summarize (descriptive) and test
(inferential) evidence
• Always ask: Could something else explain these
results? How was bias controlled?
• Ethics matter: protect participants and animals; be
honest and transparent