MODULE 2: PSYCHOLOGY AS A SCIENCE

SCIENTIFIC METHOD, RESEARCH DESIGN, BASIC STATISTICS, & RESEARCH ETHICS

OUTLINE

 What is science?
 Research methods
 Basic statistics
 Research ethics
 WHAT IS SCIENCE?
 Method of learning about reality and the world around us through systematic
observation and experimentation.

 Conclusions must be
 Based on facts
 Not personal bias or gain

 Critical thinking
 Rationally
 Independently
 Clearly

 Science will not magically solve all human problems, but is considered the
THINKING LIKE A SCIENTIST

Scientific thinking vs. everyday reasoning

 Objectivity vs. Subjectivity
 Systematic observation
 Repeatable evidence
PROCESS
 Theory OF SCIENCE
 Comprehensive framework based on existing studies
 Analysis of a set of facts in relation to one another
 Creates new predictions and helps make sense of evidence regarding a
phenomena
 Open to revision

 Hypothesis
 A tentative assumption made in order to draw out and test its logical or empirical
conclusions
 “If A happens, then B will be the result”
 Must be falsifiable and testable

 Evaluate your hypothesis

 Data collection
 Data analysis
 Can NOT prove if a hypothesis is true
 THE SCIENTIFIC
METHOD
Scientific Process
Set of rules for gathering
and analyzing information
that then allows you to test
your hypothesis.

The purpose of
psychological research is to
test ideas about behavior.
https://www.youtube.com/watch?v=_7sSuhQ1_24
 REPLICATION
 Repeating an experiment to compare results from
the first study.
 Helps determine degree to which findings can
generalize across situations and time.
 Direct Replication vs. Conceptual Replication
 Replication Crisis
 Currently going on in many scientific fields,
psychology included
 Some prominent studies and concepts cannot
be replicated by other researchers
 Power poses (and now in mainstream media);
https://www.youtube.com/watch?v=C4ACeoq
EjeA
RESEARCH METHODS
SELECTING PARTICIPANTS FOR A RESEARCH STUDY
 Subjects of psychological research = “Participants”
 Sample vs. Population
 Random sample
 Subset of population
 Every member of population has equal chance of being
selected
 Convenience
 Participants are taken from a certain population then
randomized into conditions
DESCRIPTIVE METHODS
Case Study
 Intensive examination of a particular individual being
studied or treated resulting in a detailed description
 Large samples are not possible
 Participant has a very unique quality
 HM: amnesia patient
 Participant has experienced something so abnormal that its
unrealistic to find others who have experienced it
 Genie: https://www.youtube.com/watch?v=gvSMgi23F3o
 Pro: Extensive information
 Con: Limited generalizability
DESCRIPTIVE METHODS (CONT.)
Observation
 Researcher carefully and systematically observes
and records behavior without interfering with the
behavior
 Example: https://www.youtube.com/watch?v=aAY2kBGVwSc

 Naturalistic Observation
 Observe behavior unobtrusively in natural context
 Pro: ecological validity
 Con: difficult to control environmental aspects

 Laboratory Observation
 Observations while people are engaging in a specific task
 Pro: structured with more situational control
 Con: less ecological validity

 Observer bias
 Skewed observations based on knowledge of research goals
DESCRIPTIVE METHODS (CONT.)

Natural Observation Laboratory Observation

 More realistic of how the  The control of the lab could bias
subject interacts the subject
 More pure data  May not act as though they would in
natural setting
 Subject may experience
 Controlled, so no outside
distraction or change in the
influences
situation that may detriment
 Researchers can easily measure
the observation
their variables
 Therefore the subject may stop
displaying or overreact, which
would make the observation
DESCRIPTIVE METHODS (CONT.)
Narrative Analysis
 Study of stories and personal accounts of a particular group,
culture, or person
 Researchers look at themes in their story, along with the
structure of their narrative
 Stories could be audio-recording, video-recording, or written.
 Examining what the person says in their narrative, as well as
how they said it
SURVEY RESEARCH
Surveys
 Questionnaires/interviews that ask people directly about their
experiences, attitudes, or opinions.

 Self-report
 Social acceptability: individuals want to make themselves look
good and accepted by the researchers/society.

 Operational definitions
 What are you trying to measure?
 Example: Love

 Advantages vs. Disadvantages

CORRELATIONAL RESEARCH
 A descriptive study that looks for
patterns of relationship between
two phenomena (variables)

 No manipulation by the researcher

  Correlation does NOT equal
causation
 Variables may influence
simultaneously
 Third variable problem: any
outside variable that may be
affecting variables 1 and 2
CORRELATIONS (CONT.)
Direction and strength of the relationship between two
variables.
 Correlation coefficient
 Statistical summary of the relationship between two variables
 Indicates strength of relationship and direction of relationship

 Positive correlation
 Variables move in the same direction

 Negative correlation
 Variables move in the opposite directions

 No correlation
CORRELATION EXAMPLES
CORRELATIONS: CORRELATION COEFFICIENT
CORRELATIONS: THIRD VARIABLE PROBLEM
EXAMPLE
CORRELATIONS (CONT.)

Why use correlations?

 When the researcher cannot randomly manipulate certain
variables of subjects
 Certain populations
 Example: Pregnant women
 Certain situations
 Example: Those in prison
EXPERIMENTAL RESEARCH
Experiment
 A controlled test of a hypothesis in which the researcher
manipulates one variable to discover its effect on
another

 Independent Variable (IV)

 The variable being manipulated/changed by the
experimenter
 “If A happens…”

 Dependent Variable (DV)

 The outcome variable—impact of change from IV
 “…B will…”
EXPERIMENTAL RESEARCH (CONT.)
 Experimental Condition
 Presented the independent variable (manipulation)
 Intervention or variable being tested

 Control Condition
 A condition that would be natural to the situation (no
manipulation)
 Special type of control group: Placebo

 Random assignment
 Participants have an equal chance of being in the experimental or
the control condition
 Key to cause-and-effect conclusions
EXPERIMENTAL RESEARCH (CONT.)
EXPERIMENTAL RESEARCH (CONT.)

Confounds
 Individual differences
 Random assignment usually controls this

 Situational Confounds
 Minimize by running experiments under constant circumstances

 Participant demand

 Experimenter bias
 Minimize by using double-blind design
STUDY DESIGNS
 Cross-sectional
 Compares different segments of the population at the same time.
 For example, comparison of different groups of people by age at a
given timepoint
 Cons: cannot control for generational differences unrelated to age

 Longitudinal
 Study in which the same people (or animals) are followed and
periodically reassessed over a period of time
 Cons: time-consuming, costly, attrition

 Mixed longitudinal
 Combines cross-sectional and longitudinal design
BEHAVIORAL STATISTICS
MEASUREMENT, DESCRIPTIVE STATISTICS, & INFERENTIAL STATISTICS
RELIABILITY AND VALIDITY
Reliability
 The consistency of scores from a test (or outcome of a measure),
from one time and place to another
 Every time someone takes a test their score stays roughly the
same every time
RELIABILITY AND VALIDITY
Validity
 Ability of a test or instrument to measure what is was designed to
measure
RELIABILITY AND VALIDITY
Reliability and Validity
 You need both for a good study design  cannot have validity without
reliability
Good
Design

Poor
Design
DESCRIPTIVE STATISTICS
 Identify meaningful patterns and summaries in large sets
of data.
 Statistical procedures that organize and summarize
research data
 Cannot extend these statistics to a larger population
 Descriptives only pertain to our sample
 Types:
 Measures of central tendency
 Measures of spread/dispersion
DESCRIPTIVE STATISTICS: CENTRAL TENDENCY
 Mean
 Average of a set of scores
 Problem: easily skewed by outliers

 Median
 The halfway mark score in the data, half of the scores are above
the median and half are below
 Much less affected by outliers

 Mode
 Most frequent score in the data
DESCRIPTIVE STATISTICS: CENTRAL TENDENCY

Example 1: University Professor Salary (k per year)

9 numbers in data set:
90k, 95k, 96k, 99k, 100k, 105k, 105k, 115k, 116k
Mean: $102,333.34
Median: $100,000
Mode: $105,000
DESCRIPTIVE STATISTICS: CENTRAL TENDENCY
Modified Example 1: University Professor Salary (k per
year)
9 numbers in data set:
90k, 95k, 96k, 99k, 100k, 105k, 105k, 115k, 116k
800million
Mean: $88,978,333.30
Median: $100,000
Mode: $105,000
DESCRIPTIVE STATISTICS: MEASURES OF
DISPERSION
 Standard Deviation (SD)
 Measure of how clustered a group of
scores are around the mean

 Normal Curve
 Symmetrical
 Bell-shaped
 Frequency polygon representing a
normal distribution
 Normal distribution has special
characteristics:
INFERENTIAL STATISTICS

Allow experimenters to draw

conclusions from samples to the
larger population
 Generalization
 Need representative sample of
the population—ideally random
sample

 Statistical Significance
INFERENTIAL STATISTICS: HYPOTHESIS TESTING
Hypothesis Testing…when to reject?

 Statistical Significance: term used to refer to a result that is

extremely unlikely to occur by chance
 p-value (aka probability values): probability of the occurrence of a given event
 Significance: cut-off value
 p < 01: less than 1% chance the findings were because of chance

 Null hypothesis statistical testing (NHST):

 Testing to reject the null hypothesis
 Compare expected finding (the probability) with actual findings to determine if the
null can be rejected
 Null hypothesis
 Alternative hypothesis
NHST OUTCOMES

p-values set a threshold for type 1 or type II errors

RESEARCH ETHICS
ETHICS

 IRB: Institutional Review Board

 Review studies with human participants to assess risk of harm
 Required by all studies with human subjects

 IACUC: Institutional Animal Care and Use Committee

 Review studies with animals to assess humane treatment
 Required by all studies with animal subjects

 If a study is not approved by the IRB or IACUC then they will

not be allowed to collect data on a college campus and/or
present findings
ETHICS: HUMAN SUBJECTS

“Psychologists must
carry out
investigations with
respect for the
people

who
participate and with
concern for their
dignity and welfare”
(American
Psychological
Association, 2010a).
ETHICS:

ANIMAL SUBJECTS
Research with animals must demonstrate a clear
purpose, such as benefiting the health of humans or
other animals
 Must provide excellent housing, food, and veterinary
care
 Minimize pain and suffering
 Standards are more strict than the standard humane
society
ETHICALLY QUESTIONABLE RESEARCH:
THE TUSKEGEE SYPHILIS EXPERIMENTS

Prisoners, soldiers,
and mental patients
were deliberately
exposed to syphilis
and gonorrhea to
test the
effectiveness of
penicillin
ANOTHER KIND
OF RESEARCH
ETHICS
VIOLATION:
PLAGIARISM

Cite your
references!