Research Methodology
AP Psychology
Why do we need research?
Why? Because just asking people how/why they felt and acted the way they did can be misleading.
Common sense isn’t all it is cracked up to be.
A Demonstration
Group B close your eyes.
Group A: Psychologists have found that separation weakens romantic attraction. As the saying goes “out of sight, out of mind.
Write down why this might be the case.
A Demonstration
Group A close your eyes.
Group B: Psychologists have found that separation strengthens romantic attraction. As the saying goes “absence makes the heart grow fonder.”
Write down why this might be the case.
A Demonstration
Find a member of the opposite group. Discuss the question below, being sure to provide evidence.
What impact does separation have on romantic relationships?
When both a supposed finding and its opposite seem like common sense, we have a problem.
Hindsight Bias
The problem we have is hindsight bias, or the feeling that because something has already happened, it is inevitable. I-knew-it-all-along phenomenon
What’s the point?
Hindsight bias and overconfidence often lead us to overestimate our intuition (common sense). Scientific inquiry can help us move from illusion back to fact.
Experimental Research
Research Hypothesis (Directional): A prediction stating expected results
Research Hypothesis (Non-Directional): A prediction which does not specifically state the expected difference
Null Hypothesis: A statement that there will be no statistically significant difference as a result of the experimental condition.
Variables
Independent Variable: What is manipulated by the researcher
Dependent Variable: What is being measured Confounding Variable: An unintended factor
which influences the outcome of an experiment
Groupings
Experimental Group: the participants who are exposed to treatment, one version of the independent variable
Control Group: contrasts with the experimental group, serves as a comparison for evaluating the effect of treatment
Placebo Group: experimental results caused by expectation alone
Other Experimental Considerations
Researcher (Experimenter) Bias or Expectancy
Demand Characteristics: A cue that makes participants aware of what a researcher expects to find
Single-Blind Experiments Double Blind Experiments
Experimental Designs
Independent Measures – between subjects design
Matched Pairs Repeated Measures – within subjects design Single participant/subject
Sampling Techniques
The sample selected should be representative of the target population
Avoid selection bias
Assign participants to each condition randomly
Sampling Techniques
Random Selection: each participant has an equal chance
Representative Sampling: each participant exhibits characteristics of the target population
Stratified Sampling: use a subset with each of the common characteristics, then randomly assign to each condition
Sampling Techniques
Systematic Sampling: every nth person is selected
Purposive Sampling: participants are chosen on the basis of particular characteristics
Convenience/Opportunity Sampling: participants are a pre-existing group
Snowball Sampling: referrals from initial participants generate additional participants
Interview Methods
E-mail and Telephone Focus Groups (small group interviews) One-to-One Interviews
E-Mail & Telephone Interviews
ADVANTAGES High response rate,
sample bias More people enjoy e-
mail Allows for diverse
sample Low risk for researcher
DISADVANTAGES Difficult to achieve
rapport Literacy of participants Lack of visual clues
may cause problems in interpretation
Focus/Small Group Interviews
ADVANTAGES Efficient generation of
large amounts of data Participants tend to
enjoy the experience Relatively inexpensive
and flexible
DISADVANTAGES Not easy to conduct
well Limited number of
questions covered Confidentiality can be a
problem Participants may not
give personal information in a group
One-to-One Interviews
ADVANTAGES Participants more
willing to divulge personal information
More data rich Allows for a relationship
between interviewer and participant
Allows for follow-up
DISADVANTAGES Time consuming and
expensive Difficult to conduct well Researcher must keep
subjectivity in check Report and reflect
interview accurately
Methods of Interviewing
Conversational Interviews Structured Interview Semi-Structured Interview
Conversational Interviews
ADVANTAGES Participants relax and
speak freely No specific set of
questions Participant can be
interviewed without knowing they are being studied (ethics)
DISADVANTAGES Not standardized Participant may control
the interview
Structured Interview
ADVANTAGES Very straight forward Easy to evaluate and
compare candidates Less time consuming to
analyze
DISADVANTAGES Does not yield rich
information, requires a structured response
Does not engage participant on a personal level
Limits information obtained
Semi-Structured Interview
ADVANTAGES Answers criticism of
structured interview Pre-set topics but no
specific questions Allows for more
interaction
DISADVANTAGES Places limits on what is
asked Not fully
conversational, may not allow for obtaining personal information
Traditional Transcription
Recorded on tape and transcription focuses on words only
Quick and easy, good for market research Doesn’t account for tone of voice, speech
rate
Post-Modern Transcription
Non-verbal cues are recorded and taken into account when transcribing
Allows for full interview experience Produces rich data and detailed analysis Difficult to maintain reliability Time consuming Distracting to participant Some people use more body language than others Must be transcribed quickly
Questionnaires/Surveys
ADVANTAGES Simple way to gather
data Standardized Generalizable
information Low cost, efficient,
takes less time Allows anonymity,
frankness
DISADVANTAGES Inaccurate responses Characteristics of
participants affect responses
Low response rate Ambiguous questions Interviewer bias
Participant Observation
ADVANTAGES Researcher blends in
with participants Researcher
subjectively experiences what participants are experiencing
DISADVANTAGES Observer might
influence behavior of the group
Interpretations of behavior may vary
Non-Participant Observation
ADVANTAGES Researcher blends in
and does not influence the behavior of group
DISADVANTAGES Researcher doesn’t
have same experience as group
Researcher may still influence group
Researcher may alter behavior of group
Observation Data Collection
Event Sampling: record key events each time they occur
Time Sampling: observe for a specific amount of time or timed units
Point Sampling: observe one individual in the group at a time
Content Analysis
Researcher determines media to be sampled then devises code units to be measured
Printed material: themes, characters, gender Television: race, gender, stereotypes Advertising: subtle influences on consumers Internet/E-mail: bias Objectivity of code units selected
Content Analysis
Raw Data ThemesHigher Order ThemesInterpretation
Case Study
ADVANTAGES Detail and depth High ecological validity May be only method for
studying rare behavior
DISADVANTAGES No determination of cause
and effect May be an unrepresentative
sample Reliability may be
questioned Difficult to replicate Time consuming and
expensive
Verbal Protocols
A record of what people say when they are asked to think out loud as they perform a task
Useful for information that might not be revealed in an interview
May not include all vital information; participants need to be trained for protocol to be useful; protocol may hinder performance of task; requires articulate participants
Q Sort
A subjective measure of personality characteristics used only in the humanistic perspective
One hundred cards are ranked from most descriptive to least descriptive in groups of 5,8,12,16,18,16,12,8,5
Ranking can be done by the participant or by another observer
Psychological Testing
Several types of tests to measure ability, interests, achievement, personality
Convenient way to gain insight into aspects of ability and behavior
Can be expensive, may not provide a complete representation of an individual’s true abilities or personality
Psychological Test Validity
Validity: the extent to which a test measures or predicts what it is supposed to
Content validity: the extent to which a test samples the behavior that is of interest
Criterion validity: the behavior the test is designed to predict
Predictive validity: the success with which the test predicts the behavior it is designed to predict
Psychological Test Reliability
Reliability: the extent to which a test yields consistent results, as assessed by the consistency of scores on two halves of the test, alternate forms of the test or on retesting
Triangulation
Combining data from different methods to collect richer data or to strength or test conclusions
Triangulation
Data Triangulation: using different times, locations and/or participants
Investigator/Researcher Triangulation: using multiple observers
Theory Triangulation: using theories from more than one perspective in interpreting data
Method Triangulation: using more than one method
Reflexivity
Personal Reflexivity: ways in which the researchers values, beliefs, experiences influenced the research also, how the research influenced the investigator
Reflexivity
Epistemological Reflexivity: thinking about the way in which knowledge has been generated in the study. Suggestions for further research
Descriptive Statistics
Measures of Central Tendency: Mean, Median, Mode
Measures of Dispersion: Range, Variance, Standard Deviation,
Inferential Statistics
Is the difference in the data significant or is it due to chance?
Used with quantitative data Tests of Statistical Significance:
chi square
Mann - Whitney U Test
Wilcoxon Signed Rank Test