l 5 Flashcards
(34 cards)
Factorial design
more than one IV
– Each IV is called a factor
Notation system in Factorial design
– Digits represent IVs
– Values represent the # of levels
• e.g. 2x3 factorial ( “two by three”)
– 2 IVs, one with 2 levels, one with 3 = 6 conditions
• e.g. 2x4x4 factorial
– 3 IVs, with 2, 4, and 4 levels = 32 conditions
Main effects
the separate effects of each
independent variable on the measure
e.g. main effect of training type
Research example
Gladue & Delaney (1990)
Gender differences in perception of attractiveness of men and women in bars
• Possible alcohol consumption confound?
– Found no relationship between alcohol consumption
and attractiveness ratings
• Possible confound of actual attractiveness
differences at different times of night?
– Replicated their result using standard photographs
– Note: chose photos of moderate attractiveness to
avoid floor or ceiling effects
interaction effect
When the effect of one independent variable depends on the level of another
• e.g. the effect of psychotherapy is larger when
the receiver is highly motivated to change
• e.g. using a cell phone affects driving
performance more at night than during the day
• Some effects may not be detectable without
using a factorial design
– e.g. a study of lecture vs lab emphasis alone would
find no effect, but only because there is an opposite
effect for science and humanities majors
• In the presence of an interaction, one must
interpret main effects with caution
Grant et al interaction example
“Context dependency for meaningful text material
was examined using two standard academic
testing techniques: short answer (recall) and
multiple choice (recognition). Forty participants
read an article in either silent or noisy conditions;
their reading comprehension was assessed with
both types of test under silent or noisy conditions”
person by environment design
we have both unmanipulated variables (ie. person) and manipulated variables (ie. environment)
mixed design
we have both betweensubjects and within-subjects variables
Example mixed design
• Cohen et al. (2004). Fatal Attraction: The Effects of
Mortality Salience on Evaluations of Charismatic, TaskOriented, and Relationship-Oriented Leaders.
• A study was conducted to assess the effects of mortality
salience on evaluations of political candidates as a
function of leadership style … we hypothesized that
people would show increased preference for a
charismatic political candidate and decreased preference
for a relationship-oriented political candidate in response
to subtle reminders of death.
• “Following a mortality-salience or control induction, 190
participants read campaign statements by charismatic,
task-oriented, and relationship-oriented gubernatorial
candidates; evaluated their preferences for each
candidate; and voted for one of them.”
• 2x3 mixed factorial, with counterbalancing
• IV1: mortality induction or control (between-subjects)
• IV2: leadership style of political candidates: charismatic,
task-oriented, and relationship oriented (within-subjects)
• DV: evaluation of candidates
Example PxE Design
• Inzlicht & Ben-Zeev (2000). A threatening intellectual
environment: Why females are susceptible to experiencing
problem-solving deficits in the presence of males.
• Is a situational cue, such as gender composition,
sufficient for creating a threatening intellectual
environment for females—an environment that elicits
performance-impinging stereotypes?
• Male or female subjects completed a difficult math test in
3-person groups, each of which included 2 additional
people of the same sex (same-sex condition) or of the
opposite sex (minority condition)
Example Mixed PxE Design
• Strayer & Drews (2004). Profiles in driver distraction:
effects of cell phone conversations on younger and older
drivers.
• “Our research examined the effects of hands-free cell
phone conversations on simulated driving
• 2x2 PxE mixed factorial
• IV1: age of drivers (subject variable)
• IV2: driving with and without cell phones
(manipulated variable and a repeated measure)
• DV: reaction time
• Two main effects, no interaction
– Younger drivers outperformed older drivers
– Cell phone use impaired driving
Summary of designs
• Factorial designs allow us to evaluate the effects
of multiple IVs on the DV
• There are different types of factorial designs,
depending on how you manipulate your IVs.
– Between-subjects, repeated measures, mixed, PxE
• Main effects of each IV and interactions among
IVs are the key results from factorial designs.
APA Ethical Principles and Code of Conduct
• Guidelines for ethical behavior for the practice of
research, clinical work, and teaching in psychology
• “Designed to guide and inspire psychologists to the very
highest ideals of the profession”
• Code contains:
– 5 general principles
– 89 standards of practice
Ethical History
• Nuremberg Code (1947) – Nazi doctors put on trial • Declaration of Helsinki (1964) – Revision of the Nuremberg Code • Belmont Report (1979) – In part due to Tuskegee Syphilis Study – Ethical principles and applications that inform APA Ethical Guidelines
Five general principles in APA Ethical Principles and Code of Conduct
- Beneficence and Nonmaleficence
– Constantly weigh costs & benefits; protect from harm;
produce for greatest good - Fidelity and Responsibility
– Be professional; constantly be aware of responsibility to
society - Integrity
– Be scrupulously honest - Justice
– Always treat people fairly - Respect for Peoples’ Rights and Dignity
– Safeguard individual rights; protect rights of privacy and
confidentiality
Guidelines for research with humans
• Several particulars. General logic:
– Identify potential risks
– Protect participants from physical/psychological harm
– Justify remaining risks
– Obtain informed consent
– Take care of participants after the study (debriefing)
• Can be boiled down to three general principles
– Make sure the benefits outweigh the costs
– Obtain appropriate informed consent
– Treat volunteers well
The Institutional Review Board
• The IRB determines whether the project meets ethical guidelines – Some research is exempt from review – Some research gets an expedited review – Some research requires a full review • Key factor: degree of risk to subjects – No risk (could be exempt) – Minimal risk (expedited review) – At risk (full review)
Informed consent
• Sufficient information must be provided to
participants to decide whether to participate
– Study’s basic description
– How long participation will take
– May quit at any time
– Confidentiality and anonymity ensured
– Contact information given (researcher, IRB)
– Opportunity to obtain final results of the study
– Signatures
Challenges with consent
• Sometimes we have to use deception
– e.g. Milgram’s obedience study
• Cover story: effect of punishment on learning
• Real purpose: limits of obedience to authority
• Consent can be difficult with special populations
– e.g. children: parental assent also needed
– e.g. prisoners: need to avoid feelings of coercion
Debriefing
• Debriefing participants after the study serves two
important purposes:
1. Dehoaxing: revealing the true purpose of the expt
2. Desensitizing: reducing any stress or negative
feelings that may have resulted from the expt
• Not essential that participants be debriefed on
all aspects of the study
– Sometimes we have scientific reason to prevent
participants from sharing information with eachother
– Called participant crosstalk (e.g. Edlund et al, 2009)
Animal research
• This has been an contentious issue for decades
– Hundreds of animal facilities have been vandalized or
destroyed over the years
– Some animal researchers have been directly targeted
– 1992: Animal Enterprise Protection Act
– 2006: Animal Enterprise Terrorism Act
• Proponents argue that allegations of abuse are
exaggerated and the research is justified by its
value to both humans and non-human animals
Ethical guidelines for animal research
• Justify the study
– Must demonstrate that the research is of sufficient
potential significance to outweigh any harm or
distress to the animals used
• Care for the animals
– Must demonstrate expertise with the species, have a
veterinarian check the facility and be on call
– Euthanasia must be justified as the necessary or
most humane way to conclude the research
• Minimize use
– Must consider alternatives. Especially relevant when
using animals for educational purposes
Case Studies in Ethics
• Little Albert (Watson & Rayner, 1920)
• Justification: he would learn such fears anyway
• Chose to test whether the fear would persist
rather than whether it could be extinguished
Case studies in Ethics
• Milgram’s (1974) shock study
• “[I] observed a mature and initially poised
businessman enter the laboratory smiling and
confident. Within 20 minutes he was reduced to
a twitching, stuttering wreck, and was rapidly
approaching a point of nervous collapse”
• Experimenters were trained to say things like
“The experiment requires that you continue” and
“It is absolutely essential that you continue”
