1/59
Looks like no tags are added yet.
Illusory correlation
when two variables are perceived to have a relationship when, in fact, there is no logical reason for them to be correlated
Pseudoscience
The use of seemingly scientific terms and demonstrations to substantiate claims that are not based on scientific research.
What are the 4 primary goals of behavioral science?
Describe Behavior
Predict Behavior
Determining Causes
Explaining Behavior
Explain and differentiate between temporal precedence, covariation of cause and effect and alternate explanations.
Temporal Precedence: the order of events in which the cause precedes the effect.
Cause precedes effect.
Television viewing occurred first, and aggression followed
Covariation of Cause and effect: when the cause is present, the effect occurs; and when the cause is not present, the effect does not occur.
Children who watch are aggressive.
Children who don’t watch are not likely to be aggressive
Alternate explanations must be eliminated.
More violent because left alone more often (lack of parental supervision) and can choose violent video games; or more violent because of an unknown genetic predisposition
Define and differentiate between basic and applied research techniques
Basic: answers fundamental questions about human behavior
all
Applied: addresses practical problems
Specific people/contexts
Define and differentiate between a research question and a hypothesis
Research Question: General question/topic to be addressed
Hypothesis: A formal statement specifying how the variables being studied are related
What is a variable?
any event, situation, behavior, or characteristic that varies
Operational definitions
Tells you exactly how to measure a concept or variable
Choose a concept
Concept: hive productivity
Problem: “productive” is too vague
Create the operational definition:
Operational Definition: the total net weight in kilograms of harvested capped honey frames collected from one hive during the primary foraging season (April 1 - Sept 30)
Who: hive
What: net weight of capped honey
When: foraging season
Measure it: operational part
Count it
Distinguish between positive, negative, curvilinear and no relationship between variables
Positive: as time awake increases, tiredness increases
Negative: as runners' speed increases, race time decreases
Curvilinear: motivation and test scores
No Relationship: Color of your shirt and your test scores
Distinguish between independent, dependent, and confounding variables
Independent: the cause
The variable you can manipulate or change to observe its effect
Example: light exposure on plant growth
Manipulating light levels
Dependent: the effect
The variable you measure or observe to see how it responds
Example: plant height (the outcome)
Measuring differences in height due to light
Confounding:
An extra variable that is not controlled or accounted for
Affects both IV and DV
Any factor besides the independent variable(s) that could influence the dependent variable(s)
Example: solid quality (the hidden factor)
Distinguish between nonexperimental and experimental methods. Which one is best for determining cause and effect?
Nonexperimental Methods: Relationships are studied by observing variables of interest
Also known as the “correlation method” because it allows us to observe covariation between variables
not ideal when we ask questions about cause and effect.
can be difficult to tell the direction of cause and effect - which variable affects the other (directionality problem)
May be an issue of temporal precedence
Researchers face the third-variable problem
the possibility that one or more extraneous variables exist in the relationship being studied.
When we know that an uncontrolled third variable is operating, we can call the third variable a confounding variable.
Examples:
Will the students who expect to get a higher grade in the class do better than those who expect to get a lower grade?
Would the class feel more energized if Kit Kats suddenly fell from the sky
Experimental Methods: Involves direct manipulations of variables
One variable is manipulated (Independent variable - IV), and the other is measured (Dependent variable - DV)
Reduces ambiguity, and thus uncertainty, in the interpretation of results
The manipulated variable is called the independent variable.
The variable that is measured is the dependent variable.
Ensures that any differences between groups are due to the manipulated variable.
Examples:
Do people who take Drug A feel less nausea than people who take the placebo?
Did owls come up with more science- based suggestions than otters after watching a science tutorial?
Experimental methods is best for determining cause and effect
Experimental control
Extraneous variables are controlled by making sure every feature of the environment except the manipulated variable is held constant; accomplished by treating all participants in all groups the same.
Randomization
Any variable that cannot be held constant is controlled by making sure that the effects of the variable are random
What is validity?
truth/accuracy
A matter of degree
We can’t prove validity, but we can provide evidence to suggest that something is valid
Distinguish between internal and external validity
Internal: How accurate are the findings within the current study
A study has high Internal validity when strong inferences can be made that one variable caused changes in the other variable.
Must have:
Temporal precedence.
IV before DV.
Covariation.
Experimental shows effect, control does not.
Eliminate plausible alternate explanations
Key Questions:
How were constructs measured?
How were data analyzed?
Do the conclusions match the results?
Are there any better way to explain the results than the conclusions drawn?
Lab research
Experimenter controls internal validity
The stronger our internal validity, the weaker our external validity is
Example: A researcher tests a new reading app by keeping 30 kids in a quiet, distraction-free laboratory room with identical tablets, exact 20-minute timers, and no background noise. Kids using the app score higher on a reading test than kids using a standard e-reader in the same room
External: how accurate are the findings beyond the current study?
A study has high external validity when the results can be generalized to other populations and settings
Key Questions:
Does the sample represent the target population?
Would people behave similarly in real life as they did in the study?
Would the same thing happen if you did the study at a different time or place?
Field research
Natural contexts control external validity
The stronger our external validity, the weaker our internal validity is
Example: A team tests a financial budgeting tool by asking 500 households of different income levels, ages, and family sizes across five different states to use it for six months alongside their normal everyday spending habits
Construct validity
measures the theoretical concept accurately
Anxiety test measuring clinical anxiety vs stress
What is generalization?
extending the findings beyond the current study
Example: the one person I interviewed said that they love this class, so everyone must love this class!
How does generalization relate to external validity?
External Validity: the extent to which findings may be generalized to other populations and settings.
Distinguish between the Nuremberg trials, the declaration of helsinki, the tuskagee study, and the belmont report. What is the timeline for these events (when did each occur)?
Nuremberg Trials: 1946
Gave us 10 rules of research known as “the Nuremberg Code”
Developed in response to Nazi research atrocities
Declaration of Helsinki: 1964
Expanded on the Nuremberg Code and required journal editors to ensure published research was compliant
Tuskegee Study: 1972
IRB (Institutional Review Board) formed in 1974
Belmont Report: 1979
Still some questionable studies, so a report was created that defines the principles and applications that apply to medical and behavioral research
Distinguish between beneficence, justice, and respect for persons (autonomy)
Beneficence:
Two Rules:
The study should not harm participants
The study should maximize possible benefits while minimizing the risks
How to achieve Beneficence:
Minimize risks by debriefing the participants and alerting them of the study details and purpose
IRB can conduct a Risk-Benefit analysis
Common risks include: physical harm, stress, confidentiality
Benefits: individual, societal, incentives
Justice:
Definition: each person in the study has an equal chance of receiving the burdens and benefits offered
How to achieve justice:
We could randomly assign participants to different groups
This way, we make sure certain types of people don’t receive preferential treatment in the study
Avoid exploitation of vulnerable groups
Tuskegee Syphilis Study
Respect for persons:
Two Basic Ideas:
Participants should be treated as autonomous
Additional protections are provided for people with diminished autonomy
How to achieve Respect for Persons:
Give everyone information on the risks and benefits so they can make an informed decision on their participation
Informed consent
Appropriate action should be taken for those with diminished autonomy
<18, impaired mental capacities, terminally ill, comatose, etc.
What are the 5 principles of the APA ethics code? Distinguish between each. Which 2 are new/different compared to the Belmont report
Beneficence and nonmaleficence: do good and avoid harm
Promote welfare of clients
Maximize benefits, minimize risk
Fidelity and responsibility: Build trust and be responsible
Researchers do what they promise, such as reports, and do so in a professional manner
Establish trust and maintain standards
Uphold professional conduct
Integrity: Be Honest and Accurate
Promote truthfulness and accuracy
Avoid deception or misrepresentation
Psychologists should seek to promote accuracy
Added due to concerns of fraud, fabrication of data, and plagiarism, misrepresenting another's work as your own
Justice: Fairness and equality
Treat all people equally
Ensure access to benefits of psychology
Respect for People’s Rights and Dignity: Respect Dignity and rights
Honor privacy, autonomy, and confidentiality
Respect cultural differences
Which 2 are new/different compared to the Belmont report
Integrity
Fidelity and responsibility
What is the IRB? What do they do?
Institutional Review Board
Any institution that receives federal funds must have an IRB that is responsible for reviewing research conducted at that institution
Must have a minimum of 5 members
One must be an external member
What is an informed consent? Why is it important?
The idea that potential participants in research should be provided with information that might influence their active decision to participate
What is a debrief? Why is it important?
The procedure for revealing the true purpose of a psychological study to a research participant at the conclusion of a research session
What is reliability?
The consistency or stability of a measure
Pearson-product moment correlation coefficients (r)
What is this?
What scores can this range between?
How do we know if the relationship is strong, moderate or weak?
What is the difference between a positive and negative correlation?
What is this?
Describes the strength and direction of a linear relationship between two variables
What scores can this range between?
Can range from -1 to +1
How do we know if the relationship is strong, moderate or weak?
A correlation close to -1 or +1 tells us there is a strong relationship
What is the difference between a positive and negative correlation?
The - and + signs are simply there to tell us if it’s a negative linear relationship or a positive linear relationship
Negative relationships mean one variable is increasing while the other is decreasing
Positive relationships mean both variables are increasing
Test-retest reliability
Measure the same individual at two points in time
High correlation between test and retest indicates reliability
Alternate forms reliability
Measure the same individual at two points in time, but use different measures (different questions)
This avoids respondents remembering their answers from before
Takes a lot of time and effort
Internal consistency reliability
The assessment of reliability using responses at only one point in time
Because all items measure the same variable, they should yield similar results
Split-half reliability
Correlate one half of a test with the other half
What is Cronbachs alpha used for? What score is needed for an "excellent" cronbach's alpha?
Cronbach's alpha
Reliability testing
What score is needed for an "excellent" cronbach's alpha?
.9 or higher
Interrater reliability
The extent to which both raters agree on their observations
When two researchers make very similar observations, there is a high interrater reliability
What is validity? How is different from reliability?
Validity: the extent to which a test measures what it is supposed to measure
Asks “is the test measuring what you think it's measuring?”
Reliability tells us about the measurement error, but it doesn’t tell us about whether we have a good measure of the variable of interest
Define and distinguish between the 6 key types of validity
Face Validity: The content of the measure appears to reflect the construct being measured
If the new measure of depression includes items like “I feel sad” or “I feel down” or “I cry a lot,” this would be evidence that the measure has face validity
Example: Reviewing a math test and agreeing that the questions look like math problems
Content Validity: The content of the measure is linked to the universe of content that defines the construct.
Depression is defined by a mood and by cognitive and physiological symptoms. If the new measure of depression has content validity, it will include items from each domain
Example: a history exam covering key events from all taught centuries, not just one
Predictive Validity: Scores on the measure predict behavior on a criterion measured at a future time
If the measure of depression predicts a future diagnosis of depression, this is evidence that the measure has predictive validity.
Example: a high school GPA accurately predicting success in college
Concurrent Validity: Scores on the measure are related to a criterion measured at the same time (concurrently)
If two groups of participants were given the measures, and they differed in predictable ways (for example, if those in therapy for depression scored higher than those in treatment for an anxiety disorder), this would be evidence for concurrent validity
Example: a new anxiety scale giving similar scores to an existing well-established one
Convergent Validity: the degree to which a measure related highly to other measures of the same or similar constructs
Scores on the measure are related to other measures of the same construct.
If scores from the new measure, collected at the same time as other measures of depression (for example, Beck Depression Inventory or Duke Anxiety-Depression Scale), were related to scores from those other measures, then it could be said to have evidence for convergent validity
Example: scores on an intelligence test correlating highly with other IQ tests
Discriminant Validity: the extent to which a measure does not relate to constructs that are theoretically different
Scores on the measure are not related to other measures that are theoretically different.
If the new measure, collected at the same time as other measures of anxiety (for example, state/trait anxiety), was unrelated to those measures, then it could be said to have evidence for discriminant validity because it would indicate that what was being measured was not anxiety
What is measure reactivity?
A measure is reactive if awareness of being measured changes an individual's behavior
Define and distinguish between nominal, ordinal, interval, and ratio measurement scales
Nominal: placing data into categories
Examples: College majors, male/female, has red/blond/brown/green hair
Ordinal: classifying data based on a specific order or rank
Examples: color spectrum (ROYGBIV), position in a race (1st, 2nd, third)
Interval: used for numeric data that is expressed in intervals of some kind
Ordered, equal differences
Example: temperature
Ratio: used to express the ratio of some of the values of interval data
Has true zero
Examples: how much money you have, weight, frequency of behavior
What are Surveys? What is helpful about them?
What it is: survey research involves using questionnaires and interviews to ask people to provide information about themselves
Survey data provide useful information for making public policy decisions.
Social desirability response sets
People want to present themselves favorably and give the “good” or “correct” answer
The answer that reflects well on the test-taker
Avoid this by asking the same question in different ways, indirect questions
Be able to explain/identify survey questions which are: double- barreled, loaded, use negative wording, or use complicated language (instead of simple)
Double-barreled questions
Double-barreled questions ask about two things at once, which confuses respondents and muddles your data.
Examples:
“Do you feel happy and like eating marshmallows when you go camping?
“What do you do at home to make sure you eat well and to help the other people you live with eat well?”
“Learning new information is generally hard, interesting, overwhelming, and sometimes builds character”
Loaded questions
“Don’t you think it’s really crazy how some people don’t recycle, because they obviously don’t care about the environment? Yes/No”
Negative wording
“I don’t not want to be in class” (Agree/Disagree)
Complicated Language:
Exhausted instead of helpful
Distinguish between open and close ended questions
Open Ended:
Take more time to code
More complex to interpret
Can give valuable insights into what people are thinking
Example:
How was the presentation?
Answer in your own words
Close-ended:
limited number of response options
Easier to code and assign values to responses
Response options are the same for everyone
Example:
How was the presentation? (select one)
Great
Pretty good
Ok
Pretty bad
Terrible
What is a rating scale? Distinguish between graphic, semantic, and pictorial scales
Assign scores along some numerical dimension
Graphic Rating Scale: Requires a mark along a continuous line anchored with descriptions at the end
I feel: 1. Poor to 5. Excellent
Semantic differential scale: a measure of the meaning of concepts respondents are asked to rate on a series of bipolar adjectives
Example: Practical to impractical
Pictorial Scales: appropriate for populations who may have trouble understanding other scaled
Example: children
What is a sample? What is a population?
Sample: who you collected data from
The people you asked to respond, and then actually did respond to your survey
Population: the defined group of people from which a sample is drawn
Who you want to generalize the results of your study to
Probability sampling - define and distinguish between the 3 type of probability sampling
Probability Sampling: each member of the population has a known, non-zero probability of being included in the sample
Helps reduce bias
You can generalize to the population defined in the sampling frame
Statistical tests can be used to make conclusions
Avoids convenience sampling (when you pick a sample that is easiest to access)
Simple Random Sampling: Every member of the population has an equal chance of being selected
Example: surveying student opinions on campus dining
1. Generate random numbers
students are assigned numbers
2. Select corresponding students
A computer picts 100 random numbers for the sample
Stratified Random Sampling: The population is divided into subgroups (also known as strata), and random sampling techniques are then used to select sample members from each stratum.
Example: studying major-specific needs
1. Create strata
Divide students by major
2. Sample randomly within each stratum proportional to size
10 students randomly picked from each major are surveyed
50 total
Cluster Sampling: Existing groups or geographic areas, called clusters, are identified; samples are taken from those clusters
Example: assessing mental health suppor
1. Divide campus into clusters
Dorm buildings
2. Randomly pick clusters and survey all students inside
Every student in chosen clusters participates
NonProbability sampling - define and distinguish between the 4 types of nonprobability sampling
Non-probability sampling: Selecting a sample from a population without using statistical probability theory
Sample is biased
Some types of people are systematically left out
You cannot generalize beyond the sample
Adequacy of the sample cannot be known (prone to bias and sampling error)
Convenience Sampling: high accessibility
Ask people who are convenient to use in your study
Also known as Haphazard or “take-them-where-you-find- them” sampling
Introduces bias, but is easy
Ex. Go to a neighborhood and survey the first 10 people who answer the door
Purposive Sampling: specific criteria
Get a sample of people who meet a predetermined criterion
Sample does not need to be random because researchers are targeting specific population
Ex. Studying intervention to help with alcohol addiction, post flyers at AA meetings
Snowball Sampling: referral network
Procedure in which one or more current research participants recruit others to become part of the sample
Existing study participants recruit future participants from among their aquaintances
Particularly useful for hard to reach or specialized populations where members are difficult to locate
Ex: undocumented immigrant experiences
One participant who trusts the researcher might connect them with other undocumented immigrants
Quota Sampling: controlled proportions
The sample reflects the numerical composition of various subgroups in the population.
Choose a sample that reflects subgroups of a population
Like stratified random sampling, BUT people within subgroups chosen by convenience
What is the simplest possible experimental design?
Basic Experiments
What does it mean to say that an IV can have different levels?
Experimental and control groups
Compare posttest only design to pretest-posttest design
Posttest only:
A researcher using a posttest-only design must:
Obtain two equivalent groups of participants.
Manipulate the independent variable.
Measure the effect of the independent variable on the dependent variable.
Selection differences: the people selected to be in the conditions cannot differ in any systematic way.
Pretest-Posttest:
With a pretest-posttest design, a pretest is given before the experimental manipulation.
This makes it possible to see if the groups are equivalent at the beginning of the experiment
A pretest is usually not necessary if participants are randomly assigned and the sample is sufficiently large.
Attrition and mortality
Attrition: participant dropout
Participants leave a study before it finishes
Example: a long study on mindfulness
Participants find it time-consuming ot get bored
They stop coming to sessions
Data is lost
Mortality: differential Attrition
A specific type of attrition where one group drops out more than another, creating a bias
Example: a tough new fitness program study
The “high intensity” group drops out because its too hard
The “light stretching” group stays
It looks like the program works only for super fit people
Solomon 4-group design
Solomon four-group design: The experimental and control groups are studied with and without a pretest.
Half the participants receive only the posttest, the other half receive both the pretest and the posttest
If there is no impact of the pretest, the posttest scores will be the same in the two control groups and in the two experimental groups.
Define and distinguish between repeated measures designs and independent groups designs
Repeated measures design: All participants are in ALL conditions of the experiment
Also called a within-subjects design because comparisons are made within the same group of participants.
Independent groups design: Participants are randomly assigned to the various conditions so that each participates in only one group
Also known as a between- subjects design because comparisons are made between different groups of participants.
Participants are randomly assigned to ONLY 1 condition of the IV
Cons: Need more participants
Pro: No carryover, practice, fatigue, or order effects
Random assignment
Random assignment: The decision about which group each participant is assigned to is completely random and not controlled by the researcher.
Prevents systematic biases.
Yields equivalent groups.
Removes participant differences as an explanation for results
Define/distinguish between order, practice, fatigue, and carryover effects
Order effect: The order of presenting the treatments affects the DV
Practice Effect: Participants do well on DV because they’ve had practice
Fatigue effect: Performance deteriorates because of tiredness, boredom, or distraction.
Carryover Effect: Effect of first treatment carries over to second
Matched pairs design
Match people on a participant variable that could be a confounding variable, such as age or personality, then randomly assign
Good for small samples where randomization may not work
Costly and time-consuming
Straightforward vs. staged manipulation
Straightforward Manipulation: present written/verbal/visual material to participants to introduce IV
Staged Manipulation: stage events during experiment to manipulate the IV
Ex: when you want participants to be frustrated to simulate real world situations
Often involves a confederate: someone who pretends to be in the experiment
Ex. Milgram obedience experiment, Asch conformity experiment
Ceiling vs. floor effect
Floor Effect: occurs when a test or measurement is too difficult, causing most participants to score near the minimum possible score
Difficulty level is extreme
Scores cluster at the “floor”
Obscures true differences among individuals
Example: giving a graduate-level physics exam to intro psychology students
What it means: the test can’t distinguish between different levels of knowledge
Ceiling Effect: occurs when a test or measurement is too easy, causing most participants to score near the maximum possible score
Difficulty level is trivial
Scores cluster at the “ceiling”
Prevents identifying top performers or improvement
Example: asking college students to identify basic alphabet letters
What it means: the test can’t distinguish among participants with good reading skills or show improvement over time
Demand characteristics
when subjects pick up cues during an experiment and modify their behavior, thereby possibly affecting and altering the results of the study
Clues given to the participant about the hypothesis of the study
We don’t want them to try to confirm it
Job interview, meeting significant others’ parents
Possible Solutions:
Deception
Filler items
Debrief w/discussion
Only way to take these away fully: when participant isn’t aware
Ex. Observations, field settings
Placebo group
A way to deal with demand characteristics
Especially with drugs
Placebo Effect: we don’t know if the DV changed due to drug or participants’ expectations of the drug
Use placebo group to control for this
Single-blind vs. double-blind experiment
Single-blind experiment: occurs when participants are unaware of their assignment to the control group, but the researcher administering the study is aware
Used to reduce participant bias
Example: testing a new study aid software
Participants don’t know if they are using the new software or a placebo version, but the researcher knows and can inadvertently treat them differently
What it means: limits participant bias but not researcher bias
Double-blind experiment: occurs when neither the participants nor researchers collecting the data know who belongs to the control group and the experimental group
Kept hidden by a third party
Gold standard to eliminate both participants and researchers bias
Example: testing a new focus medication
Neither the participants taking the poll nor the administering the cognitive tests know who received which, until the codes are broken
What it means: eliminates both participants and researchers bias for highly reliable results
Pilot study
researcher does a trial run with a small number of participants
Manipulation check
directly measure whether IV manipulation has intended effect