Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
126 Cards in this Set
- Front
- Back
Manipulation |
occurs when an experimenter systematically alters the levels of a variable (IV) too see if changed in the IV lead to corresponding changes in the DV. |
|
Random assignment |
a technique for assigning participants to different conditions in an experiment. The use of random assignment means that every participants in the study has an equal chance of being assigned to any of the conditions of the study. The use of random assignment makes it highly likely that the groups of Ps in different experimental conditions are highly similar to one another. Equates groups by randomly distributing extraneous variables over the treatment groups making it unlikely that an extraneous variable will serve as a confound. |
|
Matching |
a common approach adopted in the old days to overcome the barrier of individual difference. Consisted in matching each participant in the experimental condition with a very similar participant in the control condition. (hard to do and didn't work) |
|
Random sampling |
a technique for deciding which participants in a population are selected for inclusion in a study. When true random sampling is used, every person in a population has an equal chance of being selected into the study. The use of random sampling makes it highly likely that the Ps who take part in an investigation are highly similar to the entire population of people whom they were sampled. |
|
Confound |
any variable other than the IV that varies systematically with the IV and might be casually related to the DV. It is a threat to internal validity. (Ex: angry mood vs overcrowding experiment shows that the relationship could be due to angry experimenter, number of people in room is confounded with experimenter behavior/ tattoo example) |
|
Procedural confounds |
like environmental confounds in correlational studies occur when an experimenter accidentally manipulates two or more things at once |
|
Semantic priming effect
|
people recognize most words more quickly than usual when they have just been exposed to words that have a similar meaning.
|
|
Noise
|
an extraneous variable in an experiment that influences the DV but that is evenly distributed across experimental conditions. This is not a threat to validity, but it may decrease a researcher's ability to detect the effect in which he or she is interested.
|
|
Artifact
|
A variable that is held constant in a study but which influences the relation between the IV and the DV. Might represent a restricted context under which the effect is observed. Consider a drug study that includes only males. If the drug being studied works for men and not for women, gender would be an artifact. It is a threat to external validity.
|
|
Mudane realism
|
the degree to which the physical setting in an experiment is similar to the real-world setting in which the experimenter's independent and dependent variables are more likely to operate.
|
|
Experimental realism
|
degree to which the subjective experiences of research participants are realistic or psychologically meaningful. Well-designed experiments can be high in experimental realism even when they bear little physical resemblance to the real world. Ex: making Ps anxious or fearful by telling them they are about to receive electric shocks.
|
|
Manipulation checks
|
A measure designed to see if a manipulation truly puts people in the psychological state that the experimenter wishes to create.
Help you establish construct validity. Self-report/Behavioral It can help you interpret the cause of a significant effect or rule out an unsuccessful manipulation of the IV as an explanation for no effect of the IV on the DV. Ex: check for physiological arousal (IV) by measuring heart rate |
|
Optimistic bias
|
refers to the fact that most people think that they will have better outcome in life than they usually have
|
|
Planning fallacy
|
based on the idea that most people think that most things they do will go better than they actually go. (time to complete a task)
|
|
Null hypothesis
|
experimental group and control group are not really different after manipulation and that any apparent difference between the two groups is simply due to chance
|
|
Statistical significance
|
the likelihood that an observed effect would occur by chance alone. This number is typically expressed with a p-value to indicate the number of time out of 100 that the observed effect would occur by chance.
|
|
Alternative hypothesis
|
the hypothesis that any observed differences between the experimental and control group are real.
|
|
Alpha level
|
the probability the value that serves as the standard for rejecting the null hypothesis in a statistical test. Alpha is mostly set at 0.05, meaning that researchers conclude that their findings are real only if findings as as extreme as theirs would have occurred fewer than five times in a hundred.
|
|
Probability
|
the number of specific outcomes that qualify as an event divided by the total number of all possible outcomes. Can never be lower than zero or higher than 1
|
|
Inferential statistics
|
statistics used to make inferences, to interpret or draw general conclusions about a set of observations
|
|
Type I error
|
incorrectly rejecting the null hypothesis when it is in fact correct, the likelihood of making a type I error is a direct function of where we set out alpha level.
|
|
Type II error
|
occurs when we fail to reject an incorrect null hypothesis, we fail to realize our study has revealed something meaningful.
|
|
One-Way Designs
|
experiments with a single independent variable
|
|
Factorial designs
|
experiments with two or more independent variables that are completely crossed. Every level of the IV appears in combination with every level of of every other IV.
|
|
Two-groups designs
|
A type of one-way experimental design in which there is only one independent variable and only two levels of this variable. Simplest experimental design
|
|
Multiple-groups design
|
A type of one-way experimental design in which there is one independent variable with three or more levels.
|
|
True experiment
|
a research design in which the experimenter manipulates one or more variables and then assesses the impact of this manipulation. Experiment isolates the effect of a variable by holding all other variables constant while this variable is manipulated. Enables us to make inferences about cause and effect.
|
|
Cause (causation)
|
A contextually dependent event that makes something else exist
In an experiment the cause is the variable that the variable that the researcher manipulates: the independent variable |
|
Effect (causation)
|
The effect of the IV is observed in the variable the researcher measures: the dependent variable
Identifying a true effect requires that Ps simultaneously be and not be exposed to something (use of control conditions) |
|
Causation Criteria (John Stuart Mill)
|
1- Covariation: changes in one variable (cause) must correspond with changes in the other (effect)
2- Temporal sequence: cause must precede the effect 3- Eliminating confounds: no other explanations must exist for the effect |
|
Key Features of a True Experimen t
|
1- manipulation of the IV (at least two levels)
2- random assignment of participants to groups |
|
Manipulations of the IV
|
1- Presence vs absence (drug vs no drug)
2- Amount of stimulus (0 mg of drug vs 10 mg) 3- Type of stimulus (happy vs scary film) 4- Instructional manipulation (tell Ps that puzzle is solvable or not solvable when the same puzzle was given in both conditions) 1-3 are event manipulations |
|
Extraneous variable
|
any variable other than the IV that may influence the DV (age, IQ, race, last time you had ice cream, personality)
|
|
Between-subjects design
|
Each participant serves in one and only one condition of an experiment ( tmnt 1 OR tmnt 2)
|
|
Within-subjects design/ Repeated Measures
|
Each participant serves in one and more than one condition (tmnt 1 AND tmnt 2).
|
|
Advantages of Between-Subjects Design
|
1- Since Ps are only assigned to one group, they are less likely to figure out the scientific hypothesis under study
|
|
Disadvantages of Between-Subjects Design
|
Error variability due to individual differences
This means decreased sensitivity to detect significant differences between groups This means you need more Ps |
|
Systematic variability
|
variability attributed to the effects of manipulations of an IV (and any confounds)
|
|
Error variability
|
Variability attributed to the effects of extraneous variables (individual differences: subject variability)
too much error variability gives a small t which means that it is less likely the difference between groups will be statistically (t=systematic/error) |
|
Advantages of Within-Subjects Design
|
- fewer participants are needed to detect changes
- eliminate person confounds, no individual differences are contributing to error variablity so there is greater sensitivity |
|
Problems of Within-Subjects Design
|
- Sequence effects: the simple passage of time begins to take a toll on people's responses (boredom, tired). This problem can be remediated by using counterbalancing.
- Carryover effects: Occurs when participants' responses in one condition of a study influence their responses in a subsequent condition. (Ex: the humor response to the dog video affected response to the cat video) - Increase likelihood Ps will find out experimenter's hypothesis |
|
Counterbalancing
|
A method of control in within-subjects studies whereby the researcher varies the order in which participants experience the different experimental conditions (by presenting the conditions in different orders to different participants).
|
|
Types of Carryover Effects
|
- Order effect
- Practice effect - Interference effect |
|
Order effect
|
occurs when a question takes on a different meaning when it follows one question than when it follows another (ex: asking a dating question before SWL question)
|
|
Practice effect
|
occurs when people's experience in one within-subject condition facilitates their performance in a later condition. (same as testing effects and the opposite of interference effects)
|
|
Interference effect
|
occurs when people's experience in one within-subject condition disrupts or diminishes their performance in later within-subject condition.
|
|
Solutions to Within-Subject Design Problems
|
- Counterbalancing (complete counterbalancing)
- Incomplete counterbalancing (reverse & partial counterbalancing) - Structured debriefing
|
|
Complete counterbalancing
|
a comprehensive form of counterbalancing in which the researchers present every possible order of the different experimental conditions in a within-subjects experiment
|
|
Incomplete counterbalancing
|
any method of counterbalancing in which the researcher using a within-subjects manipulation does not present every possible order of the within-subjects variables across Ps
|
|
Reverse counterbalancing
|
a form of counterbalancing the order of presentation of the experimental conditions in a within-subjects experiment. Half of Ps are presented with one specific order, and other half of Ps are presented with the opposite order.
- average serial position is the same, but some conditions always occur in the same serial position. |
|
Partial counterbalancing
|
a category of incomplete counterbalancing in which several different orders are presented and in which each condition appears with equal or roughly equal frequency in every possible ordinal position in the experiment. Two forms are choosing several orders at random from pool of all possible orders or using Latin square procedure.
|
|
Balanced Latin square procedure
|
A form of partial counterbalancing in a within-subject design in which each condition appears with equal frequency in every possible ordinal position.
Generates a fixed number of orders where each condition appears exactly once in each serial position and where each condition must precede and be followed by every other condition an equal number of times. |
|
Structured debriefing
|
an interview conducted with Ps immediately after they have completed a study to determine exactly what Ps thought the researcher expected to find. This always occurs prior to the point at which experimenter conducts full debriefing. Important for within-subjects design.
|
|
Effect
|
the different between conditions with respect to the DV
|
|
Sensitivity
|
- the statistical power to detect significant difference between conditions when there is a real difference to be detected
- Within-subject designs have greater sensitivity than btw-subjects because they eliminate error variability due to individual differences. |
|
Error variability
|
composed of individual differences between Ps and error variability of measures used (measurement error)
Smaller error variability gives a larger statistic |
|
Demand characteristic
|
Characteristic of an experiment that subtly suggests how people are supposed to behave
|
|
Qualification
|
an approach to scientific hypothesis testing in which a researcher attempts to identify the conditions under which a hypothesis is and is not true. Closely linked to interactions in factorial designs (show that the effect of an IV on a DV depends on the level of the second IV).
|
|
Main effect
|
the overall influence of one IV, as reflected in the difference between the means
Refers to the overall effect of the IV, averaging across all levels of the other IVs a statistical term indicating that, on the average, an IV in a factorial design had a significant effect on the DV. Main effects can be taken at face value only when there is no interaction between the IVS in the study. |
|
Interaction
|
Interactions exist when the effect of one IV on a DV depends on the level of the second IV.
A statistical term indicating that the effect of an IV on the DV is different at different levels of another IV. The presence of an interaction means that one cannot make a simple, blanket statement about the independent and dependent variables in a factorial study. |
|
Ordinal/Spreading Interaction
|
When an effect exists at one level of a second IV but is weaker or nonexistent at a different level of the second IV.
|
|
Crossover/Disordinal interaction
|
Occurs when there are no main effect of either IV and when the effects of each IV are opposite at different levels of the other IV
Effects of both IVs are opposite and roughly equal at the different levels of the other IV |
|
Simple effect tests
|
a set of follow-up tests that are conducted when the statistical analysis in factorial design yields a significant interaction. Simple effect tests clarify the precise nature of an interaction.
|
|
Advantages of Factorial Designs
|
- more efficient in that they allow us to look for more than one main effect at a time in study
- more comprehensive in that it allows us to see how different variables work together to influence a phenomenon - external validity, main effects observed in a factorial design will generalize across whatever levels of the other IVs |
|
Mixed-Model designs
|
designs in which at least one IV is manipulated on a between-subjects basis and at least one other IV is manipulated on a within-subjects basis
|
|
Template for Describing an Interaction
|
In the ___ condition (IV 1, Level A),___ (DV) was___ (lower/higher/the same) in the ____ condition (IV 2, Level X) compared to the ___ condition (IV 2, Level Y). However...
In the ___ condition (IV 1, Level B), ___ (DV) was____ (lower/higher/the same) in the___ condition (IV 2, Level X) compared to the ____condition (IV 2, Level Y). |
|
Risk-benefit analysis
|
a comparison of the potential negative effects (risks) in a study to its potential positive effects (benefits). Rule: Benefits must always exceed the costs. To receive accreditation by the APA, colleges and universities must have an IRB perform a high risk-benefit analysis of all studies involving human subjects.
|
|
Internal review boards
|
perform risk-benefit analyses to ensure that all studies meet consensual community standards of ethical behavior. committees are made up of a group of instructors and researchers from university, one or more staff university members with experience in the area of research protocol, one or more people of local community.
|
|
Informed consent
|
- the ethical principle of advising Ps about any forseeable risks that are posed by a specific study and getting Ps permission to take part in the research prior to the beginning of the study
- needs to contain sufficient info that Ps can make an informed decision, nature of experiment, overview of procedures, time, potential risks and benefits. |
|
Freedom from coercion
|
the ethical principle of respecting Ps rights to drop out of a study if they choose to do so. This mainly consists of making it clear to Ps that they have the right to stop participating without fear of negative consequences.
|
|
Confidentiality
|
ensures anonymity, assigns identification numbers to Ps.
|
|
Debriefing
|
the ethical principle of educating Ps about the nature and design of an investigation to be sure that when they leave the study, their frame of mind is at least as favorable as when they arrived. When deception is used, debriefing at the end includes informing Ps about the nature of and reasons for the deception.
Learning experience: importance of psychological science and exposure to common methods |
|
Ethics
|
a set of principles guiding right conduct, the rules or standards governing the conduct of a person or the members of a profession
|
|
Syphillis study
|
Researchers wanted to study the long-term effects of syphilis. Even when penicillin was found to cure syphilis the researchers prevented the infected Ps from receiving treatment. The Ps were all Black and the researchers were all white.
|
|
Guidelines Exception from Informed Consent
|
- Approval by IRB
- Consultation with community - Public disclosure of study's design before the study begins and when the study is over to share the results -Notification of patients who were involved in the research - Oversight by an independent group of experts charged with monitoring the research for safety. |
|
General Ethical Principles
|
1- Beneficence and nonmaleficence: strive to benefit and do no harm2- Fidelity and responsibility: establish relationships of trust, accept responsibility 3- Integrity: promote accuracy and honesty 4- Justice: equal benefits and quality for all5- Respect people's right and dignity: respect dignity and worth, everyone had right to keeping personal info private and self-determination
|
|
APA Ethical Standards
|
1- Freedom from coercion
2- Protection from physical and psychological harm 3- Risk-benefit rule 4- Informed consent 5- Debriefing |
|
Minimal risk
|
any harm or discomfort while participating is no greater than that experienced in daily life or during routine physical or psychological tests
|
|
Deception
|
refers to witholding information from Ps or intentionally misinforming them about an aspect of the research
not divulging the exact hypothesis does not necessarily count as deception |
|
Pros of Deception
|
- allows researchers to study individual's natural behavior
- allows opportunities to investigate behaviors and mental processes not easily studied using nondeceptive methods |
|
Cons of Deception
|
- contradicts the principle of informed consent
- the relationship between researcher and P is not open and honest - frequent use may make individuals suspicious about research and psychology |
|
Protocol for IRB includes:
|
- a careful summary of the goals of research
- a detailed description of the procedure - consent and debriefing forms |
|
Benefits of Multiple-Group Designs
|
- Increases external validity
- Helps you identify the shape of the relationship between the IV and the DV - Helps you improve construct validity |
|
Pseudo-experiment
|
a false experiment, a research design in which the investigator exposes one or more people to a variable of interest and notes that the people exposed to this treatment felt, thought, or behaved as expected.
|
|
Control group
|
a group that is used as a standard of comparison for assessing the effects of an experimental manipulation or psychological treatment
|
|
Selection bias
|
Choosing research Ps from a nonrepresentative sample by using imperfect sampling techniques rather than true random sampling. Threat to external validity.
|
|
Nonresponse bias
|
the bias that occurs in research when substantial proportion of those invited to take part in a study refuse to do so. If those who agree to take part are different from those who refuse, the resulting bias is similar to selection bias and represents a threat to external validity.
|
|
History
|
changes that occur over time in a very large group of people such as those living in a city, nation, or culture. When an investigator conducts a pretest-postest study in which all Ps receive a tmnt, changes due to history may mask tmnt effect. History represent a threat to internal validity.
An event occuring between tmnt and outcome that could affect the DV (ex: heath ledger death on drug education programs) |
|
Maturation
|
changes that occur over time in a specific person or group of people due to normal development or experience. When an investigator conducts a pretest-postest study in which all Ps receive a tmnt, changes due to maturation may mask tmnt effects. Maturation represents a threat to internal validity.
|
|
Regression toward the mean
|
the tendency for people who receive high or low scores on a particular measure to score closer to the mean on a subsequent testing.
Regression toward the mean occurs because performance is influenced by error or luck as well as by a person's true score. Threat to internal validity. Pretest-postest studies without a control group are most likely influenced by RTM. Observed score = true score (skill) + random error (luck); luck changes but ability does not |
|
Hawthorne effect
|
the increases in productivity that may occur when workers believe that their behavior is being studied or believe they they are receiving special tmnt. Because Ps who are receiving tmnt are more likely to believe these things than are Ps in control condition, Hawthorne effects may be mistaken for tmnt effects and are thus threat to internal validity.
|
|
Mere measurement effect
|
ex of Hawthorne
refers to the tendency for Ps to change their behavior simply because they have been asked how they will act in the future. external validity |
|
Testing effect
|
threat to internal validity
a problem in the pretest-postest designs that have no control group tendency of most Ps to perform better on a test the second tie they take it - can be fixed by adding a pretested control group or true experiment only doing postest |
|
Attrition/Experimental Mortality
|
- the failure of some research Ps to complete an investigation. The longer the study the greater risk of attrition
- homogenous attrition: when attrition rates are equal across experimental conditions (external validity threat) - heterogenous attrition: when attrition rates are different across experimental conditions (internal validity threat) |
|
Participant Reaction Bias
|
the bias that occurs when research Ps realize they are being studied and behave in ways in which they normally would not behave.
3 types: expectancies, reactance, apprehension (internal validity mostly) |
|
Participant expectancies
|
form of participant reaction bias that occurs when Ps conciously or unconciously try to behave in way they believe to be consistent with the experimenter's hypothesis. P expectancies are a threat to internal validity.
|
|
Participant reactance
|
the form of participant reaction bias that occurs when Ps attempt to assert their sense of personal freedom by choosing to behave in way they believe to be in opposition to the experimenter's expectations.
threat to internal validity. |
|
Evaluation apprehension
|
the form of participant reaction bias that occurs when Ps attempt to behave in whatever way they think will portray them most favorably. Threat to internal validity.
|
|
Unobtrusive observations
|
observations or measurements that are made secretly without asking Ps any direct questions and without letting them know their behavior is being observed. Helps counteract participant reaction biases (evaluative apprehension)
|
|
Experimenter bias
|
the bias that occurs in research when investigator's expectations about Ps lead to false support for these expectations. Interpret ambiguous behaviors in ways that are consistent with their expectations, actually treating Ps differently in different conditions inducing Ps to behave in hypothesis consistent way
|
|
Double-blind procedure
|
a method of controlling for both participant expectancies and experimenter bias by keeping both research Ps and experimenter unaware of Ps treatment conditions during an experiment
|
|
Instrumentation
|
Problem with the way the researcher measured the DV or the observers behavior
(machinery may drift, human raters may get more reliable with practice, human rates may get bored, the task may be way to easy or too hard) |
|
Selection threat
|
Different criteria used for assigning Ps to conditions
1- first names that end in vowel or consonant: more women's names end in vowels 2- their last names begins with A-M or N-Z: more abundant in certain ethnic groups 3- smokers vs non smokers: smokers are overall less healthy 4- sitting front or back of class: different GPA |
|
Quasi-experiment
|
a research design in which the researcher has only partial control over his or her IVs. Ps are assigned to one or more conditions by some means other than random assignment. Quasi-experiments include both natural experiments that are based on archival data and person-by-tmnt quasi-experiments that are typically conducted in the lab.
can pose internal validity problems because the researchers has only partial control over IV, assigned non-randomly and no control group |
|
Person-by-treatment quasi experiment
|
a research design in which the researcher measures at least on IV and manipulates at least one other IV. Use a lab and random assignment. Always factorial designs.
|
|
Prescreening
|
identifying the Ps that will take part in the person-by-tmnt quasi experiment. Researchers who use this technique give an individual difference measure to a large group of people prior to the time that they run their lab studies. People who receive specific scores, extreme, on the individual difference measure are then recruited to take part in the lab study.
|
|
Extreme groups
|
groups of people taken from the upper and lower end of the distribution of an individual-difference measure, similar to their group but different from the restmore test sensitivity than median split
|
|
Median split
|
an approach to selecting people for inclusion in lab study. Experimenters recruit people who score either the top half or bottom half on an individual difference measure of interest. Worse than extreme groups because a lot of info is lost with the similar people who lie at the cutoff
|
|
Natural experiment
|
a type of quasi-experiment in which the researcher makes use of archival data documenting the consequences of a natural manipulation such as a natural disaster or a change in traffic laws. The best natural experiments typically involve arbitrary or near chance event that affect a large group of people.
|
|
Natural groups with experimental tmnt
|
Rather than exposing each group to both levels of a manipulation, a researcher might expose each group to a different level of a single manipulation. involves taking 2 naturally occurring groups of people and treating them differently in a very precise way.
|
|
Comparability
|
events occur to some people more or less at random, it is not always clear where to find a group of people who are in all other ways exactly like these people.
the fact that it is hard to identify the real comparison group |
|
Patching
|
testing for the influence of specific confounds, a research method in which a researcher adds new conditions to a quasi-experiment to help establish the size of an effect, to test for the influence of conceivable confounds, or both. Addition of new control groups to clarify meaning of any effects
run internal analysis |
|
One-group design
|
a very simple research design in which all the Ps are in a single group that received a natural or experimental manipulation. Pseudoexperiment, it is very difficult to draw conclusions
|
|
One group, pretest-postest design
|
a quasi-experimental research design in which measures are taken from a single group of research P both prior to and after the Ps receive a natural manipulation. Does not include control group so may be alternative explanations.
|
|
posttest-only design with nonequivalent group
|
a quasi-experimental design in which a researcher compares 2 similar but non-identical groups after one and only one of the groups experience a tmnt condition.
- simple experiment without random assignment threat to internal validity, selection |
|
pretest-postest design with nonequivalent groups
|
a quasi-experimental design in which data are collected from 2 presumably comparable groups of research Ps both prior to and after one of the groups receives a natural manipulation.
|
|
Time-series design
|
a quasi-experimental research design in which researchers collect multiple waves of data from 2 presumably comparable groups of Ps. If the 2 groups of P receive highly similar scores on a dependent measure prior to a natural manipulation but begin to receive different scores immediately after the natural manipulation, this strongly suggests that the manipulation may be responsible for the observed differences.
|
|
Internal analysis
|
a form of patching in which a researcher break a group of Ps into one or more subgroups to test for subgroup differences that are consistent with the focal theory or with competing theories.
|
|
Reasons to conduct Quasi-Experiment
|
1- some things that are important to understanding behavior cannot be experimentally manipulated (gender, IQ)
2- some topics are very difficult to study experimentally (unaffordable equipment, resources) 3- Some topics would be unethical to study experimentally (HIV, racial bias) |
|
Single N Designs
|
Quasi-experimental designs that allow you to indentify possible threats to internal validity by measuring the DV multiple times before and after a manipulation
|
|
AB Design/ Interrupted Time Series
|
quasi-experimental design in which a single P or group of Ps is tested repeatedly before and after an intervention
-intervention can be a manipulation or natural event - allows you to establish a stable baseline, pattern wouldve continued without the intervention |
|
ABA Design
|
ABA: Baseline, Treatment, Baseline
Baseline: measuring the behavior of interest in the absence of any tmnt Tmnt: administration of the IV Return of the behavior to pretreatment baseline level is an important characteristic |
|
Three-horned dilemma
|
Research varies on three desirable dimensions: precision, generalizability to situation, generalizability to people.
The dilemma lies in the fact that efforts to maximize any one of these 3 things usually results in an unwanted reduction of at least one of the other things. |
|
Multi-method approach
|
solution for the 3-horned dilemma
employ a variety of different methods, each of which addresses the same research question while operationalizing the IV and DV in different ways. |
|
ABAB/ABABA design
|
can further extend the number of A and B phases in order to further reduce the possibility of threat to validity
|
|
Multiple Baseline Design
|
treatment effect demonstrated by a change in behavior only when treatment is give
A: baseline and B: treatment Different subjects administered different patterns |