• Shuffle
    Toggle On
    Toggle Off
  • Alphabetize
    Toggle On
    Toggle Off
  • Front First
    Toggle On
    Toggle Off
  • Both Sides
    Toggle On
    Toggle Off
  • Read
    Toggle On
    Toggle Off
Reading...
Front

Card Range To Study

through

image

Play button

image

Play button

image

Progress

1/677

Click to flip

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;

677 Cards in this Set

  • Front
  • Back

Two hurdles that get in the way of the research process:

1. Making sure that people are considering the right question. (ensuring that research participants are thinking about the same question that the researcher was thinking about)



2. making sure that once participants have considered the right question, they are able to confer their reaction to the question into a meaningful answer. (ensuring that participants are able to translate their internal psychological state into some kind of value on a response scale)

What are the two mental steps all responders go through when they answer a survey question?

1. judgment phase


2. response translation phase

Judgment phase definition

First step to answering a self-report question where participants determine what question is being asked, and they form some initial response to the question.



-familiar language, pilot testing

In what phase can researchers ensure that most, if not all, participants who respond to a specific self-report question are answering the specific question that the researcher hopes they are answering?

Judgment

In what phase can researchers help people translate their subjective judgments into categories or numbers?

Response translation

Structured self-report questions

Questions that require participants to respond in a specific format.

Keys to translating a hypothetical construct well is to write questions that clearly direct people to the specific psychological state or experience that you wish to measure. How can this be achieved?

1. Ask questions using familiar language that participants will easily understand.


2. Read the research literature to find out how others have solved this problem.


3. If new grounds, use pilot testing.


4. Start with focus groups and open-ended questions.

Pilot Testing

The use of practice studies that are designed to help researchers refine the measures or manipulations they wish to use in the full-blown version of a real study.

Focus Group

Can be used in pilot testing. A small but representative smile of participants from the group a researcher wishes to understand meet together to discuss their experiences. Questions are led by the researcher.

Open-ended Questions

Questions that allow people to respond in their own words.

Two ways open-ended questions can be used:

1. often generate unpredictable reposes hat allow researchers to refine structured ratings scales


2. some qualitative researchers use these questions as their primary source of data and completely steer away from structured questions that require participants to use rating scales.

Problems with open-ended questions:

-Can only be used as data once the answers are coded into the same kind of numbers that participants provide.


-Some participants will provide little or no information about a particular construct.


-Harder to score.


-Introduce judgment error.


-Writing Skills.


-Legibility.

Benefits of open-ended questions:

-Sometimes researchers know so little about a question that they may need to being by asking these questions.


-Useful tools for developing and refining more structured questions.

Uses of pilot testing:

- Comparing open-ended and structured self-report questions.


-Preliminary version of the full-blown study.


-Help researchers develop new self-report questions that truly measure what they are supposed to.

Tips for writing self-report questions that are likely to be as clear as possible:


1. Keep it simple.


2. Use informal language.


3. Avoid negations.


4. Avoid double-barreled questions.


5. Avoid forced-choice items.


6. Avoid questions that do not yield any variance.


7. Avoid loaded questions.


8. Make sure your questions are relevant to everyone in the study.


9. Write multiple questions to assess the same construct.


10. Mix it up.


11. Establish judgmental context.


12. Ease into socially sensitive questions.


13. Ask sensitive questions sensitively.


14. Guarantee anonymity.

How to keep a question simple:

Use words that are familiar to people and use them correctly. Don't ask questions that force people to try to hold a lot of thoughts in their heads.

What type of informal language should be used when writing a question?

Do not use psychological jargon or catchphrases.


Should make comfortable reading for people several grade levels lower than your intended audience.


Why avoid negations when asking questions?

They lead to confusion. People process affirmations more quickly and efficiently than negations. When people process negative statements they sometimes misinterpret them.

What type of questions involving negation should especially be avoided?

Double-negatives.

How to ask questions that require a negation:

try to use words that express it in a familiar way.



Example- use naked instead of "not clothed"

What are double-barreled questions?

A question that asks you to evaluate two different things using a single response.

How should double-barreled questions be resolved?

Break them up into two (or more) separate questions.

Forced choice questions and example:

Questions that ask participants to select one one of two or more options?



Ex. Do you like apples or oranges?

What type of questions do not yield any variance?

Questions that are answered by almost anyone in the same why.

Why avoid questions that do not yield variance?

Researchers only get useful information from participants when they different from one another on the questions asked. If you get no variability in your responses then you cannot use them to predict variability in other responses.

What two things are important to avoid when it comes to avoiding questions that do not yield any variance?

1. Floor effects


2. Ceiling effects

Floor Effects

Occur when almost everyone in a sample responds at the same low level on a question or dependent measure.

Ceiling Effects

Occur when almost everyone in a sample responds at the same high level on a question or dependent measure.

What are floor and ceiling effects examples of?

Restriction of range.

Restriction of Range

Occurs anytime people's scores on a measure have little or no variation.

How to avoid using questions that do not yield any variance?

Know your sample and adjust your questions, and sometimes your response scales, accordingly.

How to minimize the problem of asking loaded questions?

Writing questions in ways that do not indicate which response the researcher considers most desirable.

What is the problem with not having questions that are relevant to everyone in the study?

If it ignores the thoughts and feelings of a significant portion of your sample, that portion is unlikely to provide useful data.

How to establish a judgmental context:

1. Include instructions that establish the appropriate context.


2. To be sure that participants who have not read their instructions carefully still have a good sense of the appropriate judgmental context. (Researchers can include warm-up or practice items at the beginning of the survey.)

Ways to guarantee anonymity:

Tell them that you have gone to great lengths to preserve the anonymity of their responses. Remind them in writing and verbally that you are not interested in the responses of any single person, that they should not include any personally identifying information, and that their responses will only been seen my researchers.

Three issues to consider when designing a numerical rating scale:

1. How many numbers to use.


2. What anchors to build into your numeric scale.


3. The best numbering system.

What does a good response scale do:

allows participants to express a wide range of opinions.

What is the problem with including too many numbers on a scale and the problem with too few?

May create confusion, judgment is too refined, asks you to split judgmental hairs.



Crude, does not allow you to give any detail about your life satisfaction.

What is the optimal number to use for scales?

between 3 and 10

Anchors

Adjectives that lend meaning to the numbers on a scale.



Equal Appearing Intervals

A scale has this whenever the psychological distance implied by a single-unit difference on the rating scale remains constant across the entire range of the scale.

Bipolar scale

asks respondents to rate a quantity that deviates in both directions for a zero point.



ex. -5 to 5

Unipolar scale

asks respondents to make ratings on dimensions that begin at some very low value (often zero) and move upward to a subjective maximum point on the dimension of interest.



ex. 0-7

EGWA Scale

Empirically Grounded, Well-Anchored response scale. A very simple rating scale that can be see in many different research situations.

Five Steps to Designing Questions:

1. Step back and think. (what do you want to measure? talk to other people, look up previous measures)



2. Write lots and lots of questions. (for an 8 item measure you might want to start with 40 questions. you'll eventually abandon most of them)



3. Analyze your scale and derive the best items. (Give your questions to a large sample of people similar to those you intend to study. Make sure they are reliable.



4. Administer and Factor Analyze- do they correlate?



5. Validity- does it correlate with other established questionnaires?

Absolute scales

Report magnitude in a direct and obvious way.


Raises important questions about how well people can remember their own behavior, but it does not usually require researchers to go through all the other machinations.

The Semantic Differential

The majority of adjectives in language fall nearly into one of three different categories that can be used to evaluate almost any object in almost any culture. Adjectives of evaluations, potency, and activity.

Thurstone Scale

Use a scale that shows a spectrum. Scales are designed so that different items imply different levels of evaluation. They then compute scores that reveal whether a rater most strongly endorses the items that conveys low, medium, or high evaluations.

Guttman Scale

Once you say no to an item you are expected to continue saying no to similar items.

What do the Thurstone and Guttman Scales do?

Use scoring procedures that reveal which items are being endorsed and which items are not. They can thus learn more about the people than they would learn if they only focused on computing a scale total.

Problems with Thurstone and Guttman Scales?

1. Extremely difficult to design.


2. Both can produce some complex scoring issues. (what happens if people answer on both ends of the spectrum but not the middle?)

4 types of Measurement:

Direct Questioning


Observational Methods


Collateral Reports


Physiological Measures



Pscyhometrics

The study of psychological measurement.

Memory Telescoping

Tendency to recall events as more recent than their actual dates.

Types of Direct Questioning

Self Report


-paper and pencil


-face to face interviews


-telephone or internet interviews


-experience sampling (diary methods)

Collateral Reports

Third party responses to a questionnaire or interview.

Observational Measures

Direct observation of behavior

Physiological measures

Internal processes that are not directly observable.

Three threats to the validity of a research study:

1. People are different.


2. People change.


3. The process of studying people changes people.

Pseudo-experiment

a research design in which someone tests a claim about a variable by exposing people to the variable of interest and noting that these people feel, think, or behave as expected.

Problem with Pseudo-experiments:

-There is no control group, so we have no idea what would have happened if the same person had engaged in the same behavior without being exposed to the variable of interest.


-Threat to internal validity.

Problems with "People are Different"

1. Individual differences and third variables


2. Selection bias and nonresponse bias


What type of validity is "People are different" a threat to? And example.

Internal and external

Selection bias

Sampling people from an unrepresentative sample.

Nonresponse bias

closely related to selection bias except that in this case, the respondents are the source of the bias.

Problems with Nonresponse Bias-

People who choose to answer surveys are systematically different from people who choose not to do so, therefore low response rates may yield information that is highly misleading.



Threat to external validity.

Way to reduce Nonresponse bias:

Weight individual respondents based on known population values.


Problems with Selection Bias

Threatens external validity but not internal.



Knowing that a cause and effect relationship is present for a given sample does not offer assurances that it generalizes to other populations.

Problems with "People Change"

1. History and Maturation


2. Regression Toward the Mean



If people change for reasons that have nothing to do with a researcher's treatment, then these changes can lead to some inappropriate conclusions.

What threat does history and maturation present?

Threat to internal validity which is likely to pose a problem when a researcher conducts a pretest-posttest study.



Pose problems even when a researcher attempts to control for individual differences by assessing the responses of the same person before and after treatment.

History

Changes that occur more or less across the board in a very large group of people such as a nation or culture.

Maturation

Specific developmental or experiential changes that occur in a particular person, or a particular age cohort, over time.

How to correct problem of history and maturation?

Conduct an experiment where individuals are randomly assigned to an experimental condition or a control condition, then compare.

What is regression toward the mean 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.

True Score

an underlying ability or trait that the observed score presumably reflects

Observed Score

influenced not only by true scores but also error

How to separate regression toward the mean from an experimental treatment:

Divide the group in half at random. Make one the experimental group and the other the control group. Then compare.

Hawthorne Effect

Increases in productivity that occur when workers know they are being studied.

Mere measurement effect:

the tendency for participants to change their behavior simply because they have been asked how they will act in the future.

What is the problem with the mere measurement effect?

threatens external validity. If a study documents an effect on behavior it is possible that the result would not generalize to people who have not been in the study

Problems with "the process of studying people changes people":

1. Testing Effects


2. Experimental Mortality


3. Participant Reaction Bias


4. Experimenter Bias


Testing effects:

problem in pretest-posttest designs that have no contra group. the tendency for most participants to perform better on a test or personality measure the second time they take it.

What type of validity is testing effects a threat to?

Internal

Attitude polarization

allowing people to give a little thought to their attitudes often leads them to become more extreme in these attitudes

Ways to correct the problem of testing effects:

1. Conduct a true experiment with a pretested control group. Will not eliminate the test effects but will allow you to separate them from your experimental treatment.


2.Eliminate the pretest. Half would receive an experimental intervention- without measuring anyone prior.


3. Waiting as long as possible.

Experimental Mortality

The failure of some of the participants in an experiment to complete the study.

What type of validity can experimental mortality cause a threat to?

1. internal


2. external


3. both

What is necessary to maintain internal validity when there is experimental mortality?

The proportion of the people who dropped out of the study is the same int he two different conditions.

Homogeneous attrition and what type of validity it is a threat to

there is an equal level of attrition across all of the experimental conditions



external validity

Heterogeneous attrition and what type of validity it is a threat to

Occurs when the attrition rates in two or more conditions of an experiment are noticeably different.



internal validity

How to reduce problems with mortality?

1. Do anything to get people to not drop form the study- communicate the importance, warn people ahead of time about what they should expect, schedule breaks, offer rewards


2. expose both groups to the same conditions, one before (experimental) and one after (control)


Participant reaction bias and type of validity it is a threat to:

Bias that occurs when people realize they are being studied and behave in ways that they normally wouldn't.



internal

Types of participant reaction bias

1. People may try to do what they think the researcher expects.


2. The opposite of what they think the researcher expects.


3. Whatever will make them look good.

Participant expectancies

occur when participants try to behave in ways they believe to be consistent with the experimenter's hypothesis.

Demand characteristics

characteristics of an experiment that subtly suggest how people are expected to behave

Participant reactance

the tendency of participants to try to disconfirm an experimenter's hypothesis

Evaluation apprehension

people's concerns about being judged favorably or unfavorably by another person causing people to do whatever they expect will portray them in a favorable light.

How to avoid participant reaction bias:

1. assure anonymity


2. give participants the same expectancy in both groups that has little or nothing to do with the real predictions


3. cover story


4. keeping participants int he dark


5. convince participants that researchers can read their minds


6. Make use of indirect measures of people's attitudes and opinions

cover story

a false and often elaborate story about the nature and purpose of the study

Unobtrusive observations

secret observations where participants do not realize that they are being studied.

Key to the success of indirect measures of attitudes?

when participants fill out questionnaires containing indirect measures of attitudes, they don't realize what is actually being measured

Experimenter Bias and two forms

When experimenters' expectations about their studies bias their experimental observations.



1. researchers make biased observations in an experiment.


2. treat their participants differently based on expectations

Confound

Used to identify any situation in which some additional variable (a) varies systematically with the independent variable and (b) also varies systematically with the dependent variable.



The researcher hasn't thought about or couldn't control but could lead to a false association between an independent and dependent variable.

What type of validity are confounds a threat to?

Internal

How to eliminate confounds?

Use clean and well-established manipulations, confusing replication studies in which you use different manipulations of the same basic construct, and measuring and controlling for confounds using statistical techniques.

Artifact

A variable that is held constant in a study or series of studies and that might represent a redistricted context under which the effect with be observed.

What type of validity do artifacts threaten?

External

Example of a confound

People with tattoos die sooner than those without, must be the tattoos.

Example of an artifact

Most college students engage in risky drinking however the questions in the study caused students to think that heavy drinking, and this caused them to drink more

Quasi–Experiment
Research designs in which researchers have only partial control over their independent variables.

Participants are assigned to one or more conditions in a study by some means other than random assignment.
What type of validity can quasi–experiments pose problems for?
Internal
Why use quasi–experiments?
1. you can't randomly assign or manipulate things like gender
2. cost (pricy equipment)
3. ethical reasons (manipulating diseases, forcing someone to smoke)
Person–by–treatment quasi–experiment
Designs in which the researcher maintains partial control over the world by measuring at least one independent variable and manipulates at least one other independent variable.
Pre–screening
Screening people on an individual–difference measure prior to running a lab study.
Extreme groups
Groups of people taken from the upper and lower ends of the distribution of an individual–difference measure.
What does the use of extreme groups accomplish?
They create categories of participants who closely resemble the other members of their own groups on the dimension of interest but who differ greatly from the members of the other group. The two groups are qualitatively different from one another.
Two possible pretest splits:
1. extreme groups
2. median split
Median split
Label those above the median as high and those below it as low.
Problem with median split:
People who fall very near the cutoff score for the two groups are much more similar to one another than they are to the extreme members of their own groups.
What is reducing power (with the use of a median split) preferable to?
1. drawing inferences based on very small numbers of people
2. allowing one or two people with really extreme scores to play a disproportionate role in the findings
Test sensitivity
Statistical power to deter real effects if they exist.
What type of pretest method generally finds real effects?
Extreme groups method
What do researchers conducting a person–by–treatment quasi–experiment expect to find?
Different persons to respond differently to different treatments (experimental manipulations).
Self–verification theory
A person's preference for positive or negative feedback should depend not only on the favorability of the feedback but also upon the self–concept of the person receiving the feedback.
Problem of induction
It is possible that researchers will correct for a large number of important confounds but fail to consider the one confound that is the true cause of a commonly observed effect.
Natural experiments
An experiment where there is no use of true random assignment at all and there is no experimental control. They involve naturally occurring manipulations. Often random life events that were caused by factors that had nothing to do with the people who experienced them (house destroyed by earthquake).
A way to gain control over an independent variable in a natural experiment:
measure any differences that do exist between people exposed to two different levels of a manipulation (measure any existing confounds)
What type of studies often qualify as natural experiments?
archival
Natural groups with experimental treatment design:
Opposite of person–by–treatment.

Involves taking two naturally occurring groups of people and treating them differently in a very precise way.
Natural groups with experimental treatment design yield useful data only when the two groups:
1. are extremely similar in important ways to begin with
2. are subsequently exposed to different levels of an experimental treatment, without otherwise being treated differently
Nature–by–treatment study:
Combination of natural and laboratory manipulations, or by assessing individual differences among people and then waiting for them to experience truly arbitrary natural events.
What is very important to the success of a quasi–experiment?
comparability
What is a way to identify an ideal comparison group in many quasi–experiments?
patching
If you want to show that an IV is influencing a DV you must _________. This requires ______.
hold everything else but the independent variable constant across experimental conditions; random assignment and procedural control
Random assignment deals with _____ confounds and procedural control deal with ____ confounds.
person; operational
Patching
Occurs when a researcher adds new conditions to a study to help establish the size of a quasi–experimental effect, to test for the influence of conceivable confounds, or both.
Patched–up designs:
Occur when researchers continually add control groups to a quasi–experimental design
Outcome of a carefully patched up study:
often resemble a patchwork quilt, with many different conditions sewn in to the original design, each to deal with specific concerns.
What is the researcher responsible for when using patching?
Thinking of all the possible control groups that should be included to help clarify the exact meaning of any effects that are found.

Focus attention not only on the groups or conditions that are of interest but also on the groups or conditions that are not of interest.
Example of a patched–up study:
The study guide quasi–experiment.
One–group design
All participants are in one group, pseudo experiment, no comparison
Treatment confound example from study guide quasi–experiment
Two of the same class but taught at different times might make the groups innately different based on the students who choose an early morning class v. an afternoon class.
Time–series designs
researchers look at long runs of data to show equivalence prior to a quasi manipulation
one group design study guide example
give class study guide for final, take average score
one group, pretest–posttest design study guide example
test midterm, give study guide, test final
posttest–only design with nonequivalent groups study guide example
take average grade from previous class without study guide, and compare to current class with study guide
pretest–posttest design with nonequivalent groups study guide example
compare midterm and final grades from two classes, see if midterm are the same but final are different
Comparative time–series design study guide example
look at quizzes from before the study guide and after
internal analysis
Break one or more groups into additional subgroups to test for differences that are consistent with the focal theory or with competing theories
internal analysis study guide example
Come up with two theories– one that study guides are just morale boosters, two that study guides help student study more fictively. Check to see who uses to the study guide and who doesn't.
Name–letter effect
The finding that people like letters that occur in their own names quite a bit more than they like letters that do not.
Implicit egoism
If people prefer things that they associate with the self, then by extension people should prefer people, places, and occupations whose names or titles resemble their own first or last names
Quasi–experiments' advantages over true experiments:
1. They can determine whether effects observed in the lab generalize to the real world
2. useful for testing basic theories
Image appeal
A persuasion technique designed to change a person's attitudes or beliefs by capitalizing on people's desire to adopt positive images and a avoid negative images.
Coaches influences the self–esteem of young players. One league is taught behavioral coaching principles and the other league is the control group. Is this an experiment or a quasi–experiment? Why?
Quasi–experiment, because coaches were not randomly assigned to be trained or in the control group.
Two ways to devise a study on self–esteem:
1. bogus feedback (tell someone they are good or bad at something)
2. self–esteem questionnaire (use to determine levels of self esteem)
Bogus feedback is a way to start a(n) __________ on self–esteem while a self–esteem questionnaire is a way to start a(n)___________.
Experiment; quasi–experiment
How did experimenters end up nailing the tobacco company without using human experiments?
They did experiments on monkeys.
Main problem with Pearson–by–Treatment quasi–experiment:
person confounds
EXAMPLE–

IVs – mood (from bogus feedback), depressed or not (based on questionnaire)
DV– explanatory style (do you blame yourself or someone else).

What type of study is this? Why?
Person–by–treatment; Mood is manipulated, depression is measured
EXAMPLE–

Study on those who won the lottery. Groups are predetermined by those who won the lottery and those who did not.

What type of study is this?
Natural Experiment
EXAMPLE–

You want to assess a new way to give incentives to workers. Firemen and police are used because they are in different but similar groups.

The new system is used with the police and not with the firemen. Productivity is measured for the police and the control group (firemen).

What type of study is this?
Natural Groups with Experimental Treatment
Ceteris paribus
All else being equal
Possible confounds for a one–group, pretest–posttest design
1. overall improvement
2. experimenter bias
3. got used to a teaching style
4. testing effect
EXAMPLE–

No–fault divorce laws were seen as being a problem for the institution of marriage. People believed that more people would get divorced if divorce was easier. This test ____________ concluded that this was the case.

Why was this not the case?
One group, pretest–posttest design.

Although divorce rates increased after the no–fault laws were introduced, they did not increase any more than the general trend. The test did not show this.
Four common threats to internal validity in design that study one group before and after the quasi–indepdent variable:
1. history
2. maturation
3. regression to the mean
4. pretest sensitization
Comparative time–series design
researchers look at a long run of data to show equivalence prior to quasi–manipulation
Two differences between true experiments and quasi–experiments:
1. manipulation and prescreening are not always the same and won't always yield the same results
2. experimental contamination sometimes drives researchers to prefer quasi–experiments even when a true experiment is feasible and moral
EXAMPLE–

Plants are given to people in a nursing home. One group (on one floor) was asked to take care of the plant, and the other group (on a different floor) was told that other people would take care of the plant. Those asked to take care of the plant were happier and more active than the other group.

Is this a true experiment or quasi–experiment?
Quasi–experiment because there was not random assignment.
Do quasi–experiments have time sequence, covariation, and/or randomization or experimental control?
yes;yes;no
How many IVs do one–way designs have? How many levels are there?
One; one
How many IVs do two–groups designs have? How many levels are there?
One; two
Two groups design
Often consists of an experimental group and a control group.
How many IVs in one–way, multiple groups designs? How many levels?
One; three or more
Physical attractiveness stereotype
Most people assume that physically attractive people possess a specific set of personality characteristics, most of which are favorable.
Problems with multiple–group designs:
It almost always takes more time, more resources, and more research participants to conduct experiments that have more than two conditions.
Major limitation of one–way designs:
Allow researchers to look at only one IV at a time.
Who popularized the use of factorial designs?
R.A. Fisher
Factorial designs
designs that contain two or more IVs that are completely crossed that allow researchers to answer questions about more than one IV at the same time
What does it mean to say "two IVs are completely crossed"?
Every level of every IV appears in combination with every level of every other IV.
The label a researcher gives a factorial design specifies both ____ and _____.
1. how many IVs exist in the design
2. how many levels of each IV exists in the design
How many IVs and levels exist for the following factorial designs?

1. 2 X 2
2. 2 X 3
3. 2 X 2 X 3
4. 2 X 4
5. 2 X 2 X 2 X 3
1. 2; 2 and 2
2. 2; 2 and 3
3. 3; 2, 2 and 3
4. 2; 2 and 4
5. 4; 2, 2, 2, and 3
Interaction occurs when...
the effect of one IV differs depending on the level of a second IV.

They are always about two or more variables.
Main effects
The simple, straightforward effects of IVs in factorial studies that has to do with only one variable. The overall effect of an IV, averaging across all levels of the other IV(s).
To determine whether your factorial study yielded a main effect of a specific variable you....
collapse across all the levels of all other IVs in a factorial design.
Factorial designs allow researchers to detect the presence of
statistical interactions
Ordinal or spreading interaction
The observed pattern when an effect exists at one level of a second IV but is weaker or nonexistent at a different level of the second IV.

In this interaction the lines DO NOT cross.
Disordinal or crossover interaction
Occurs when there are no main effects of either IV and when the effects of each IV are opposite at different levels of the other IVs

In this interaction the lines DO cross.
Nonparallel lines on a graph for a factorial study indicate:
that there is an interaction
Crossover interaction
Nonparallel lines cross over one another at or very near the middle of two lines.
Simple effects tests
A set of follow–up tests that are conducted when the statistical analysis in a factorial design yields a significant interaction.

Used to see which specific mean comparisons are significant in a factorial story.

Clarify the precise nature of the interaction.
In a true crossover interaction simple effects tests will be ____.
significant
Three advantages of factorial designs according to Fisher:
1. they are more efficient (fewer participants, less noise)
2. more comprehensive
3. and produce more external validity than one–way designs
Factorial designs are efficient in the sense that they...
allow us to look for more than one main effect at a time in a single study
Factorial designs are comprehensive in the sense that...
unlike one–way designs, they tell us more of the whole story behind a specific phenomenon by allowing us to see how different variable work together to influence the phenomenon
Explain the relationship between factorial designs and external validity.
A researcher who observes main effects in a factorial study can be reasonably sure that any observed main effects will generalize across whatever levels of the other IVs exist in the study.
Between–subjects designs
Designs in which each participant serves in one and only one condition of the experiment.
Within–subject designs
Each participant serves in more than one (perhaps all) of the conditions of a study.
What is this an example of:

There are two groups. One that shows squares that are the same color but appear different, and one that shows them appearing to be the same color.

Five participants are randomly assigned to rate the darkness of square A and five other participants rate the darkness of square B on a scale of 1–10.
a between–subjects design
Advantages of within subject designs:
1. Require fewer participants, especially when it comes to complex factorial designs
2. Eliminate person confounds
Disadvantages of within subject designs:
introduce biases that are unlikely to emerge in between–subject designs:

1. sequence effect (getting tired)
2. carryover effect (response to stimulus that influences response to a second)
a. order effect (order questions are asked)
b. practice effect
c. interference effect
3. demand characteristics– increase the likelihood that participants will be able to figure out an experimenter's hypothesis
Sequence effect
occur when the simple passage of time begins to take its toll on people's responses (bored or tired)
Carryover effects
Occur when people's responses to one stimulus in a study directly influence their responses to a second stimulus.
Order effects
Type of carryover effects that occur when a question takes on a different meaning when it follows one question than when it follows another.
Practice effects
Type of carryover effects that occur when participants' experience with one task makes it easier for them to perform a different task that comes along later.
Interference effects
Type of carryover effects that occur when performing one task disrupts people's performance on a second task.
Demand characteristics and evaluation apprehension cause...
participants to try to please the experiment and participants to report whatever they think portrays them in the most favorable light possible
Solutions to within–subject design drawbacks:
Counterbalancing
1. complete counterbalancing
2. incomplete counterbalancing
a reverse counterbalancing
b partial counterbalancing
3. Latin Square
Counterbalancing
Method of control where the researcher varies the order in which participants experience the different kinds of within–subjects studies
Complete counterbalancing
All of the possible treatment sequences are used the same number of times. Each goes through one of the sequences.

Example– Treatments A, B, and C. Six possible sequences ABC, BCA, CAB, CBA, BAC, and ACB.
Incomplete counterbalancing is used when
it is impractical to use complete
Reverse counterbalancing
All treatments are presented to each subject twice. First in one order then in another.

ABBA or ABCCBA order––When subjects experience conditions more than once, they first experience the conditions in one order, and then the reverse order.
Partial counterbalancing
Only some of the possible condition sequences are used because there are more possible sequences than there are subjects.
Latin Square

Procedure that generates fixed number of orders to balance a within–subjects study.

Example

1– A B C D
2– B A D C
3– C D A B
4– D C B A

Latin Squares could be used to _______________, such that each of them appears _______ and ________.
generate four orders; exactly one in each possible serial position; exactly twice before and exactly twice after each of the other three unique conditions
Counterbalancing does not __________ but it does _________. The effects are ________.
erase sequence or carryover effects; unconfound them with particular treatment conditions; balanced
Structured debriefing
Interview conducted with participants immediately after they have completed a study to determine what they thought the researcher expected to find.
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.
Numerosity heuristic
The tendency to estimate quantity or magnitude by basing one's judgments disproportionately on the number of units into which a stimulus is divided– without fully adjusting for other important variables.
Which study in our class uses the two–groups design?
Study 2– warm v cold
The toilet goer's experiment using no confederate, a confederate present one urinal away, and a confederate in the next urinal is an example of?
A one–way three–groups design
Taxonomy
1. Overall number of numbers– how many IVs exist
2. Numbers in the formula– how many levels of each IV exist
Cells
Number of unique conditions of the study
How many cells in a 2 X 2 and 2 X 3 study?
4; 6
Study 3

1. What are the IVs?
2. How many levels are in each IV?
3. Taxonomy and # of cells?
4. What is the DV?
5. What is the design?
6. experiment or quasi?
1. IVs– Valance (positive or negative traits), Mood (positive or negative mood)
2. 2
3. 2 X 2; 4
4.# of traits recalled
5. factorial, mixed–model
6. experiment
Study 3

1. What is the hypothesis?
2. What is the main effect of A?
3. What is the main effect of B?
4. What is the interaction?
1. If you are placed in a positive mood you will recall more positive traits.
2. Valance– negative traits are remembered more than positive traits
3. Mood ??
4. When placed in a positive mood we will remember more positive traits. When placed in a negative mood we will remember more negative traits.
EXAMPLE–

Researchers approach members of other sex on college campus and ask them to have sex.

Three questions, would you go out wit me tonight, would you come over, would you go to bed with me?

Ask two groups, male and female

1. What are the IVs?
2. How many levels?
3. What is the DV?
4. Experiment or quasi–experiment?
1. type of question, gender
2. 3, 2
3. saying yes
4. quasi because you can't manipulate gender
Why are complex designs (2 X 2 X 2 X 2) not popular?
1. You need a lot of participants
2. qualification (unnecessary to have more than two IVs usually)
3. hard to understand
Example–

Rat, fetal alcohol syndrome, short term and long term mental effects from rats who drink during pregnancy and those who don't.

IVs?
DV?
Taxonomy?
Is this an experiment?
IV– Maternal diet, age of rate
DV– passive avoidance
Tax– 2 x 2
Yes
In rat experiment:
What is Factor A?
What is Factor B?
What is the AB interaction?
1. addresses whether maternal diet affects PA learning
2. addresses whether age is related to PA learning
3. addresses whether the effects of maternal diet depend on the age of the offspring
How to compute marginals?
Add up the row or column, then average.
How to find main effect?
Look at marginals, if they are the same there is no main effect.
What type of interaction is involved with study 3?
Crossover
A main effect is qualified by the interaction when?
The interaction explains something more (that changes the meaning) that the main effect does not.
Graph when main effect of factor X and interaction is significant:
There is one straight line and one line that crosses through it
Graph of a interaction and both main effects:
//fce-study.netdna-ssl.com/2/images/upload-flashcards/84/19/18/7841918_m.jpg
EXAMPLE–

In an experiment one group of participants performed a motor task with and without practice. Another group performed a spatial task with and without practice.

The research design is:

1. 2 X 2 between–subjects
2. 2 X 3 mixed model
3. 2 X 2 within subjects
4. 2 X 2 mixed model
5. 2 X 3 between subjects

The between–subject factor is _____; the within subject factor is ______
4; task type; practice
Stapel did what?
Fabricated findings for train research and vegetarian research.
Journals with ______ are more likely to publish retractions that those with ________.
high–impact factors, lower impact factors
Approaches to ethical decisions:
1. Deontology
2. Ethical Skepticism
3. Utilitarian
Deontology
Ethics should be judged in light of a universal moral code.
Ethical Skepticism
Rules are relative to culture and place. The final arbitrator is the researcher's conscience.
Utilitarian
The potential benefit should be weighed against the potential costs.
APA Ethical Principles of Psychologists and Code of Conduct Potential Benefits
1. basic knowledge
2. improvements of research or assessment techniques
3. practical outcome
4. benefits for researchers
5. benefits for research participants
APA Ethical Principles of Psychologists and Code of Conduct Potential Costs
1. time, effort, money
2. harm (psychological or otherwise)
3. confidentiality compromised
What does the IRB do?
Decides if an experiment is ethical before it is begun.
What members are on the IRB?
1. Nonscientific disciplines.
2. At least one member that is a representative of the local community.
What is required of IRB?
1. Written proposal is submitted
2. research cannot be conducted without prior approval
Six crucial aspects of ethics:
1. informed consent
2. invasion of privacy
3. coercion to participate
4. physical and mental stress
5. deception
6. confidentiality
Informed consent requirements:
Must notify people of what is being studied and what will happen.

Must include
–simple langauge, signature, potential risks and benefits, penalties in case of withdrawal, special considerations for kids
What research is allowed without informed consent?
1. minimal risk researcher (nothing more than what happens in daily life)
2. will not impact the rights and welfare of the participants
3. cannot feasibly be carried out if required
Examples of studies that are or might be an invasion of privacy?
1. toilet study
2. asking for sex study
Example of a study that might cause physical or mental stress?
Simulated rape study
Kinds of deception
Active and passive
How can deception be used?
–Experimental confederates
–provide false feedback
–present two related studies as unrelated
–give incorrect info regarding stimulus materials (placebo)
–double deception
Criticism of deception
1. moral ground
2. participants lose their regard to science
3. perseverance effects
How to resolve problems of deception?
debriefing
Debriefing
explaining the nature of the study, the motives, obtain reactions, and show appreciation
Confidentiality means
the data may only be used for purposes of research and should be obtained by maintaining anonymity
Quasi–Experiment

Research designs in which researchers have only partial control over their independent variables.

Participants are assigned to one or more conditions in a study by some means other than random assignment.

What type of validity can quasi–experiments pose problems for?
Internal
Why use quasi–experiments?
1. you can't randomly assign or manipulate things like gender
2. cost (pricy equipment)
3. ethical reasons (manipulating diseases, forcing someone to smoke)
Person–by–treatment quasi–experiment
Designs in which the researcher maintains partial control over the world by measuring at least one independent variable and manipulates at least one other independent variable.
Pre–screening
Screening people on an individual–difference measure prior to running a lab study.
Extreme groups
Groups of people taken from the upper and lower ends of the distribution of an individual–difference measure.
What does the use of extreme groups accomplish?
They create categories of participants who closely resemble the other members of their own groups on the dimension of interest but who differ greatly from the members of the other group. The two groups are qualitatively different from one another.
Two possible pretest splits:
1. extreme groups
2. median split
Median split
Label those above the median as high and those below it as low.
Problem with median split:
People who fall very near the cutoff score for the two groups are much more similar to one another than they are to the extreme members of their own groups.
What is reducing power (with the use of a median split) preferable to?
1. drawing inferences based on very small numbers of people
2. allowing one or two people with really extreme scores to play a disproportionate role in the findings
Test sensitivity
Statistical power to deter real effects if they exist.
What type of pretest method generally finds real effects?
Extreme groups method
What do researchers conducting a person–by–treatment quasi–experiment expect to find?
Different persons to respond differently to different treatments (experimental manipulations).
Self–verification theory
A person's preference for positive or negative feedback should depend not only on the favorability of the feedback but also upon the self–concept of the person receiving the feedback.
Problem of induction
It is possible that researchers will correct for a large number of important confounds but fail to consider the one confound that is the true cause of a commonly observed effect.
Natural experiments
An experiment where there is no use of true random assignment at all and there is no experimental control. They involve naturally occurring manipulations. Often random life events that were caused by factors that had nothing to do with the people who experienced them (house destroyed by earthquake).
A way to gain control over an independent variable in a natural experiment:
measure any differences that do exist between people exposed to two different levels of a manipulation (measure any existing confounds)
What type of studies often qualify as natural experiments?
archival
Natural groups with experimental treatment design:
Opposite of person–by–treatment.

Involves taking two naturally occurring groups of people and treating them differently in a very precise way.
Natural groups with experimental treatment design yield useful data only when the two groups:
1. are extremely similar in important ways to begin with
2. are subsequently exposed to different levels of an experimental treatment, without otherwise being treated differently
Nature–by–treatment study:
Combination of natural and laboratory manipulations, or by assessing individual differences among people and then waiting for them to experience truly arbitrary natural events.
What is very important to the success of a quasi–experiment?

comparability

What is a way to identify an ideal comparison group in many quasi–experiments?
patching
If you want to show that an IV is influencing a DV you must _________. This requires ______.
hold everything else but the independent variable constant across experimental conditions; random assignment and procedural control
Random assignment deals with _____ confounds and procedural control deal with ____ confounds.
person; operational
Patching
Occurs when a researcher adds new conditions to a study to help establish the size of a quasi–experimental effect, to test for the influence of conceivable confounds, or both.
Patched–up designs:
Occur when researchers continually add control groups to a quasi–experimental design
Outcome of a carefully patched up study:
often resemble a patchwork quilt, with many different conditions sewn in to the original design, each to deal with specific concerns.
What is the researcher responsible for when using patching?
Thinking of all the possible control groups that should be included to help clarify the exact meaning of any effects that are found.

Focus attention not only on the groups or conditions that are of interest but also on the groups or conditions that are not of interest.
Example of a patched–up study:
The study guide quasi–experiment.
One–group design
All participants are in one group, pseudo experiment, no comparison
Treatment confound example from study guide quasi–experiment
Two of the same class but taught at different times might make the groups innately different based on the students who choose an early morning class v. an afternoon class.
Time–series designs
researchers look at long runs of data to show equivalence prior to a quasi manipulation
one group design study guide example
give class study guide for final, take average score
one group, pretest–posttest design study guide example
test midterm, give study guide, test final
posttest–only design with nonequivalent groups study guide example
take average grade from previous class without study guide, and compare to current class with study guide
pretest–posttest design with nonequivalent groups study guide example
compare midterm and final grades from two classes, see if midterm are the same but final are different
Comparative time–series design study guide example
look at quizzes from before the study guide and after
internal analysis
Break one or more groups into additional subgroups to test for differences that are consistent with the focal theory or with competing theories
internal analysis study guide example
Come up with two theories– one that study guides are just morale boosters, two that study guides help student study more fictively. Check to see who uses to the study guide and who doesn't.
Name–letter effect
The finding that people like letters that occur in their own names quite a bit more than they like letters that do not.
Implicit egoism
If people prefer things that they associate with the self, then by extension people should prefer people, places, and occupations whose names or titles resemble their own first or last names
Quasi–experiments' advantages over true experiments:
1. They can determine whether effects observed in the lab generalize to the real world
2. useful for testing basic theories
Image appeal
A persuasion technique designed to change a person's attitudes or beliefs by capitalizing on people's desire to adopt positive images and a avoid negative images.
Coaches influences the self–esteem of young players. One league is taught behavioral coaching principles and the other league is the control group. Is this an experiment or a quasi–experiment? Why?
Quasi–experiment, because coaches were not randomly assigned to be trained or in the control group.
Two ways to devise a study on self–esteem:
1. bogus feedback (tell someone they are good or bad at something)
2. self–esteem questionnaire (use to determine levels of self esteem)
Bogus feedback is a way to start a(n) __________ on self–esteem while a self–esteem questionnaire is a way to start a(n)___________.
Experiment; quasi–experiment
How did experimenters end up nailing the tobacco company without using human experiments?
They did experiments on monkeys.
Main problem with Pearson–by–Treatment quasi–experiment:
person confounds
EXAMPLE–

IVs – mood (from bogus feedback), depressed or not (based on questionnaire)
DV– explanatory style (do you blame yourself or someone else).

What type of study is this? Why?
Person–by–treatment; Mood is manipulated, depression is measured
EXAMPLE–

Study on those who won the lottery. Groups are predetermined by those who won the lottery and those who did not.

What type of study is this?
Natural Experiment
EXAMPLE–

You want to assess a new way to give incentives to workers. Firemen and police are used because they are in different but similar groups.

The new system is used with the police and not with the firemen. Productivity is measured for the police and the control group (firemen).

What type of study is this?
Natural Groups with Experimental Treatment
Ceteris paribus
All else being equal
Possible confounds for a one–group, pretest–posttest design
1. overall improvement
2. experimenter bias
3. got used to a teaching style
4. testing effect
EXAMPLE–

No–fault divorce laws were seen as being a problem for the institution of marriage. People believed that more people would get divorced if divorce was easier. This test ____________ concluded that this was the case.

Why was this not the case?
One group, pretest–posttest design.

Although divorce rates increased after the no–fault laws were introduced, they did not increase any more than the general trend. The test did not show this.
Four common threats to internal validity in design that study one group before and after the quasi–indepdent variable:
1. history
2. maturation
3. regression to the mean
4. pretest sensitization
Comparative time–series design
researchers look at a long run of data to show equivalence prior to quasi–manipulation
Two differences between true experiments and quasi–experiments:
1. manipulation and prescreening are not always the same and won't always yield the same results
2. experimental contamination sometimes drives researchers to prefer quasi–experiments even when a true experiment is feasible and moral
EXAMPLE–

Plants are given to people in a nursing home. One group (on one floor) was asked to take care of the plant, and the other group (on a different floor) was told that other people would take care of the plant. Those asked to take care of the plant were happier and more active than the other group.

Is this a true experiment or quasi–experiment?
Quasi–experiment because there was not random assignment.
Do quasi–experiments have time sequence, covariation, and/or randomization or experimental control?
yes;yes;no
How many IVs do one–way designs have? How many levels are there?
One; one
How many IVs do two–groups designs have? How many levels are there?
One; two
Two groups design
Often consists of an experimental group and a control group.
How many IVs in one–way, multiple groups designs? How many levels?
One; three or more
Physical attractiveness stereotype
Most people assume that physically attractive people possess a specific set of personality characteristics, most of which are favorable.
Problems with multiple–group designs:
It almost always takes more time, more resources, and more research participants to conduct experiments that have more than two conditions.
Major limitation of one–way designs:
Allow researchers to look at only one IV at a time.
Who popularized the use of factorial designs?
R.A. Fisher
Factorial designs
designs that contain two or more IVs that are completely crossed that allow researchers to answer questions about more than one IV at the same time
What does it mean to say "two IVs are completely crossed"?
Every level of every IV appears in combination with every level of every other IV.
The label a researcher gives a factorial design specifies both ____ and _____.
1. how many IVs exist in the design
2. how many levels of each IV exists in the design
How many IVs and levels exist for the following factorial designs?

1. 2 X 2
2. 2 X 3
3. 2 X 2 X 3
4. 2 X 4
5. 2 X 2 X 2 X 3
1. 2; 2 and 2
2. 2; 2 and 3
3. 3; 2, 2 and 3
4. 2; 2 and 4
5. 4; 2, 2, 2, and 3
Interaction occurs when...
the effect of one IV differs depending on the level of a second IV.

They are always about two or more variables.
Main effects
The simple, straightforward effects of IVs in factorial studies that has to do with only one variable. The overall effect of an IV, averaging across all levels of the other IV(s).
To determine whether your factorial study yielded a main effect of a specific variable you....
collapse across all the levels of all other IVs in a factorial design.
Factorial designs allow researchers to detect the presence of
statistical interactions
Ordinal or spreading interaction
The observed pattern when an effect exists at one level of a second IV but is weaker or nonexistent at a different level of the second IV.

In this interaction the lines DO NOT cross.
Disordinal or crossover interaction
Occurs when there are no main effects of either IV and when the effects of each IV are opposite at different levels of the other IVs

In this interaction the lines DO cross.
Nonparallel lines on a graph for a factorial study indicate:
that there is an interaction
Crossover interaction
Nonparallel lines cross over one another at or very near the middle of two lines.
Simple effects tests
A set of follow–up tests that are conducted when the statistical analysis in a factorial design yields a significant interaction.

Used to see which specific mean comparisons are significant in a factorial story.

Clarify the precise nature of the interaction.
In a true crossover interaction simple effects tests will be ____.
significant
Three advantages of factorial designs according to Fisher:
1. they are more efficient (fewer participants, less noise)
2. more comprehensive
3. and produce more external validity than one–way designs
Factorial designs are efficient in the sense that they...
allow us to look for more than one main effect at a time in a single study
Factorial designs are comprehensive in the sense that...
unlike one–way designs, they tell us more of the whole story behind a specific phenomenon by allowing us to see how different variable work together to influence the phenomenon
Explain the relationship between factorial designs and external validity.
A researcher who observes main effects in a factorial study can be reasonably sure that any observed main effects will generalize across whatever levels of the other IVs exist in the study.
Between–subjects designs
Designs in which each participant serves in one and only one condition of the experiment.
Within–subject designs
Each participant serves in more than one (perhaps all) of the conditions of a study.
What is this an example of:

There are two groups. One that shows squares that are the same color but appear different, and one that shows them appearing to be the same color.

Five participants are randomly assigned to rate the darkness of square A and five other participants rate the darkness of square B on a scale of 1–10.
a between–subjects design
Advantages of within subject designs:
1. Require fewer participants, especially when it comes to complex factorial designs
2. Eliminate person confounds
Disadvantages of within subject designs:
introduce biases that are unlikely to emerge in between–subject designs:

1. sequence effect (getting tired)
2. carryover effect (response to stimulus that influences response to a second)
a. order effect (order questions are asked)
b. practice effect
c. interference effect
3. demand characteristics– increase the likelihood that participants will be able to figure out an experimenter's hypothesis
Sequence effect
occur when the simple passage of time begins to take its toll on people's responses (bored or tired)
Carryover effects
Occur when people's responses to one stimulus in a study directly influence their responses to a second stimulus.
Order effects
Type of carryover effects that occur when a question takes on a different meaning when it follows one question than when it follows another.
Practice effects
Type of carryover effects that occur when participants' experience with one task makes it easier for them to perform a different task that comes along later.
Interference effects
Type of carryover effects that occur when performing one task disrupts people's performance on a second task.
Demand characteristics and evaluation apprehension cause...
participants to try to please the experiment and participants to report whatever they think portrays them in the most favorable light possible
Solutions to within–subject design drawbacks:
Counterbalancing
1. complete counterbalancing
2. incomplete counterbalancing
a reverse counterbalancing
b partial counterbalancing
3. Latin Square
Counterbalancing
Method of control where the researcher varies the order in which participants experience the different kinds of within–subjects studies
Complete counterbalancing
All of the possible treatment sequences are used the same number of times. Each goes through one of the sequences.

Example– Treatments A, B, and C. Six possible sequences ABC, BCA, CAB, CBA, BAC, and ACB.
Incomplete counterbalancing is used when
it is impractical to use complete
Reverse counterbalancing
All treatments are presented to each subject twice. First in one order then in another.

ABBA or ABCCBA order––When subjects experience conditions more than once, they first experience the conditions in one order, and then the reverse order.
Partial counterbalancing
Only some of the possible condition sequences are used because there are more possible sequences than there are subjects.
Latin Square
Procedure that generates fixed number of orders to balance a within–subjects study.

Example

1– A B C D
2– B A D C
3– C D A B
4– D C B A
Latin Squares could be used to _______________, such that each of them appears _______ and ________.
generate four orders; exactly one in each possible serial position; exactly twice before and exactly twice after each of the other three unique conditions
Counterbalancing does not __________ but it does _________. The effects are ________.
erase sequence or carryover effects; unconfound them with particular treatment conditions; balanced
Structured debriefing
Interview conducted with participants immediately after they have completed a study to determine what they thought the researcher expected to find.
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.
Numerosity heuristic
The tendency to estimate quantity or magnitude by basing one's judgments disproportionately on the number of units into which a stimulus is divided– without fully adjusting for other important variables.
Which study in our class uses the two–groups design?
Study 2– warm v cold
The toilet goer's experiment using no confederate, a confederate present one urinal away, and a confederate in the next urinal is an example of?
A one–way three–groups design
Taxonomy
1. Overall number of numbers– how many IVs exist
2. Numbers in the formula– how many levels of each IV exist
Cells
Number of unique conditions of the study
How many cells in a 2 X 2 and 2 X 3 study?
4; 6
Study 3

1. What are the IVs?
2. How many levels are in each IV?
3. Taxonomy and # of cells?
4. What is the DV?
5. What is the design?
6. experiment or quasi?
1. IVs– Valance (positive or negative traits), Mood (positive or negative mood)
2. 2
3. 2 X 2; 4
4.# of traits recalled
5. factorial, mixed–model
6. experiment
Study 3

1. What is the hypothesis?
2. What is the main effect of A?
3. What is the main effect of B?
4. What is the interaction?
1. If you are placed in a positive mood you will recall more positive traits.
2. Valance– negative traits are remembered more than positive traits
3. Mood ??
4. When placed in a positive mood we will remember more positive traits. When placed in a negative mood we will remember more negative traits.
EXAMPLE–

Researchers approach members of other sex on college campus and ask them to have sex.

Three questions, would you go out wit me tonight, would you come over, would you go to bed with me?

Ask two groups, male and female

1. What are the IVs?
2. How many levels?
3. What is the DV?
4. Experiment or quasi–experiment?
1. type of question, gender
2. 3, 2
3. saying yes
4. quasi because you can't manipulate gender
Why are complex designs (2 X 2 X 2 X 2) not popular?
1. You need a lot of participants
2. qualification (unnecessary to have more than two IVs usually)
3. hard to understand
Example–

Rat, fetal alcohol syndrome, short term and long term mental effects from rats who drink during pregnancy and those who don't.

IVs?
DV?
Taxonomy?
Is this an experiment?
IV– Maternal diet, age of rate
DV– passive avoidance
Tax– 2 x 2
Yes
In rat experiment:
What is Factor A?
What is Factor B?
What is the AB interaction?
1. addresses whether maternal diet affects PA learning
2. addresses whether age is related to PA learning
3. addresses whether the effects of maternal diet depend on the age of the offspring
How to compute marginals?
Add up the row or column, then average.
How to find main effect?
Look at marginals, if they are the same there is no main effect.
What type of interaction is involved with study 3?
Crossover
A main effect is qualified by the interaction when?
The interaction explains something more (that changes the meaning) that the main effect does not.
Graph when main effect of factor X and interaction is significant:
There is one straight line and one line that crosses through it
Graph of a interaction and both main effects:
//fce-study.netdna-ssl.com/2/images/upload-flashcards/84/19/18/7841918_m.jpg
EXAMPLE–

In an experiment one group of participants performed a motor task with and without practice. Another group performed a spatial task with and without practice.

The research design is:

1. 2 X 2 between–subjects
2. 2 X 3 mixed model
3. 2 X 2 within subjects
4. 2 X 2 mixed model
5. 2 X 3 between subjects

The between–subject factor is _____; the within subject factor is ______
4; task type; practice
Stapel did what?
Fabricated findings for train research and vegetarian research.
Journals with ______ are more likely to publish retractions that those with ________.
high–impact factors, lower impact factors
Approaches to ethical decisions:
1. Deontology
2. Ethical Skepticism
3. Utilitarian
Deontology
Ethics should be judged in light of a universal moral code.
Ethical Skepticism
Rules are relative to culture and place. The final arbitrator is the researcher's conscience.
Utilitarian
The potential benefit should be weighed against the potential costs.
APA Ethical Principles of Psychologists and Code of Conduct Potential Benefits
1. basic knowledge
2. improvements of research or assessment techniques
3. practical outcome
4. benefits for researchers
5. benefits for research participants
APA Ethical Principles of Psychologists and Code of Conduct Potential Costs
1. time, effort, money
2. harm (psychological or otherwise)
3. confidentiality compromised
What does the IRB do?
Decides if an experiment is ethical before it is begun.
What members are on the IRB?
1. Nonscientific disciplines.
2. At least one member that is a representative of the local community.
What is required of IRB?
1. Written proposal is submitted
2. research cannot be conducted without prior approval
Six crucial aspects of ethics:
1. informed consent
2. invasion of privacy
3. coercion to participate
4. physical and mental stress
5. deception
6. confidentiality
Informed consent requirements:
Must notify people of what is being studied and what will happen.

Must include
–simple langauge, signature, potential risks and benefits, penalties in case of withdrawal, special considerations for kids
What research is allowed without informed consent?
1. minimal risk researcher (nothing more than what happens in daily life)
2. will not impact the rights and welfare of the participants
3. cannot feasibly be carried out if required
Examples of studies that are or might be an invasion of privacy?
1. toilet study
2. asking for sex study
Example of a study that might cause physical or mental stress?
Simulated rape study
Kinds of deception
Active and passive
How can deception be used?
–Experimental confederates
–provide false feedback
–present two related studies as unrelated
–give incorrect info regarding stimulus materials (placebo)
–double deception
Criticism of deception
1. moral ground
2. participants lose their regard to science
3. perseverance effects
How to resolve problems of deception?
debriefing
Debriefing
explaining the nature of the study, the motives, obtain reactions, and show appreciation
Confidentiality means
the data may only be used for purposes of research and should be obtained by maintaining anonymity
Scientific Misconduct examples:

1. fabrication
2. falsification
3. plagarism

Nonexperimental Research Designs can be high in what kind of validity?
External
Case Studies and Goals
Making careful analyses of the experience of a particular person or group.

Goals:
–development or refinement of theories of human behavior
–to describe a rare phenomena
Types of people case studies often involve:
remarkable people with conditions so rare it would be difficult to identity and study them in large numbers or to recreate in lab
Case studies are usually designed to:
generate, delineate, corroborate, or invalidate theories and hypotheses
Who frequently makes use case studies?
Clinical psychologists and behavioral neuroscientists.
What happened with Phineas Gage?
Had a head injury that took out a large chunk of his brain. He had the ability to speak normally and lost no senses or physical abilities. Had normal cognitive abilities but his personality changed.
What did the Phineas Gage case study lead researches to believe?
Damage to specific brain areas is associated with specific kinds of psychological deficits.

Prefrontal cortex damage was primarily responsible of the social and emotional deficits that occurred.
Trait information
Describes what type of person you are. ex. I'm introverted
Autobiographical information
Describes past actions ex. I went to a party on Friday
Episodic memory
amnesia
Cognitive dissonance
occurs when people become aware that they possess two contrary beliefs
What makes case studies scientific?
–source of ideas for theory building
–can be used to test the plausibility of existing theories
–used to test theories that are open to falsification
Scientific approaches to increasing knowledge are so as long as
it proponents are willing to change their minds in the face of information that raises doubts about their original positions.
Negatives of case studies:
–not great with the use of operational definitions or statistical analysis
–not sure if they can be used to generalize
–low external validity
–failure to control extraneous variables
–observer biases
Single variable studies
designed to describe some specific property of a large group of people, focusing on a single variable rather than many

individual participants serve as the unit of analysis

often more than one participant is studies but their responses are analyzed separately
Two most important types of single–variable research studies:
–census
–surveys
Census
A body of data collected from every (or virtually every) member of a population of interest.
Surveys
Identify a subset of people in the population who are then studied. Assumption that the subset would probably represent the attitudes of the entire population.
Population survey
uses random sampling to identify a sample of people to be surveyed
Advantages of population surveys:
easier to do than conducting a census
Problems of population surveys:
–hard to find a population list
–once list is found, it is hard to test everyone who might be selected
Sampling Methods:
– simple random sampling
–systematic sampling
–stratified sampling
–cluster sampling
–handpicked sampling
–snowball sampling
–volunteer sampling
Cluster Sampling
Being by creating a manageable list of all the possible locations ins which they can find members of the population in which they are interested.

Conducted in multiple stages, allows to overcome the constraints of costs and time associated with a dispersed population.
Benefits of cluster sampling:
researchers can make some very good guesses about he attitudes of extremely large groups of people while avoiding the problems involved with a census or with a survey that uses true random selection
Sampling error
statistical calculations estimate how much error is likely to exist in a specific set of population survey findings

reflects the likely discrepancy between the results one obtains in a specific sample and the results one would have been likely to have obtained from the entire population
Two important applications of population surveys are in the areas of:
–epidemiological research
–research on public opinion
Epidemiology
the scientific study of the cause of disease
How clinical epidemiologists study disorders:
randomly sample a population of interest and conduct clinical diagnostic interviews with each person in their sample to determine the precise percentage of people who meet strict clinical criteria for the disorder
Research on Public Opinion and example
designed to determine the attitudes and preferences of specific populations

ex. voters, consumers
Marketing Research
designed to assess consumers' attitudes about and preferences for different products and services
False consensus effect
most people overestimate the proportion of other people who's attitudes and behaviors are similar to their own
Benefits of Population Surveys
–cost effective
–better than sampling everyone
– better than sampling no one and performing and intuitive assessment of people's attitudes
Limitations of Population Surveys:
–complex sampling issues
–people have grown weary of responding to surveys
–surveys and interview questions have to be written so that all different types of people can easily understand the questions
–loss of control (not in a laboratory setting)
–validity and reliability of interviews
–non–response
Example of a single variable convenience sample:
Milgram's pain study
Conjunction fallacy
Ta conjonctive or compound even (both A and B) can never be more likely than one of its component parts (A or B), so someone who assumes A and B together is more likely than A or B alone, is committing this fallacy.
Correlational Methods
–generate fixed set of observations about a group of people and test hypotheses about the associated between different variables
–does not imply causality
–multi–variable
Person confound and ex.
occurs when a variable seems to cause something because people who are high or low on this variable also happen to be high or low on some individual difference variable

ex. anxiety and depression
Environmental confound and ex.
situational

ex. stressful life events might cause people to become more depresses while simultaneously lowering their self–views
Operational confound in non–experimental design
occur when a measure designed to assess a specific construct measures something else as well

when something else is measured that 1) has nothing to do with the predictor variable that the researcher really wished to measure but 2) has a lot to do with whatever outcome the researcher wished to predicts
Example of operational confound:
definition depression as loss of taste for food and difficulty getting up, cofounded with AGE
Ways to fix an operational confound:
1. discard or device measure to remove the confound
2. you could measure a nuisance variable that you think is really accounting for the action and perform statistical tests that are designed to tell you which are really responsible
3. run a replication study in which you focus exclusively on participants who are roughly the same age
Reverse causality
A might be causing B or B might be causing A.
Way to address the problem of reverse causality:
–longitudinal designs
Longitudinal designs:
–follow people per time and make repeated assessments of the variables in which they are interested
Third variable problem
correlation research is susceptible to confounds

C might be causing A
Benefits of correlational research:
1. many things cannot be studied ethically or efficiently in the lab
2. the gain in external validity sometimes compensates for losses in internal validity
Correlational Research Methods:
–archival
–observational
–field studies
Archival research designs
research in which investigators examine naturally existing public records to test a theory or hypothesis
When are archival research designs good to use:
when a topic is ethically sensitive, when a variable is difficult or impossible to manipulate, or when a researcher is especially motivated to conduct a study high in external validity
Benefits or archival research designs:
–potential for high levels of external validity
–eliminates worries that a set of research findings occurred because the people being studies are trying to manage of positive impression with an experiment
Negatives of archival research designs:
–hard to gain internal validity
–wide range of potential confounds
–people who generate public records of real–world events often do so for reasons that have nothing to do with research
–dependent on others
Observational Research
research in which investigators record the real behavior of people in their natural environments

–always real and not manipulated by researchers
Benefits of observational research
give researchers some control over exactly what and how they observe
Unobtrusive observation
1) researchers do not interfere in any way with people's natural behavior
2) people do not realize that they are being studied
Positive and negative of observational studies:
high in external validly, low in internal
Positive and negative of single–variable studies:
good at describing the state of the world at large,
do not necessarily tell us much about why it is that way
Criticism of Group Designs and Analysis
1. generalizability
2. reliability
3. ethical issues
4. error variance
Idiographic
the study of the individual who is seen as a unique agent with a unique life history, with properties setting him or her apart from other individuals
Nomothetic
study of classes or cohorts of individuals
Criticism of Group Designs and Analysis in regard to learning curves and averages:
1. an individual's learning curve might be all or nothing, not gradual, when an average shows a curve, indicating hat the larding is gradual but it isn't

2. two sets of data might have the same averages but may vary immensely by individual
Ethical question of single–case experimental designs:
is it ethical to remove a clinical intervention?
Forms of single case experimental designs
1. A– B
2. A–B–A
3. A–B–A–B
4. A–B–BC–B
Robbies behavior record design:
1. baseline
2. operant
3. remove operant conditioning
4. reinstate operant conditioning
5. continue (assess later)

ABAB design
Irreversibility
In some withdrawal designs, once a change in the independent variable occurs, the dependent variable is affected. This cannot be undone by simply removing the independent variable.
Groupthink
1. Excessive tendency to seek concurrence among group members

2. emerges when the need for agreement takes priority over the motivation to obtain accurate information and make appropriate decisions
How to challenge groupthink? And examples
Institution of systematic procedures.

Ex. supreme court must allow everyone to state their opinion once before others than talk again, or israeli military intelligence branch hires someone to come up with an opposite prediction
Problems of a census:
can take a long time and cost a lot of money
What is sampling error influenced by?
Population size
Focus of the study
Simple Random Sampling
identifying all elements of a population, listing them, randomly selecting from list

difficult and might not capture enough elements of a particular subgroup
Systematic Sampling
selecting every nth case within a defined population, easier than random selection but cannot show a pattern
Stratified Sampling
dividing population into subgroups and taking random samples within each one

ensures sample represents key subgroups

should be proportionate
Handpicked Sampling
selection of a sample with a particular person in mind
Snowball Sampling
building a sample through referrals
Volunteer Sampling
Sampling by asking for volunteers
Single Variance Convenience Sample
demonstration of human judgment or behavior that is so interesting, counter intuitive or disturbing that it is surprising to see it in any sample
Why does correlational research not imply causality?
– third variables
–reverse causality
Cross–Sectional Designs:
Non–experiemntal research design in which outcomes are measured for each individual during one period.
Cross–Legged Design
Correlational research design that will help gauge what is causing what.

Look at the cross between time one and time too and judge the correlational coefficient to see which seems to be causing which.
The Dyadic Interaction Paradigm
How to overcome ethical problems in observational research: ask people to do something, do not tell them they are being filmed, after they do it, tell them they were filmed and ask if you can review the footage or not.
How to deal with confounds?
1. think hard and measure them
2. statistically remove influence
Lab experiments show us
exactly how the world changes when you change one and only one thing.

Solve problem of individual differences and help minimize confounds.
To be considered an experiment a study must make use of:
–manipulation
–random assignment
Manipulation
occurs when an experimenter systematically alters the levels of a variable to see if changes in the IV lead to corresponding changes in some outcome variable
Random Assignment
Placing people in conditions on a totally arbitrary basis. It gives every participant an equal chance of being assigned to any specific condition of an experiment.
Matching
Came before random assignment.

Matching each participant in an experimental condition with a very similar participant in a control condition
Problems of matching:
–really hard to do
–doesn't really work

a researcher cannot know with certainty that two matched groups have been equated on all the relevant dimensions that could influence the dependent variable
Benefits of random assignment:
a lot quicker and easier than matching and equalizes two or more groups on practically every dimension
Random sampling versus random assignment:
RS is used in population surveys to help increase the odds that it will generalize (external validity), RA is used in experiments to help equate different experimental conditions before manipulations are applied (internal validity)

Both maximize the likelihood that two separate groups of people will be as similar as possible.
Ways to fix problems in Random Assignment
–Statistical testing
–replicating study
Random assignment only fixes what type of confound?
person
Procedural confound
Logical equivalent of the environmental confounds of non experimental research.

Occur when an experimenter unwittingly manipulates two or more things at once.
When an experimental manipulation suffers from an operational confound we still know that...
the IV caused the observed changes in the DV but we are worried that the IV might not represent the construct that the experimenter had in mind.
Researchers can avoid confounds if they
1. randomly assign participants
2. develop manipulations that are free of procedural and operational confounds
3. move studies to the lab where great control can be exerted
Semantic priming effect
finding that people recognize most words more quickly than usual when they have just been exposed to words that have similar meaning
True Experiments:
Provide researchers with clear and compelling info about causality, are especially high in internal validity, use random assignment to eliminate person confounds, take steps to eliminate other confounds, and police noise.

Allow researchers to determine whether the IV has an impact on some outcome of interest.
Noise
extraneous variables that influence the dependent variable but are evenly distributed across experimental conditions
Solution to potential artificial nature of experimental design:
1. mundane realism
2. experimental realism
Mundane realism
make a study as similar as possible to the real–world setting you care about

looks like the real world
Experimental realism
the degree to which a research study is psychologically meaningful to research participants

feels like the real world
How do we know when a study is high in experimental realism?
1. manipulation checks
2. deception
Manipulation check and example
a measure designed to see if a manipulation truly puts people win the psychological state that the experimenter wishes to create

in experiment on testosterone and winning, asked people to remember how many times they won to make sure they are in the intended state of mind.
Lab experiment advantages over correlational or observational studies:
1. superior info about causality
2. may yield important insights into the psychological mechanisms that are responsible for a research finding documented in the field
3. allow to observe unobservable
4. eliminates confounds
5. minimize noise
Pilot Tests
Good to conduct before the full experiment.
Experimental Study with Two Independent Variables
Cockroach study, watched or not watched
Two types of matching:
1. frequency distribution
2. precision matching
Precision matching:
matching one variable such as age
Frequency distribution:
accounting for more than one characteristic such as age and gender
When to use matching?
When research is so important the drug( for example) is so costly that you would like to have matching because you know the important parameters and you know that the groups are similar going into it
Level
The amount of independent variables. The least amount you can have is two– control group and experimental, the max is as many as you want.
Qualitative IV
express a qualitative attribute such as hair color, eye color, religion, favorite movie, gender, and so on. The values of a qualitative variable do not imply a numerical ordering.
Quantitative IV
those variables that are measured in terms of numbers. Some examples of quantitative variables are height, weight, and shoe size.
3 types of Independent Variables:
1. environmental
2. instructional
3. invasive
Invasive IV
Injecting something
Environmental IV
Changing something about the environment like the room color
Instructional IV
Teaching something
Dependent Variable
Variable measured by and experimenter. Experimenter expects the depender variable to be influenced by the IV.
Independent Variable
An experiment that is manipulated by the experimenter. The experimenter expects the IV to influence the dependent variable.
Types of Dependent Variables
1. correctness– right or wrong
2. rate or frequency
3. degree or amount
4. latency or duration– how long does it take to start doing something
Internal validity:
causation
External validity:
generalization

How to conduct a good lab study:

1. setting the stage, optimistic bias/planning fallacy
2. be professional

metaphysical
supernatural explanations
Three categories of metaphysical explanation
1. animism 2. mythology and religion 3. astrology
animism
belief that the natural phenomena are alive and influence behavior
mythology and religion
make the assumption that deities play and important role in human behavior.
astrology
human behavior is determined by the celestial bodies
Are metaphysical explanations scientific?
NO
Precursors for Psychology
1. philosophy 2. physiology
Philosophy
the study of knowledge, behavior, and the nature of reality by making use of logic, intuition, and empirical observations.
Positivism
study of human behavior should be based only on observations that can be made with absolute certainty
Empiricism
the idea that the best way to learn about the world is to make observations -we do not know a priori
Physiology
-the study of the functions of and interrelations between different parts of the brain and body -experimental psychologist owe a lot of their discipline to the traditions and methods developed by physiologists. -logic, intuition and empirical observation
Example of the experimental method...
using frogs to dispute that nerves were tiny pipelines for animal spirits
Year experimental psychology was invented and where and by who
Germany 1800s, Wilhelm Wundt
Psychology
the study of human behavior
Canons
Fundamental principles that are more or less accepted on faith.
Four canons accepted by scientists:
1. Determinism 2. Empiricism 3. Parsimony 4. Testability
Determinism
-the doctrine that the universe is orderly -the idea that all events have meaningful, systematic causes -theory driven -research is not meaningful without it
illusory correlation
-people falsely infer a connection or correlation when none truly exist
B.F. Skinner's pigeon experiment shows
-non contingent reinforcement or superstitious conditioning -falsely conditioned pigeons to think that a behavior causes them to get food
Superstitious conditioning
The "false" conditioning that often occurs when an organism is provided with reinforcements at random intervals
Theory
-a statement about the casual relation between two or more variables -typically stated in abstract terms and usually has some degree of empirical support -cannot be tested as it -should be deterministic (logical and orderly) -should be testable -boundary conditions- there are times when they do not apply -possible for more than one to be true
Parsimony
-recommendation about the kid of theory or explanation that a good scientist should prefer -if we are faced with two competing theories that do an equally good job of handling a set of empirical observations, we should prefer the simpler one
Testability
-most important canon of science -the assumption that scientific theories should be testable (confirmable or disconfirmable) using currently available research techniques -should strive to prove yourself wrong, if you can't your theory gains more support
Testability is closely associated with what?
Falsifiability
Falsifiability
The idea that scientists should go a step beyond putting their theories to some kind of test by actively seeking out tests that could prove their theories wrong.
Logical positivism
-the belief that science and philosophy should be based solely on things that can be observed with absolute certainty -many believe that the way to go about testing theories and hypotheses is to try to disconfirm them
Operational definitions
-definitions of theoretical constructs that are stated in terms of concrete, observable procedures -solve the problem of understanding things like love by connecting unobservable traits or experiences to things that can be observed -makes abstract hypothetical concepts measurable and thus testable
Example of an operational definition of attraction
those who are attracted to someone will choose to sit closer to them
What are the four ways of knowing about the world?
1. intuition 2. logic 3. authority 4. observation
Intuition
A gut feeling not necessarily supported by research
Who uses faith and intuition?
religious and political thinkers
Logic
Describes the use of valid reasoning
Who uses logic?
scientists and philosophers
How religion and government knows about the world:
1. authority 2. intuition 3. logic 4. observation
How philosophers knows about the world:
1. logic 2. observation3. intuition 4. authority
How scientists know about the world:
1. observation 2. logic 3. intuition 4. authority
Barnum effect
The observation that people tend to believe in descriptions of their personality that supposedly are descriptive of them but could in fact describe almost anyone.
Astrology
human behavior is determined by the activity of celestial bodies
a priori
Before experience
Rationalism
what seems logical is
Positivism
theory of knowledge based on observations with certainty
Galton's View On The Power of Prayer and criticism
if prayer wrks then the Queen should live longer than anyone else -need a basis for comparison -commoners not good because they don't have a great quality of life -royalty have incest and die earlier -shows prayer has no effect
Research Hypotheses
-specific proposition that logically follows from the theory -a priori -almost all can be reduced to if-then -must be falsifiable
Outcomes of theories should either
enforce it or lessen the belief
Limitations on empiricism and example
people might change their behavior if they know they are being watched ex. Hawthorne effect (workers performed better knowing they were being observed based on performance)
Confederate
an individual involved with researcher without other participants knowing
Two principles that explain animal and human behavior
1. classical conditioning 2. operant conditioning
classical conditioning
A type of learning in which one learns to link two or more stimuli and anticipate events
operant conditioning
A type of learning in which behavior is strengthened if followed by a reinforcer or diminished if followed by a punisher
Conceptual definition
similar to a dictionary definition
Pierce's Method of "The Fixation of Belief" Involves
1. intuition 2. authority 3. consensus 4. a prior method (logic) 5. the scientific method
Problems with method of intuition
false-consensus effect
False consensus effect
-if you agree to something yourself you estimate more people also agree -our own point of view changes the way we see the world
Problems with method of authority
Milgram Study- authority can be wrong
Milgram Study
shocking "participants", psychiatrists predicted that very few would go all the way but 2/3 did
Problems with a priori method
support for hypotheses is not a weakness or a strength
Three Points of Scientific Approach
1. empiricism2. public verification (data should be available to all)3. solvable problems
Three Goals of Behavioral Research
1. describing 2. predicting 3. explaining
Can we prove in science?
we can confirm or support but not prove
Strategies of Behavioral Research
1. descriptive 2. correlational 3. experimental 4. quasi-experimental
Descriptive Research
Has the purpose of observing and recording behavior.
Correlational Research
A research strategy that identifies the relationships between two or more variables in order to describe how these variables change together. The direction and degree of the association is described by a correlation coefficient.
Experimental Research
Gathering primary data by selecting matched groups of subjects, giving them different treatments, controlling related factors, and checking for differences in group responses
Quasi-experimental Research
Research in which naturally occurring situations are used rather than experimental manipulations; usually used to avoid practical and ethical problems.
Descriptive laws
what people ought to do
Prescriptive laws
what people actually do
Law
-a universal statement of the nature of things that allows reliable predictions of future events -comprehensive, fundamental statements
Equifinality
the notion that the same behavior is often produced by many different causes
Hypotheses
specific predictions that can be readily tested that are derived from a theory
Method of Induction
-making many observations under controlled conditions and arriving at a general statement about how things are -reasoning from specific instances to general principles
How to deal with potential outcomes of theories:
if new observations are consisted with the statement then the statement survives, if not, the statement is either discarded or revised and tested again
Problem of Induction
You can never know how many observations are good enough to know that a law is true.
Deduction
-reasoning from general to specific -when a general statement (theory) is used to develop predictions (hypotheses) and then tested against observations
Can you prove a theory false or true?
Yes to false, no to true.
Falsifying theory
much can be learned about something from failure. sometimes it is more useful to know what is not true that to know what is
Induction and problems
problem- not guaranteed to be the same. we can't know what is going to happen, you can't know how many tests are necessary
Three basic ways scientists test hypotheses:
1. Validation 2. Falsification 3. Qualification
Positive test bias
the tendency for people who are evaluating hypotheses to attempt to confirm rather than to disconfirm these hypotheses
behavioral confirmation
the tendency for social perceiver to elicit behaviors from a person that are consistent with their initial expectancies of the person
Validation
an approach to hypothesis testing in which researchers attempt to gather evidence that supports or confirms a theory or hypothesis
Cognitive Dissonance
When a person simultaneously holds two beliefs that are dissonant the person will experience an aversive state of arousal and the person will be highly motivated to reduce this aversive arousal.
Falsification
researchers attempt to gather evidence that invalidates or disconfirms a theory or hypotheses
Qualification
researchers try to identify the boundary conditions under which a theory or hypothesis is and is not true. can lead to the integration of two contradictory theories by specifying the conditions user which each of the theories is correct
Problems with Qualification
it is inherently complicated and requires the researchers to have some sophisticated ideas about the world, etc.
Advantages of Qualification
-combines the desirable features of both validation and falsification -research will represent a closer approximation to reality
Experimental Paradigm
-an ideal set of research procedures that serves as the model for almost all psychological research -grounded in the assumption that experimentation is the most useful tool for figuring out the true causes of human behavior
Case Studies
carefully documented observations of a specific group or person
Paradoxical incidences
puzzling or nonsensical observations
Analyzing the practitioner's rule of thumb
analyzing things that experts in a particular area do to achieve certain outcomes
Serendipity
good luck or fortune
List the inductive techniques
-cause studies -paradoxical incidences -practitioner's rule of thumb -serendipity
List the deductive techniques
-analogy -applying a functional or adaptive analysis -hypothetico-deductive method -accounting for conflicting results -accounting for exceptions
reasoning by analogy
reasoning that draws a comparison between two ideas, things, or situations that share some essential common feature
applying a functional or adaptive analysis
Analyzing what an organism has to do to successfully master an environment or achieve a desired end state
hypothetico-deductive method
The ability to use deductive reasoning manipulate several variables, test the effects in a systematic way, and reach correct conclusions in complex problems; scientific reasoning.
accounting for conflicting results
attempting to come up with reasons why studies on the same topic have yielded different findings
accounting for exceptions
Attempting to generate exceptions or limiting conditions to well-established psychological principles or empirical findings.
Validity
relative correctness of the statement -tries to answer whether the measurements are appropriate
Types of validity
-internal –external –construct -conceptual
Internal Validity
extent to which a set of research findings provides compelling information about causality
Covariation
-for one variation to cause another, changes in one variable must correspond with changes in the other -shows a relationship between A and B
temporal sequence
the changes in the first variable must precede the changes in the second ex. eating ice cream and drowning, if A is causing B first there will be changes in A and then changes in B. B can't change first if we are going to say that A is causing B.
third variable problem
two variables will covary with one another and give the false appearance of a causal relation
eliminating confounds
eliminating other variables that might be the cause of something. making sure that we control for all other possible alternative explanations so that A is not said to cause B if C is actually causing B.
Laboratory experiments good and bad
good- allow researchers to access covariation, to establish the temporal sequence of events, and to eliminate a great number of confounds be means of random assignment and the manipulation of variables -control for individual differences -isolate independent variables from contamination bad- not always high in external validity -artificial
external validity
extent to which a set of research findings provides an accurate description of what typically happens in the real world
passive observational studies
high in external validity because they are out in the real world
construct validity
whether the manipulated and/or measured variables accurately reflect the variables the researcher hopes to manipulate or measure -convergent and divergent validity to test this
conceptual validity
-how well a specific research hypothesis maps onto the broader theory that is was designed to test -whether the researcher should have been interested in that specific hypothetical construct in the first place
reliability
repeatability of a measure of observation
Ways to assess the reliability of a measure?
-test-retest reliability -internal consistency -interrater reliability
Test-retest reliability
measuring a group of individuals at one time and then having them come back a second time to take the test again if scores are the same both times then the measurement is reliable -appropriate when test takers not permanently changes by taking test or interval is long enough to prevent practice effect -give the same test the the same group at different times then compare the scores of each person from both using a correlation
internal consistency
A measure of reliability; the degree to which a test yields similar scores across its different parts, such as on odd versus even items. You can do a test administration and then split it into two and see how they did on one how and how they did on the other. This is only if it is measuring one trait like multiplication
interrater reliability
the degree to which different judges agree upon an observation
What is common to all forms of reliability?
likely to increase as we increase the number of observers, observations, or occasions that go into the measure
Nominal Scale
simplest kind of measurement. involve meaningful but potentially arbitrary and nonnumerical names or categories ex. sex, ID numbers, SS numbers, is or isn't
Ordinal Scale
scales that have clear numerical properties. involve set of mutually exclusive and exhaustive labels, clear hierarchy, not sensitive to the absolute differences between things ex. ranking
Interval Scale
measurement scales that make use of real numbers designating amounts to reflect relative differences in magnitude -can go below 0... example negative degrees ex. weight
Ratio Scale
like interval scales except they have an important attentional property, they always have a true zero point. nothing under zero is present. allows us to speaking meaningfully about ratios of values
Example of Equifinality
aggression could be bio theory, could be frustration, could be social learning
social facilitation
Stronger responses on simple or well-learned tasks in the presence of others
Cockroach racing study
1. straight, audience made them go faster (easy) 2. maze, audience made them go slower (hard) nature of outcome makes a difference depending on difficulty of task -Zajon's Mere Presence Theory
fundamental retribution error
What we don't know, we assume.... although prevent to not be the case in india
Correlation coefficient
Range from 1 to -1, +/- refers to direction, 0 to 1 indicated size of the relationship, dependent on the field that you are conducting research. physical ream .883 is high, psychological realm .45 is high
Positive correlation
as one variable changes positively so does the other (gain in height, gain in weight)
negative correlation
as one variable changes negates so does the other (weight loss and calorie intake)
Who has the higher correlation coefficient value in absolute terms?
height and weight r=+.833, IQs of married couples r=+.45 height and weight
Covariance
a number that represents the degree to which two different variables change together
causality and what is needed to establish it
confidently concluding that variations in the independent variable caused the observed changes in the dependent variable
in order to establish causality one needs:
1. covariation 2. temporal sequence 3. eliminating confounds
independent variable
what we manipulate
dependent variable
what we measure
Ways to Eliminate Confounds:
1. random selection 2. random assignment 3. other confounds? - measure them and then remove them when you measure the dependent variable
Random selection
anyone in the population has an equal chance to participate in the study
Random assignment
once we have participants we assign some to control group and some to experimental group randomly
Generalizability
a property of a study in which the finding can be applied to 1. other people, 2. other physical or social environments
Crown and Marlowe studied what?
Told people things were important and measured how much they enjoyed it. Those with a high social desirability scale showed higher scores than those who were low.
Conformity Measure
how much you are influenced by others in your presence.
Face validity
the extent to which the test seems on the surface to be measuring something relevant. -does the test look like it measures what it is supposed to? -tells us nothing about what a test really measures -important for creating rapport and reducing examinee's suspicions -examinees may take it more seriously
Why is face validity important?
If it does not appear face valid it may not be taken seriously by the participants.
Can a test be reliable but not valid?
Yes

Can a test be valid but not reliable?

No
Causes of Measurement Error
1. transient states- how the person is feeling 2. stable attributes- how the person is in general (intelligence) 3. characteristics of the measure- the behavior of the researcher and atmosphere 4. mistakes in recording responses
Inter-Observer Reliability
refers to the degree to which different judges independently agree upon an observation or judgment. Violating the principle of independence of ratings can falsely increase the reliability of a set of rating without necessarily increasing their validity. Goal to achieve high reliability and not a high number of judges.
Ways to increase testing reliability

-standardized administration -clarify instructions -train observers -minimize error in coding data