Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
126 Cards in this Set
- Front
- Back
Surveys
|
-major function is to dispell myths
-look at how pp feel -quantify behavior -allow examination of possible cause-effect relationships |
|
Questionaires
|
-consider what behaviors want to quantify (purpose)
-consider types of questions going to use (open-endend or close-ended) |
|
Using previously developed questionaires
|
-aviod redesigning
-compare results w/ previous studies using the same instrument |
|
What should you ask yourself when desinging questions?
|
1. What do I expect to accomplish (purpose of doing survey)
2. What questions of interpertation may arise when I have the data |
|
What are the two types of questions?
|
1. open-ended: respondents answer in their own words (i.e. short-answer or essay)
2. close-ended: limit the respondent to certain alternatives |
|
Pros of open-ended questions
|
-more complete answer
-see reason behind answer so learn more -may get non-anticipated information |
|
Cons of open-ended questions
|
-hard to score b/c answers are in narrative form
-require more effort from respondent |
|
When do you use open-ended questions?
|
-with small groups
-with preliminary studies b/c get range of answers so can standardize alternatives into close-ended format |
|
Pros of close-ended questions
|
-easy to score (scoring is usually automated
-easy to answer (don't need to generate answer) -respondents don't have to be as articulated |
|
Cons of close-ended questions
|
-range of answers are restricted making it hard to reduce complex issue into small set of alternatives
-may not give adequate choices -doesn't reveal reason behind answers |
|
When do you use close-ended questions?
|
-with large groups
-with refined studies |
|
What does using both types of question allow for?
|
-easy coding and analyis of data
-insight into respondent reasoning for choice |
|
What are the "do's" of writing question?
|
-only address one issue per question (no loaded questions, unambigious, and clear)
-avoid bias in question (may be in wording) -make alternatives clear, concise, and easy to understand |
|
Mutually exclusive
|
-answers should be mutually exclusive
-categories defined so that membership in one ruls out membership in another -only answer one way, no overlap |
|
Exhaustive
|
-answers should be exhaustive
-categories defined so that all possible cases will fall into one of them -give entire list of possibilities |
|
Social Desirability
|
-characteristic of certain repsonses that cause ppl to choos that response even if it doesn't represent their true tendency or opinion
-aviod this in questions to avoid bias |
|
How do you avoid social desirability
|
-use 2 or more forms of a question (verification key)
-word question so each alternative appears equally socially desirable |
|
Verification Key
|
-colletion of items on a questionaire desigened to detect dishonesty
-show tendency of person to be influenced by social desireability |
|
What is a lie scale
|
-form of a verification key
-used to determine if answer questions honesty and if take questionaire seriously |
|
Acquiscence
|
-tendency to aggree w/ a statement on a questionaire regardless of its content
-T/F, agree/disagree are susceptible to bias caused by this |
|
How do you decide the format of a questionaire?
|
-depends on issue addressed
-answers may take many forms (T/F, multiple choice) so dependent on type of question asked |
|
Likert Scale
|
-question that asks for a rating of the extent of agreement or disagreement w/ a statement
-rating scale that measures magnitude of opinion -usually 7 pt scale |
|
What affect does sequence have on a questionare?
|
-sequence items carefully b/c answers to some questions may be biased if they were to follow some others
-order effects (may get tired and not respond as clearly) |
|
Branching format
|
-convient b/c permit respondent to skip inappropiate questions and move through questionaire more efficientyl
-terms used muse be clearly defined |
|
Data Analysis
|
-how going to score and analyize questionaire
-what type of statistical analysis going to run -do prior to data collection |
|
Methods of administration
|
-one that is best depends on circumstances
-face to face, telephone, written, or computarized |
|
Pros of face to face
|
-develop rapport w/ person
-motivate them to answer carefully -provide clarification if respondent misunderstands -get them to elaborate (more info) |
|
Cons of face to face
|
-introduc bias b/c of social desirability
-expensive b/c one at a time -safety (danger to interviewer) -difficult to supervise interviewer (fake data) |
|
Pros of written
|
-low cost
-impersonal (no bias) -usu can complete at leisure |
|
Response rate
|
in survey research, the percentage of individuals in the sample who return the completed survey
|
|
Cons of written
|
-lower response rate
-no clarification -don't know if person takes questionaire seriously -illeterate participants comprimise reliabilit and validity of data |
|
Pros of computerized
|
-cheap, fast collection, high response rate
-impersonal (reduce bias) -consistent adminstration (everyone takes same way) -reduces noise -check invalid responses -use branching format -avaible 24/7 (high response rate) |
|
Cons of computerized
|
-illetrate subjects
-difficult to program -since unmonitored, ppl could lie bout age or gender making hard to get truly random sample |
|
Pros of telephone
|
-cheap
-random-digit dialing (random sample) -rapid -central location for interviewer (reduce cost) -possible to use computer-assisted interviewer |
|
Cons of telephone
|
-need short questionaire to get respondent of finish
-more personal -difficult to est good rapport -intrusive -reduce external validity b/c only ppl w/ phones and patience are sample |
|
Low response rates
|
-biased in direction of the more vocal ppl
-biased in direction of those who strongly agree or disagree |
|
What is the quality of the data base on?
|
-direction function of the return rate
-at least 50% response rate -preferably 90% |
|
Response bias
|
-refusal to cooperate, failure to return questionaire, or unavaibility of target respondent shoud be recorded
|
|
Five ways to sample
|
hapazard sample
purposive sample convenience sample probability sample cluster sample |
|
Hapazard Samples
|
-population subgroup for whose selection the researcch use hit or miss methods
-almost worthlessq |
|
Purposive samples
|
-nonrandom sample that is chosen for some characteristic that is possesses
-meet particular definition or selected for reason |
|
Pros of purposive samples
|
Pro: frequently preferable to random sample b/c can be considered to constitute a population
|
|
Cons of purposive sample
|
-error in judgement by researcher selecting sample may influence results
|
|
Convience Sample
|
-nonrandom sample chosen for practical reasons
-doesn't come close to sampling all of population -most pys research uses this |
|
Probability Samples
|
-reseacher knows the probability that any given individual will appear in the sample
-able to apply various stats -usu use random sampling |
|
Random Sample
|
a sample in which every member of the population has an equal and independent chance of being selected
-controlled by chance alone |
|
Random Selection
|
-controlled by chance alone
-results from selecting some but not all pairs -any other method of selection could result in some nonindependence among members of a group -statistical independence |
|
Sampling Frame
|
-a population as it is defined for the purpose of selecting subjects for a study
-population you will work w/ in the study |
|
Element
|
-individual memeber of a sampling frame
-each individual that falls within the sampling frame |
|
Types of random samples
|
systematic sample
simple random sample stratified sample |
|
Systematic sample
|
-a probability sample that is not randomly selected
-fails eqaul-probability part of random selection definition |
|
Pros of systematic sample
|
-taking every nth person is less work than random method
-some structure to list, results nonrandom -usually survey research b/c easy |
|
cons of systematic sample
|
-no structure to list, then resulst will be random
|
|
Simple Random samples
|
-group chosen from an entire population such that every member of the population has an equal and independent chance of being selected in a signle sample
-used when pop is believed to homogenious w/ respect to question of interest |
|
Stratified random sample
|
-a random sample in which two or more subsamples are represented according to some predetermined proportion, generally in the same proportion as they exist in the population
-use if pop has identifiable subgroups that are likely to differ in repsonse -treat pop as 2 or more separate subpop. and create a separet random sample of each -accurate |
|
Use of stratified random sample
|
-oversample some subgroup of the population
-purposely include some group at greater frequency than it is rep. in the population |
|
Cluster samples
|
-group selected by using clusters or groupings from a larger population
-use if impossible or impractical to number pop. -not as accurate as random sampling |
|
Multistage sampling
|
-a form of cluster sampling in which the clusters are further broken down by taking samples from each cluster
|
|
True Experiment
|
research procedure in which the scientist has complete control over all aspects
|
|
By controlling all variables, the experimenter can...
|
-assign subjects randomly to conditions
-control way expt is conducted |
|
quasi-experiment
|
when a study does not meet the requirements of a true experiment
|
|
Factors
|
-IV's of the expt
-have at least 2 -have at least 2 levels (particular value of an IV) |
|
Condition (treatment)
|
-a group or treatment in an expt
-term to discuss IV -way in which subjects are treated |
|
One-group Posttest-Only Design
|
-research design that measures the behavior of a single group of subjects after they are given the treatment
-many threats to validity are not controlled for |
|
non-equivalent control group
|
-group of subjects that is not randomly selected from the same population as the experimental group
|
|
Post-test with non-equivalent control group
|
-non-equivalent control group is better than no control group
-quaise expt b/c subjects were not randomly assigned to groups |
|
One-group pretest-posttest design
|
-measures the behavior of a single group of subjects both before and after treatment
-still threats to validity -addition of nonequivalent control group improves control sufficiently |
|
What two elements of design are basic to all experimental designs b/c they control for many threat to validity?
|
-existence of a control group or a control condition
-random allocation of subjects into groups |
|
With-in subject experiment
|
-possible that some variable may influence data as a result of repeated testing
-avoid if think order or sequence effects will be substantial |
|
Order effects
|
-changes in a subject's performance resulting from the position in which a condition appears in an expt
|
|
Sequence effects
|
-changes in a subject's performance resulting from interactions among the conditions themselves
|
|
Counterbalancing
|
-controlling for order and sequence effects by arranging that subjects experience the various conditions in different orders
|
|
What are the two strageties to controll for order and sequence effects?
|
1. arrange the order of the conditions in such a way that order and sequence effects are controlled w/in subjects
2. control for order and sequence effects bwt subjects |
|
Ways to control for order effects in w/in-subjects design
|
randomization, block randomization, counter-balancing, reverse counter-balancing, and Latin square
|
|
Randomization
|
-randomize order of conditions for each subject
-used when each condition is given several times to each subject |
|
Block Randomization
|
-order of conditions is randomized but with each condition being presented once before any condition is repeated
|
|
When's block randomization useful?
|
-if each condition is presented at least twice and expt requires 2 or more sessions
-if conditions are presented several times to subjects |
|
Reverse counter-balancing
|
-conditions are presentedd in order the first time and then in reverse order
|
|
When do you use reverse counter-balancing
|
-relatively few subjects
-several conditions presented a few times -typical ex. is w/ 3 conditions |
|
When does counter-balancing work well?
|
-when you suspect that the possible confounding variables will act in a linear manner over conditions
|
|
Latin Square
|
-control procedure in which each subject experiences each condition in a different order from other subjects
-incomplete counterbalancing b/c doesn't meet requirement that each condition follow every other condition an equal number of times -sequence is not controlled for -more flexibility in choosing # of subjects to be tested -used when not possible to control for order and sequence effects within subjects |
|
two-conditions design
|
-simplest research design, involving only 2 conditions
-each subject serves as own control -all subjects experience both conditions in counterbalanced order |
|
Why is 2 condition design tested within subjects not used that often?
|
1. many expts involve more than 2 conditions
2. possibility of carry over effects from one condition to the other |
|
Multiple condition design
|
-research design that involves more than 2 conditions
-usually between-subject experiments |
|
When do you use multiple condition design?
|
1.compare several variables or treatments for effectiveness
2.determine the shape of the function that relates the IV and DV 3.presence of more than one rival hypotheisi must be ruled out |
|
Why use the two condition design tested between subjects?
|
-possibility of large order or sequence effects is present
-once participants experience one condition they would no longer be naive about the situation |
|
Why use the multiple conditions design tested between subjects?
|
-study the effects of several levels of one IV
|
|
When is a between-subject design used?
|
-when a significant interaction bwt conditions would occur if test within subjects
|
|
When are multiple-condtions, within-subjects experiments common?
|
-in perception research, as when one scale the brightness of different intensities of light
|
|
Observational Research
|
-observe and record ongoing behavior but do no attempt to change it
-can be experimental or non-experimental |
|
Characteristics of true experiments
|
-manipulation
-assignment -strong inference about casuality |
|
Manipulation
|
in an experiment, conditions or variables assigned or presented to a participant
|
|
assignemtn
|
pairing a subject with a condition or variable according to the experimenter's plan
|
|
observation
|
record of behavior
|
|
correlational research (non-experimental research)
|
-measures 2 or more variables to determine the degree of relationship bwt them
-dont manipulate IV, assign subjects, and cant draw strong inferences bout casuality -often used to guide later experimental research |
|
Hermeneutics
|
-nonexperimental
-study of behavior to understand behavior rather than determine casuality |
|
archival research
|
-examine existing recordes to obtain data and test hypothesis
|
|
case study
|
-exploratory study of an existing situation as a means of creating and testing a hypothesis
|
|
Theory development
|
-hypothesis to prediciton to test prediction through observation to evaluate
-if consistent w/ hypothesis keep it -if not consistent w/ hypothesis generate new hypothesis |
|
deductive reasoning
|
-premise to conclusion
-experimental research -have theoritical concept |
|
inductive reasoning
|
-specific example to conclusion
-non-experimental research |
|
survey
|
-assessing public opinion or individual characteristics by the use of questionaire and sampling methods
|
|
What are the 4 non-experimental research methods?
|
observational
archival case studies surveys |
|
observational research
(ethology: animal study) |
-naturalistic observation
-unobtrusive research -non reactive research |
|
naturalistic observation
|
observational research of subjects in their natural environment carried out to disturb the subjects as little as possible
|
|
unobtrusive research
|
another term for naturalistic observation commonly used in the social sciences
|
|
non reactive research
|
another term for naturalistic observation in the social sciences emphasizing the subjects are unaware that they are being studied
|
|
Physical trace
|
-unobtrusive measure of behavior that uses physical evidence
-indirect way of studying behavior |
|
natural (field) setting
|
-environment were behavior happens naturally
-ecologically valid -if in lab, change behavior |
|
general rules for good observational research
|
-keep careful records
-use a variety of measures -respect privacy of individuals studyign |
|
Participant observer research
|
-observational research in which the observer participates in a group to record behavior
|
|
When is participant observer research useful?
|
-small groups
-seperate from society (cults) -groups which little is know |
|
What are the problems or limitations of participant observer research?
|
-difficult to gain membership (don't like outsiders)
-may not be ethical to incorporate self into group -dangerous -affect behavior w/ presence |
|
What should you consider before starting observational research?
|
-operational definitions of constructs (define constructs in term of procedures
-content analysis |
|
How can you answer the question of operational defintions?
|
-be systematic
-be selective of type of behavior studying -use technology -safe guard data (multiple copies) |
|
content analysis
|
-ways of interperting behavior
-ideally want both manifest and latent content analysis -also ideally want 2 or more coder for coding reliability |
|
manifest content
|
number of times a word appears in text
|
|
latent content
|
way a word appears in text
|
|
protocols
|
-script
-give script to ppl running expt to make sure its run same way w/ everyone |
|
pilot study
|
-small scale study
-begin study w/ one to see if get certain results |
|
Archival research
|
-done on files already out there
-do it b/c data already exists, may be more ethical, or make more legistical sense |
|
archival data
|
factual info in existing records
|
|
limitations on archival research
|
-data collection isn't controlled by researcher so data may be inadequate
-may be impossible to rule out other hypothesis |
|
meta-analysis
|
when many expts are run, you can run analysis on expts
-ie 9 out of 10 studies show... |
|
case studies
|
-observational and/or archival methods
-examine individual instances of some phenomenia (ongoing situation) |
|
characteristics of case studies
|
-researcher must be opertunistic
-usually in less than ideal situations so researcher doesn't have control over conditions -can't draw inferences bout casuality |