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;
94 Cards in this Set
- Front
- Back
|
The random sample |
Everyone in the population has an equal chance of being chosen for the sample |
|
|
Error of Estimation |
AKA Margin of Error the amount that the data obtained from the sample are expected to vary from the population The smaller the error, the better the sample ONLY FOR PROBABILITY SAMPLING-otherwise , you're violating too many assumptions |
|
|
What 3 factors is error of estimation a factor of? |
Sample size Population size Variance of the data |
|
|
4 types of probability sampling |
simple random systematic stratified random cluster |
|
|
Simple Random Sampling |
Random samp from the entire pop MUST have a sampling frame; this is what makes this type of sampling difficult Type of probability sampling |
|
|
Sampling Frame |
the name and contact information for everyone in your population Type of probability sampling |
|
|
Systematic Sampling |
Choosing "every so many" individuals for the sample Population is unknown Because you don't need a sampling frame, this form is easier and more accessible Type of probability sampling EX: Choosing every 8th person to walk through the door |
|
|
Stratified Random Sampling |
Similar to simple random in that you need a sampling frame and allows you to choose a random set of people from the whole population Allows you to set parameters!!!!! Divide pop into strata or categories, then randomly choose within these categories Uses proportionate sampling method; guarantees that you won't bias the experiment in a certain way EX: adjusting for the number of men vs women so its not biased for gender |
|
|
Proportionate Sampling Method |
Sample is chosen in proportion to its representation in the population used in stratified random sampling |
|
|
Cluster Sampling |
A method of whittling down population BEFORE sampling; makes population a manageable size Easiest and most practical probability sampling method; do NOT need a sampling frame Sample clusters of the population first, then work your way to the sample Multistage cluster sampling |
|
|
Multistage Cluster Sampling |
Type of cluster sampling, probability sampling Move from large clusters to smaller ones Making clusters more than once |
|
|
Threats to Representativeness |
Response bias Selection bias Misgeneralization |
|
|
Response Bias |
AKA the problem with nonresponse the idea that people deciding not to participate ruins the randomness of your sample, making it biased and increasing the error Why its important to do your best to get EVERYONE to participate |
|
|
Selection bias |
method of sampling procedures produce biased sample automatic in non probability samples and can happen in probability samples |
|
|
Misgeneralization |
attempting to generalize results based on unrepresentative/biased samples The reason people include the limitations section in their research reports |
|
|
Descriptive Research |
used to describe the characteristics or behaviors of a given population in a systematic and accurate way Involves 2 ways of collecting data: watch/record (observational and physiological measures) and surveys (questionnaires and interviews) |
|
|
Approaches to Observational methods |
1) Natural vs contrived settings (articficial setting made to mimic real-life scenarios) 2) Disguised vs undisguised observation (do they know you are watching them) 3) Different methods of behavioral recording |
|
|
Natural vs Contrived Setting approaches to observational methods |
Natural: naturalistic observation Contrived: Field studies |
|
|
Naturalistic Observation |
not contrived observing an individual/group in natural habitat; cannot speculate about motives Characteristics: very clinical, used to describe behavior; can be either disguised or undisguised |
|
|
How to achieve unobtrusive observation |
hide habituation (Jane Goodall; repeated exposure) participant organization (allows you to be unobtrusive) group infiltration (become active, participating member of the group) |
|
|
Field Studies |
contrived method of observation not naturalistic; creating a contrived condition out in the real world |
|
|
Methods of behavioral recording in observation |
narratives and field notes both are difficult to analyze quantitatively content analysis, checklists, temporal measures (latency and duration), rating scales, and physiological measures |
|
|
narratives |
method of behavioral recording in observational studies unstructured notes containing full descriptions of everything done and said typically on audio or video tapes |
|
|
field notes |
method of behavioral recording in observational studies more concise descriptions of behavior |
|
|
Content Analysis |
method of quantifying info from observational data (field notes and narratives) a procedure to convert textual information to more relevant, manageable data |
|
|
Steps of Content Analysis |
1) Identify units of text or analysis 2) choose coding approach (classification, rating, counting, etc.) 3) create coding rubric (must create specific rules; often rely on computer software packages that are already developed) 4) calculate interrater reliability (>.7) |
|
|
Checklists |
Method of behavioral recording possible behaviors are decided on in advance; structured description containing a tally of specific behaviors Must have very strong operational definitions! |
|
|
Temporal measures |
when and how a behavior occurs; always in units of time 2 types: latency measures and duration measures |
|
|
Latency measures |
the amount of time that nothing is happening the time elapsed either between two behaviors or between an event and a behavior EX: measuring reaction time |
|
|
duration measures |
the amount of time that something is happening how long the behavior lasts |
|
|
Rating Scales |
another type of behavioral recording used for measuring quality or intensity of a behavior EX: on a scale of 1 to 10, how often did you observe or how intensely was the person's response... |
|
|
Physiological Measures |
another method of behavioral recording measuring internal processes that are not readily visible Neural imaging and/or activity measures such as EEG, EMG, PET, MRI, fMRI Autonomic nervouse system Blood, urine, and saliva assays |
|
|
Types of Self Report Measures |
Cognitive self-reports, affective self-reports, and behavioral self-reports |
|
|
Guidelines for asking good questions: |
- be specific, precise wording - simple wording - no loaded words (words with pos/neg connotations that would suggest certain stances) - no assumptions about respondent - present conditional elements first - appropriate response format - no double-barreled questions - pretest the questions |
|
|
self-report response formats |
free-response format rating scale fixed-alternative format (either/or) true/false response format (form of fixed-alternative) |
|
|
Problems with self-report |
reliance on participants to tell truth, but this doesn't always happen; dealing with inaccuracy Faulty memory social desirability response bias different response styles |
|
|
Faulty memory |
problematic in self-report when people remember things wrong or they don't actually remember but still record a response |
|
|
Social Desirability Response Bias |
problematic with self-report response people don't want to be negatively evaluated, have the researcher think that you're a bad person people almost always over-report their nice qualities |
|
|
Different Response Styles |
problematic with self-report response when people don't know which option to choose, some people will tend toward yes while others tend to choose no; difference in response styles |
|
|
Ways to reduce problems with self-report measures |
Experience Sampling Methods (ESM) Diary Method Computerized ESM |
|
|
Experience Sampling Methods (ESM) |
Way of avoiding problems in self-report measures has participant report what they are feeling, thinking, doing at the time that they are doing it problematic bc you have to wait for your data to come in rather than collecting data all at once when you first want it |
|
|
Diary method |
Way of avoiding problems in self-report measures Relying on participants to record their activities daily on pen/paper questionnaires |
|
|
Computerized ESM |
way of avoiding problems in self-report measures use of palm pilots that beep when they are supposed to answer questions offers more structure |
|
|
How to minimize social desirability response bias |
Ask neutral questions ensure anonymity perform unobtrusive observation (as oblique, clinical as possible) |
|
|
How to minimize problem of different response styles |
include an equal number of true/yes questions and false/no questions for the same issue AKA: ask same question multiple times in opposite ways |
|
|
Interviews |
difficult to administer in un-biased manner Good interview: - interviewer should not create bias - friendly atmosphere - maintain neutral responses to participant - conceal personal reactions - create a logical sequence - don't deviate from written questions (consistency) - don't "lead" the respondent |
|
|
Types of Survey design |
Cross-Sectional Survey Successive Independent Samples Longitudinal Design |
|
|
Cross-Sectional Surveys |
Survey design style sample = a cross section of the population one group is tested at one point in time no major drawbacks, but must ensure measures are VALID and RELIABLE sample must be representative and the survey must be reliable/valid usually used to make frequency or association claims |
|
|
Successive Independent Samples |
Survey design style 2 or more samples answer the same questions at different points in time call one sample at one point, then a DIFFERENT sample at another point; often used in election polls Drawbacks: - the two temporally different samples might not be comparable - the variance may change over time, leading to interpretation issues |
|
|
Longitudinal Design |
Survey design method single sample is questioned more than once assesses how responses change over time Drawbacks: - difficult to keep in contact with people - attrition - not very time-effective |
|
|
Attrition |
one of the drawbacks in longitudinal studies selective attrition might influence results- a lot of people of similar characteristics drop out, causing you to lose that demographic |
|
|
Advantages and disadvantages of online surveys |
Pros: - inexpensive - easy to access so more likely to get a lot of participants - no manual data entry required Cons: - always will produce a more biased sample- biased toward people who are more likely to fill out internet surveys - limited sample - have to follow directions, but people don't always do this how we want them to; unreliable |
|
|
categories of research |
demographic research epidemiological research archival research |
|
|
Demographic Research |
describes patterns of life events, experiences typically conducted by sociologists; also useful in studying psychological impacts of major life events EX: marriages, birth, deaths, voting |
|
|
Epidemiological Research |
studying disease occurrence in different groups of people EX: type A ppl have more hypertension than type b; asian indians have more heart disease than any other group; men re more likely to be developmentally disabled than women |
|
|
Archival Research |
research done using data that was collected for purposes other than research or other than the purpose of the present study; using old data to answer new questions archives used might include observable behaviors (school records, marriage records), physiological processes (medical records), or self-reports (diaries, personal letters) Extremely useful in studying behavior change over time, repercussions of events/phenomenons after they occur that are impractical or unethical to reproduce |
|
|
3 Qualities of good descriptive data |
accurate concise understandable |
|
|
Types of Descriptive data reports |
Frequency distributions (how often) Measures of central tendency (typical score of data) Measures of variability (how much does one score vary from another; spread of scores) Standard deviation/the normal curve the z-score |
|
|
Frequency distributions |
type of descriptive data report summarizes raw data by showing # of scores within categories Simple FDs: # of participants who obtained each score Grouped FDs: subsets of scores |
|
|
Measures of Central Tendency |
Mean, Median, and Mode Relies on normal distribution |
|
|
Normal Distribution |
most scores fall at/around mean most behaviors in life ARE normally distributed |
|
|
Z-Scores |
identify how you do in comparison to everyone else; in relation to other participants how far a [articular score is fro the mean in standard deviation units Equation: (participant's score-mean of the sample)/standard deviation of the sample |
|
|
Correlational Research |
research that focuses on the degree to which two variables are related allow you to predict outcomes, but not imply causation Predictive values of a correlation are determined by Pearson's Correlation Coefficient; the strength of a relationship is determined by the absolute value of r |
|
|
Pearson's Correlation Coefficient |
determines the predictive value of a correlation AKA "r" indicates the degree to which two variables are related; will always be between 1 and 0 2 Important Aspects: Sign and Magnitude Can ONLY be used in linear relationships CORRELATION COEFFICIENTS ARE NOT ON A RATIO SCALE |
|
|
Coefficient of Determination |
statistic that tells you how much of the variance in variable A can be explained by variable B; calculating for systematic variance How much are these two REaLLY related Created to turn Pearson's r (not ratio, but interval scale) into a ratio system |
|
|
Systematic Variance |
the amount of variance in one variable that can be explained by the other variable; the correlation they SHARE r^2 By squaring Pearson's correlation coefficient, we turn it into a ratio scale number AKA coefficient of determination |
|
|
Factors that distort correlations |
restricted range, outliers, and unreliable measures |
|
|
Restricted Range as a distorting factor in correlational research |
Age range/scope plays big role in determining correlation accurately Especially important to consider in curvilinear data EX: strong positive correlation between age and foot size, but when ages are restricted, it might not appear to be so |
|
|
Outliers as a distorting factor in correlational research |
outlier= score that is at least 2 standard deviations away from the mean can lead to either/both on-line outliers and off-line outliers |
|
|
on-line outliers |
exaggerate a nonexistent correlation |
|
|
off-line outliers |
make correlations appear weaker deflate or reduce an existing correlation |
|
|
Unreliable measures as a distorting factor in correlational research |
lead to lower correlation than is true allows variability that doesn't belong; introduces erroneous variability, resulting in "noisy data" Can't determine what is real/important, which scores are true; obscures the relationship between the two variables |
|
|
Third Variable Explanation |
one of the reasons we can't imply causation from correlation; the true cause could be something your weren't even looking at- the third variable |
|
|
Multivariate research |
used to determine cause; manipulation of variables while holding all others constant Correlational techniques: longitudinal designs, MR Analysis, Pattern and Parsimony Approach |
|
|
3 Causal Criteria in Multivariate Research |
1) is there covariance? 2) is there temporal precedence? 3) is there internal validity? |
|
|
Multivariate Longitudinal designs |
can provide evidence for temporal precedence by measuring the same variables in the same people at several points in time Multivariate because we are interested in multiple variables and their relationships to eachother |
|
|
Correlations Examined in Multivariate Longitudinal Designs |
Cross-sectional correlations Autocorrelations Cross-Lag Correlations |
|
|
Cross-Sectional Correlations |
investigates covariance correlations of the two variables when measured at the same time relationship between (V1T1 and V2T1) and (V1T2 and V2T2) |
|
|
Autocorrelations |
investigates the stability of traits or behaviors over time making sure that each individual variable is stable over time (V1T1 and V1T2) with (V2T1 and V2T2) |
|
|
Cross-Lag Correlations |
Investigates which variable occurs first; establishes temporal precedence examines correlation diagnally/between var 1 at beginning with var 2 at end as compared to var 2 at beginning with var 1 at end by comparing the two r values, we can see if can see if one var precedes the other (which correlation is strong?) EX: if r is high for TV violence in 3rd grade and aggression in 13th grade but not so much for Aggression in 3rd grade and TV violence in 13th grade, we can assume that TV violence in 3rd grade predicts AKA has temporal precedence over aggression in 13th grade AKA |
|
|
Limitations of Multivariate Longitudinal Designs |
still has low internal validity haven't yet ruled out alternative explanations and possible 3rd variable explanations! |
|
|
Multiple/Multivariate Regression Analysis |
used to rule out alternative explanations/3rd variables that may possibly account for the correlation increases internal validity This statistic evaluates the relationship between the 2 variables of interest while holding a third variable constant |
|
|
Multivariate Research: Pattern and Parsimony |
investigating causality by using a variety of correlational studies that all point in the same causal direction allowing explanation that most parsimoniously explains the ENTIRE pattern of data across SEVERAL STUDIES |
|
|
3 steps of determining causation through pattern and parsimony |
1) specify theory/mechanism for a causal path 2) make specific predictions from the theory 3) choose most parsimonious explanation that explains ALL of the data |
|
|
Example of a pattern and parsimony result |
all of the predictions (stemming from the one correlational theory) listed are supported by data from many correlational studies and no one single 3rd variable can explain ALL of the results as easily/parsimoniously as the ONE--- this is your cause. |
|
|
Mediators |
an explanatory variable that is used to show HOW variable A may be causing variable B; shows the why between two variables, like physical exercise between recess time and classroom behavior problems EX: In a relationship of recess time being negatively associated with classroom behavior problems, it is said that recess gives chance for physical activity and that physical activity tires kids out in order to reduce behavior problems in the classroom; physical activity would be the mediator |
|
|
5 Things for Showing Mediation |
1) test for relationship between recess (V1) and behavior problems (V2); typical correlation 2) " recess(V1) and physical activity (mediator) 3) " physical activity (mediator) and behavior problems (V2) 4) run a multiple regression, looking at the relationship between recess (V1) and behavior problems (V2) when controlling for physical activity (mediator) 5) establish temporal precedence |
|
|
Difference between testing for mediation vs for moderation |
Mediation: looking for the WHY of the correlation; Why are the variables linked? Moderation: looking for UNDER WHAT CIRCUMSTANCES for the correlation; Are the variables linked the same way for everyone, or in every situation? |
|
|
Relationship between mediation, moderation, and their-variable problem |
|
|
|
Goals of Sampling |
achieve external validity attain a sample that is representative of the population economic sample |
|
|
Types of Nonprobability Sampling |
Convenience sampling Quota Sampling Purposive sampling |
|
|
Convenience Sampling |
choosing people for your sample based on proximity and willingness easiest, quickest Problem: likely not very representative |
|
|
Quota Sampling |
similar to convenience; steps are taken to ensure representativeness- certain number/percentage of people from a particular group are made sure to be included in the sample |
|
|
Purposive Sampling |
choose certain pockets of people to be in your sample that, in the past, have reflected the views of the whole population EX: election polls |
