Reading...
Play button
Play button
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;
44 Cards in this Set
- Front
- Back
|
Qualitative research designs
|
-Participant observation - develop relationships with subjects as they go about normal activities
-Intensive interviewing - open ended, unstructured questioning -Focus groups - unstructured group interviews usually centered around specific topic of interest to the study |
|
|
Common features of qualitative research
|
-Data mostly qualitative
-Inductive reasoning; may be exploratory, do not test preformulated hypotheses but discover what people think and how they act in social settings -Allows focus on previously unstudied processes and unanticipated phenomena; new issues, hard to study groups, study the meaning people give to their lives -Focus on social context and interconnections -Focus on events leading up to particular event or outcome, rather than search for general causal explanations -Reflexive research design - design is adjusted as study progresses, based on what researcher learns -Focus on understanding from actor's point of view -Lots of data on few cases, study in depth and in detail -Attention to impact of researcher's and other's values on course of analysis -Naturalistic and uncontrolled methods |
|
|
Field Research
|
-Involves the study of human behavior, usually in social groups
-Direct observation of specific physical or social settings and the behaviors and events within them -May study informal groups and situations (gangs) or formal groups and situations (police officers) -Study behavior not amenable to simple quantification -Can also incorporate quantitative methods |
|
|
Participant observation
|
-3 types of roles based on degree of involvement
1. Complete observation 2. Participant and observation 3. Covert participation (complete participation) |
|
|
Complete observation
|
-Researchers tries to see things as they happen, without actively participating in or disrupting these events
-May be interaction with subject(s) of study -Subjects may not be aware they are being studied -Common when observing specific types of events and behaviors that occur in public -Ex. Susan Miller's study of community policing -But... will people act naturally knowing they are being watched? Reactive Effects OR research does not participate in illegal, unethical behavior but may observe it, report it? |
|
|
Participant and observation
|
-Involves some degree of active participation
-Some of the subjects aware -Allows for development of rapport -Ex. Whyte's study of gangs (established relationships, broke down social barriers, hung out with members, obtained info not otherwise available if only a complete observer) |
|
|
Covert participation
|
-Researcher operates as fully functioning member of the group or social setting
-People being studied are usually not aware they are being studied -"Going native" - researcher assumes identity of group and adopts ideas and views of subjects, dress, can result in loss of objectivity -Note taking is restricted, must rely on memory -Cannot ask questions that might reveal the fact that a research study is being conducted -Danger to researcher and others, if others suspect their identity or if researcher contributes to, impedes, group action these consequences can be adverse (change group dynamics or cause distrust of social scientists) -Ethical issues of engaging in illegal behavior or deception where subject participation is not voluntary (must determine if deception is necessary) |
|
|
Sampling in field research
|
-Not probability sampling
-Snowball sampling OR -Theoretical sampling: researcher learns of important processes during data collection and makes certain to sample enough people to investigate fully [experience sampling method (ESM) - subjects carry pagers and when paged, they fill out reports of what they are doing at the time] |
|
|
Techniques of qualitative data analysis
|
-Documentation
-Conceptualization, coding, categorizing (identifying and refining important concepts) -Examining relationships and displaying data (looks at the why aspect) -Authenticating conclusions (evaluating the validity or authenticity of conclusions; Becker's 3 Criteria or informant credibility, spontaneous answers or answers to researchers questions, presence/absense of researcher's influence on actions or statements) -Reflexivity |
|
|
Ethics in qualitative research
|
-Privacy, confidentiality, anonymity
-Intruding on people? -How identifiable are individuals/organizations studied? If confidentiality cannot be guaranteed, don't promise it -What does the researcher do when she sees harmful, wrongful, or illegal behavior during a study? -Research integrity and quality, IRBs help ensure this -Ownership of data and conclusions -Use and misuse of results |
|
|
Cause
|
-An explanation for some characteristic, attitude, or behavior of groups, individuals, or other entities (such as families, gangs, police depts) or for events
-Types: nomothetic (many cases) vs. idiographic (for one person or thing) |
|
|
Criteria for Causation
|
1. Variables must be correlated with one another (empirical association); must be observable, not assumed or believed
2. Cause must precede effect in time (time order); change in x occurs before change in y 3. Relationship between variables cannot be explained away by a third variable (nonspuriousness); example of crime and ice cream sales both increasing 4. Causal relationship strengthened by finding causal mechanism; the story about how x affects y must make logical sense 5. Causal relationship should be considered within context; must consider all the factors affecting an outcome |
|
|
Study designs
|
-One-shot case study: implementing a program and measuring its effects once, X O
-Before-and-after design: giving a pretest at start, O X O -Non-equivalent control group: uses a comparison group, O X O compared to O _ O -Randomized pretest-posttest control group (Classic experiment): random assignment to groups. R O X O vs. R O _ O |
|
|
Requirements for true experiments
|
1. Two groups (in the simplest case, an experimental and a control group)
2. Random assignment to the two (or more) comparison groups 3. Pre and post tests Ex. Prison classification and inmate behavior (some reclassified to medium security, others stayed, random assignment, looked at number of disciplinary infractions) |
|
|
Quasi-experimental designs
|
-Do not use random assignment (usually because not feasible given topic or population studied)
-Have less explanatory power and more problems with validity -3 major types: nonequivalent control group designs, before and after designs, ex post facto control group designs |
|
|
Nonequivalent groups designs
|
-Because groups not selected through random assignment, considered nonequivalent
-Comparison group selected to be as similar as possible to group that receives experiment -Individual matching: choosing individuals that match on several key characteristics and putting in opposite groups -Aggregate matching: general characteristics of control group match general characteristics of experimental group on key characteristics -Control group must be "similarly situated" to the experimental group |
|
|
Before and after designs
|
-No comparison group
-Time-series designs most powerful type (often called "interrupted time-series design), make before and after measures of some phenomenon then interrupt time by implementing an intervention, need many time points to see if trend holds (more than just 2 or 3) |
|
|
Ex post facto control group designs
|
-Identify treatment and comparison groups after the fact (after results have already occurred)
-Not really an experimental design |
|
|
Internal validity
|
-Conclusions reflect what actually occurred in the experiment
-Ability to yield valid conclusions is determined by comparability of experimental and control groups (why random assignment is best because only difference is the intervention) -5 basic sources of invalidity: selection bias, endogenous change, external events (history effects), contamination, treatment misidentification |
|
|
Selection bias
|
-Treatment and comparison groups are different at beginning or end of study
-Random assignment solves this problem at beginning of study IF it is done properly -At end groups can differ based on differential attrition (some types of people may be more likely to drop out - experimental mortality) |
|
|
Endogenous change on validity
|
-Internal change, within person
-Maturation: people are continually changing so can be problem if experiment takes place over fairly long time -Testing: process of testing and retesting will influence people's behavior (may become more sensitized to object of experiment) -Regression: naturally occurring fluctuations, somethings things just get better on their own regardless of intervention |
|
|
External events on validity
|
-Aka exogenous events
-History effect: event may occur during course of experiment that affects it such as OJ Simpson case bringing attn to domestic violence -Things happen in environment that changes results of experiment |
|
|
Contamination on validity
|
-Experimental and control groups communicate and control group gets some of the intervention or groups otherwise affect each other (aka diffusion)
-Can be intentional or unintentional -Compensatory rivalry: control group member works harder to make up for not having the intervention -Demoralization: control group member feels left out and becomes worse than they would have been without the study |
|
|
Treatment misidentification on validity
|
-Treatment does not cause outcome, but rather the outcome is caused by something that went on in the conduct of the experiment
-Expectancies of experimental staff -Placebo effect: people improve because they think they are getting something that will make them improve -Hawthorne effect: people may change because more attention is being paid to them |
|
|
Factorial surveys
|
-Used to study attitudes and beliefs
-Uses vignettes on topic of interest -Vignettes are randomly distributed across respondents and along characteristics of the vignettes -Respondents are then asked for likely responses to vignettes about different hypothetical situations -Treatment is different types of vignettes -Enhancing internal validity |
|
|
Solomon Four-Group Design
|
-Assigns subjects to at least 2 experimental groups and at least 2 comparison groups, but not all have both pre and post tests
-Examine whether testing influences outcomes by comparing pre and post test of control group 1 and control group 2 and comparing the post test of each experimental group 1 with experimental group 2, etc -Enhances external validity -Can also be randomized |
|
|
Repeated Cross-Sectional designs
|
-Data are collected at 2 or more points in time from a different sample selected from the same population
-General changes in population, usually not detailed information -Several snapshots strung together -Trend studies (several snapshots of the same thing taken at several times) -Least powerful design, may not be studying the same people each year |
|
|
Fixed-sample panel designs
|
-Study same group of people at several intervals, collect data from sample at time 1 then collect data from same people at time 2, etc
-Very expensive, rarely done because of attrition (esp in long studies), subject fatigue (drop out or dont provide valid info) |
|
|
Event-based designs
|
-Group of individual units who enter or leave defined population during specified time period (common starting point)
-Study this group over some period of time (prison reentry program, veterans of a specific war) -Similar to cohort studies where people enter or leave a defined population at the same time (ex. prison admissions in 95, followed for 5 years or everyone who entered university in 97) |
|
|
Features of survey research
|
-Versatility: can be designed to study almost any social issue, can learn about individuals and organizations
-Efficiency: data can be collected from a large number of people relatively inexpensively and quickly -Generalizability: most effective way to obtain information from representative sample of large population |
|
|
Types of questions
|
-Open-ended: allow respondent to fill in own answer
-Close-ended: aka fixed or forced choice questions, respondent chooses from list of possible responses, ordered responses or unordered |
|
|
Fence-sitting
|
-People who see themselves as being neutral, may skew the results if you force them to choose between opposites
|
|
|
Floaters
|
-Respondents who choose a substantive answer when they really don't know the answer
|
|
|
Likert Questions
|
-Usually framed as declarative statements - not questions
-React to statement by identifying intensity of feeling/belief/opinion -Strongly agree to strongly disagree with or without a neutral point |
|
|
5 Basic types of surveys
|
1. Mail (self-administered)
2. Group administered (self-administered) 3. Telephone 4. In-person 5. Electronic |
|
|
Ethical issues in survey research
|
-Dealing with sensitive subjects; self report of criminal or deviant behavior
-Web-based studies; may be designed to be anonymous but researcher cannot guarantee absolute anonymity -Harm; most harm arises from breach of confidentiality, some surveys may cause psych harm |
|
|
Secondary data analysis
|
-Use pre-existing data in a different way or to answer a different research question than intended by those who collected the data
-Surveys, official statistics, official records, historical documents -Most common sources are social science survey data collected in studies funded by feds and state govnt or official records maintained by govnt agencies for admin purposes |
|
|
Historical events research
|
-Making comparisons with other times and places
-Cross-sectional examination of something that occurred in the past -Purpose: develop general theoretical explanations of an event/process rather than compiling facts -May be a case study of important event to examine its potential long term importance (ex. Martinsville Seven found presence of capital sentencing disparities) |
|
|
Comparative methods
|
-Examine phenomena across societies
-Variability in scope; may be cross-national research, may study different social groups in single country, or differences across states or cities |
|
|
Content analysis
|
-Systematic, objective, quantitative analysis of message characteristics
-Analysis of the content of communication from sources |
|
|
Crime mapping
|
-Purpose is to illuminate relationship between some category of crime and corresponding characteristics such as poverty and disorganization across given locations
-Used to identify the spatial distribution of crime along with the social indicators such as poverty and social disorganization that are similarly distributed across areas |
|
|
Methodological challenges to secondary data
|
-Incomplete historical records (underrepresentativeness > bias)
-Different data collection systems and variable cross-national record keeping may influence -Differences in cultural/linguistic contexts -Meanings and concepts change over time -Linking current study constructs to original data -Sampling - availability, purposive |
|
|
Triangulation
|
-Using multiple research methods enables researcher to examine problem from several perspectives
-In-depth interviews (inductive) -Secondary data analysis (exploratory & descriptive) along with other methods |
|
|
Ethical issues in secondary data analysis
|
-Confidentiality not an issue because existing data from research studies is usually in public use files; identifiers removed and file has been created so that any researcher can access it without violating confidentiality
-IRB determines whether review is necessary |
