Study Guides (248,149)
United States (123,289)
Sociology (220)
Final

SO201 Final Review

8 Pages
93 Views
Unlock Document

Department
Sociology
Course
CAS SO 201
Professor
Emily Barman
Semester
Spring

Description
Final Review SO201 Causality Nomothetic: prove that change in IV lead to change in DV. If X, then Y. -criteria of causality: -association/correlation: when DV changes due to change in IV -time order: must show that change in DV occurred after change in IV; or IV changes before DV -non-spuriousness: when relationship between the IV and DV is NOT due to presence of third variable -causal mechanism: confidence in conclusion that two variables with causal relationship will be strengthened if causal mechanism (steps that link IV and DV) is identified Idiographic causality (qualitative): multiple ways from XY. Includes multiple and different causes of outcome for and across cases. Criteria of causality: True/classic experiment: where researcher is able to control value of IV across different cases in sample -manipulates IV by exposing different groups of cases to different values of IV -then measure value of DV in each group to see if DV changes as IV changes -3 parts of experiment: -two groups: treatment and comparison (no treatment) -random assignment -pre-test and post-test -limitations: randomization cannot always occur (ethics of studying child abuse), researchers can’t always manipulate value of the IV (variables of gender/poverty), pre- test/post-test can’t be conducted (effect of race or event-9/11-on DV) Alternative research designs & examples: -quasi-experimental: when researcher cannot use all 3 parts of true experiment -two-group post-test (only post-test of DV) -pre-experiments: no randomization and no treatment/comparison groups—hardest to establish three criteria of causality -one group pre-test post-test (no comparison group) -non-experiments: researcher can’t control value of IV (ex: naturally occurring) -longitudinal: gather data from cases more than once over time -cross-sectional: gather data from cases only at one point in time Example: Sherman and Berk: “The Specific Deterrent Effects of Arrest for Domestic Assault” -RQ: How does punishment affect rate of deviance? -hypothesis: labeling theory (punishing intensifies deviance) or deterrence doctrine (punishing deters deviance) -tested through true experiment -multiple groups: arrest (treatment), ordering away, advice -random assignment: calls of domestic abuse randomly assign treatment -pre-test: already know level of deviance before treatment -post-test: subsequent rates of repeat incidents -research design: when incidents of misdemeanor domestic assault occur, random application of 3 different intervention strategies: arrest, ordering offender from premises and advice. See how these different strategies affect rate of repeat incidents in 6 months -findings: found that arrest lead to lowest level of repeat incidentsdeterrence doctrine confirmed Surveys -a strategy to collect data from cases, characterized by a list of questions (generated via deductive method). Best for quantitative approach. Potential problems: -response rate: subjects’ refusal to participate and/or their incomplete participation in survey. Creates nonresponse bias -goal is response rate of 60%+ or 10 cases per variable -response error: responses inaccurate due to bad design of survey (multiple interpretations of the questions) -no way to prove it though because surveys are anonymous -these potential problems depend on type/design of survey, survey question construction, contact with respondents Types of Surveys (+/- of each): -mail surveys: mailing questionnaires to carefully selected sample respondents and requesting completion and mailing back results +: cheap -: low response rate, not best for asking detailed written responses, no chance for clarification -telephone surveys: interviews over phone and interviewer records response +: ensure clarification of question and response, can ask more complicated questions than mail survey -: not many have publicly-listed telephone numbers—creates sample bias -face-to-face interview: interviewer works directly with respondent, asks questions and writes down responses +: clarification of question and response, able to ask more complicated questions, able to encourage response -: not good for sensitive topics, interviewer bias -internet-based questionnaire: directed to website, complete survey, submit +: cheap, efficient, anonymous -: tough to find emails, low response rate, inability to ask complex questions Composing the Questionnaire: -type of question: -closed-ended: respondent asked to select from list of answers; must be exhaustive and mutually exclusive -open-ended: provide respondent’s own answer to question -screening/filter/contingency questions: sometimes must ask one question first to determine if they are qualified to answer subsequent questions—may cause data to be skewed if there is no filter question -wording: confusing/vague words, biased questions (accidentally word question so that response is largely controlled by whether respondent has similar background to designer of survey instead of accurately reflecting what the respondent feels about the topic), confusing phrasing (double negatives, double-barreled questions—2 questions 1 answer), and types of questions Designing the Survey: Principles: directions, order, grouping -clear directions: specify what is expected of respondents and how to answer questions at beginning of survey -order: leave threatening questions to end, group questions by theme, move from closed to open questions Before Administering the Survey: -ask other researchers to review survey, conduct pre-test/pilot survey to get feedback Contacting Subjects: -make sure introduce self, purpose of project, subject’s ethical rights -consider offering incentives -promise to share our results with respondents -allow them to provide additional info if they want -non-completers: reminders to do survey—up to 2 Example: Loftus: “America’s Liberalization in Attitudes Toward Homosexuality, 1973 to 1998” -RQ: Why do we see a decline in Americans’ attitudes or understanding about homosexuality? -hypotheses: Can changes in attitudes be accounted for by demographic changes in population? Are changing attitudes towards homosexuality embedded in larger cultural ideological shifts (attitudes towards GLBT pop. and/or willingness to restrict civil liberties of others)? -data and methods: -general Social Survey data from 1973-1998 -DV: index of questions (on morality of homosexuality and willingness to restrict civil liberties (free speech, university teaching, and removal of GBLT library book) of GLBT population -IV1 of changing demographics (survey questions on age, sex, education, race/ethnicity, religion, type of community, region) -IV2 of changing attitudes, questions on judgments on morality of extra/pre marital sex, restriction of liberties of atheists and Communists, political conservatism, support for school prayer and Biblical literalism (Bible is word of God) -findings: -while attitudes against homosexuality have declined since 1990, even greater decline in willingness to restrict GLBT civil rights -demographic changes—rising education—and increasingly liberal political beliefs explains about one-half of change in attitudes -author hypothesizes that other half of changes are due to success of GLBT social movement in changing attitudes towards homosexuality in and of itself Content Analysis -a quantitative research method for systematically analyzing and making inferences about the context of texts to learn about society. Data used: texts or media -goal: translation of texts into variables—capacity to generate valid and reliable measures of texts Unit of observation vs. unit of analysis (incomplete) Steps of content analysis: -mode of reasoning: employs the deductive method (RQtheoryhypothesis variable(s) of interest) -sampling: identify relevant population of texts, select sample from pop. (probability based sampling method) -measurement: figure out how to define concept, construct each variable and measure the value of the variable in each text -process of transforming raw data into a standardized and comparable form -steps: -choose unit of analysis/text units: what section of text will be coded for its meaning across the sample elements? -choose type of coding: selection of manifest or latent coding -manifest: presence of pre-specified list of particular words, phrases that convey presence of variables/attributes -latent: up to discretion of coder to identify meaning of text—more difficult, but more validity -design coding procedure: for each variable, generate list of attributes and/or indicators -create coding sheet -code texts: go through each unit of text and assign value for each variable -analysis of results Pros and Cons: +: economical, safety in case of error (reread texts), can study long periods of time and the past, no reactivity (influence on subjects in study) -: limited to recorded texts, don’t know intention of authors, time consuming, meaning of text may be unclear/open to multiple interpretations Field Research -a qualitative method for gathering data that is a study of a naturally occurring social unit: place, organization or informal group. Seek to inductively identify the culture of the place to see how things really occur by immersing self in local setting -idea that some features of social life can’t be validly studied through use of other data collection strategies Pros and cons: +: reliance on first-hand information; attention to context and multiple determin
More Less

Related notes for CAS SO 201

Log In


OR

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


OR

By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.


Submit