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;
70 Cards in this Set
- Front
- Back
content analysis
|
the study of recorded human communication (books, newspapers, websites, journals, video games, etc.)
-qualitative & quantitative components |
|
topics of content analysis
|
often focuses on popular culture & power relations; can address wide variety of research questions
|
|
steps of content analysis
|
1. identify a population of documents or other sources
2. determine the units of analysis (social artifacts) 3. select sample 4. design coding procedures for variables to be measured 5. train coders & conduct pilot study 6. code & analyze data |
|
coding
|
how you will transform raw data to standardized form; coding instructions must be as explicit as possible to insure inter-coder reliability
|
|
2 kinds of coding
|
latent: underlying meaning
manifest: visible, surface content |
|
strengths of content analysis
|
can be economical, limits researcher effects on what is being studied (qualitative interviews), can permit analysis of processes across time (ex: examining changes in cartoons over time), can be somewhat easy to correct for mistakes making it very flexible
|
|
weaknesses of content analysis
|
limited to examination of recorded communications; validity problems (b/c of context, may impose meaning on it; intercoder reliability)
|
|
goal of experiments
|
establish cause & effect between 2 variables
|
|
who participates in experiments
|
usually less than 100 people; usually college students
|
|
experiment process
|
begin w/hypothesis; modify situation; compare outcomes
|
|
IV
|
treatment (stimulus or manipulation)
|
|
DV
|
outcome
|
|
pre-test
|
measure of DV before IV
|
|
post-test
|
measure DV after IV
|
|
experimental group
|
group that receives treatment
|
|
control group
|
group that does not receive treatment
|
|
random assignment
|
assigning cases to groups
|
|
clearly defined DV & manipulable IVs
|
specific hypotheses about outcome & cause, observable in a controlled space; IV (cause) must be controllable; basic logic: create an experimental treatment/stimulus & observe if it has hypothesized effect
|
|
2 or more groups to compare
|
experimental group given treatment/experiment stimulus; control group given placebo or diff treatment; compare outcomes between experimental & control groups
|
|
randomization
|
random assignment to groups (not random sampling)
|
|
methods to establish equivalency
|
randomization: number respondents & assign groups based on odd/even numbers, w/o it can't rule out selection effects, can still have error due to chance (minimized by larger samples or repeated experiments)
-matching on imp causal characteristics (usually w/randomization, can be multiple factor matching) -pretest on outcome (sometimes): usually w/matching & randomization |
|
threats to internal validity
|
1. history: uncontrolled events coinciding w/the experiment & affecting the outcome
2. maturation/aging: subjects can naturally age & change if enough time has elapsed 3. testing: especially pretest may influence outcome: memory, social desirability, learning 4. experimental mortality: subjects drop out of study before completing 5. unmeasured group differences: chance diff even w/randomization 6. contamination: threat that participants will communicate w/one another 7. instrumentation: any change in the measurement procedures or devices |
|
threat to external validity
|
hawthorne effect: reacting differently b/c one is participating in a study
|
|
how to avoid hawthorne effect
|
double blind; field experiments
|
|
field research
|
study of people acting in natural course of daily lives (aka participant-observation research)
|
|
ethnography
|
describing people/culture & understanding people from another point of view; subset/extension of field research; explicit & tacit knowledge
|
|
features of qualitative fieldwork
|
natural settings & behaviors; interested in meaning (focus on process); researcher is primary instrument for data collection & analysis; flexibility; inductive theory construction
|
|
steps in field research
|
1. prepare oneself, read literature
2. select field site & gain access 3. enter field & establish social relations w/members 4. adopt a social role, learn the ropes, & interact w/members 5. watch, listen, & collect quality data 6. begin to analyze data & to generate & evaluate working hypotheses 7. conduct field interviews w/members & informants 8. disengage & physically leave setting 9. complete analysis & write research report |
|
complete observer
|
nonparticipant observation; observe w/o becoming a part in any way; researcher hides true identify; only can observe "public" info (physical signs, expressive movements, use of space, etc.)
|
|
observer as participant
|
reveal researcher status but observer activities are overriding; researcher intentionally draws attention to status as observer; relationship w/informants more formal
|
|
participant as observer
|
reveal researcher status, but role emphasizes participation; others relate to researcher in the participant role (need to build rapport); researcher participates more than observes
|
|
complete participant
|
conceals the observer role while becoming fully accepted into membership; hides true identity; used to study subcultures; danger of going native; ethical considerations
|
|
The General Social Survey (GSS)
|
conducted annually by NORC; data collected using a full-probability personal-interview survey; topics related to social change in the US w/various years
|
|
3 basic steps in GSS
|
data preparation; describing the data (descriptive stats); testing hypotheses
|
|
descriptive stats
|
measures of central tendency; mean, median, mode; inferential vs. descriptive stats
**modes used for nominal data, medians for interval data, & means for ratio data, not for nominal data |
|
dispersion
|
spread of the values around the central tendency
|
|
2 measures of dispersion
|
1. range: how values are spread
2. standard deviation: how much the typical data point differs from the mean, or average, data |
|
statistical significance
|
means that results aren't likely due to chance; tells us what is likely; tells us how likely results are to recur if we repeated the study through the p-value
|
|
p-value rules of thumb
|
when a test stat is significant = differences larger than would expect by random chance
small sample: if significance level <= .10 (1 in 10 chance) large sample: if significance level <=.05 (1 in 20 chance) very large sample: if significance level <= .01 (1 in 100 chance) |
|
.10 level means?
|
results are to chance factors only 10 in 100 times; 90% chance that the sample results are not due to chance alone (reflect population accurately); the odds that results are due to chance alone is 10%; we are 90% confident that results are due to a real relationship in the population & not chance factors
|
|
dealing w/potential harm
|
disclose who you are & what you're doing; voluntary participation; anonymity & confidentiality; avoid high risk topics & participants
|
|
informed consent
|
allowing the participant's voluntary participation to be based on a full understanding of the possible risks involved
|
|
confidentiality
|
when a researcher can identify a person's responses, but promises not to do so publicly (e.g., using pseudonyms for interviewees)
|
|
anonymity
|
neither the researcher nor the readers can identify a response w/the respondent
|
|
lit reviews
|
the way we learn what's already known & not known; in order to see where we're going, we must understand where we've been
|
|
steps in reading
|
1. read abstract
2. check the conclusion 3. skim article; pay attention to section headings, tables, graphs & structure 4. read the whole article |
|
functions of scientific reporting
|
communicate a body of specific ideas that is clear & detailed enough to be evaluated by others; contribute to the general body of scientific knowledge; stimulate & direct further inquiry
|
|
plagiarism
|
presenting someone else's words or thoughts as though they were your own, constituting intellectual theft
|
|
unobtrusive research
|
methods of social research that study social behavior w/o affecting it; these techniques can involve both qualitative & quantitative methods of data collections & analysis
|
|
3 major types of unobtrusive research
|
1. analyzing existing stats: using available official/quasi-official stats; you aren't re-working the stat analyses (secondary analyses), rather you are working w/their stat findings
2. comparative/historical research: examination of societies over time & in comparison w/one another 3. content analysis |
|
appropriate questions for experiments
|
purpose of research (descriptive, explanatory, exploratory); manipulable; breadth (broad or narrow scope or scale)
|
|
12 steps of an experiment
|
1. being straightforward w/hypothesis
2. decide on experimental design 3. decide how to introduce treatment (IV) 4. develop reliable & valid measure of DV 5. set up experimental setting & set up pilot test of the treatment & DV 6. locate appropriate subjects 7. randomly assign subjects to group & give careful instructions 8. gather data for pretest measure of DV for all groups 9. introduce treatment to experimental group only 10, gather data for posttest measure of DV 11. debrief subjects (esp crucial if using deception) 12. examine data collected & make comparisons between groups |
|
iterative process of inductive research
|
starting point (open-ended research questions about specific social phenomena or events); observe people & behavior; find patterns & draw initial conclusions; more observation & revision of ideas; interpret in light of researcher's theoretical paradigm
|
|
double blind experiment
|
an experimental design in which neither the subjects nor the experimenters know which is the experimental group & which is the control group; guards against experimenter bias
|
|
randomization
|
technique for assigning experimental subjects to experimental & control groups randomly
|
|
matching
|
in connection w/experiments, the procedure whereby pairs of subjects are matched on the basis of their similarities on one or more variables & one member of the pair is assigned to the experimental group & the other to the control group
|
|
internal validity
|
refers to the possibility that the conclusions drawn from experimental results may not accurately reflect what went on in the experiment itsel
|
|
external validity
|
refers to the possibility that conclusions drawn from experimental results may not generalize to the "real" world
|
|
stages of social research
|
1. formulation of research prob
2. preparation of research design 3. measurement 4. sampling 5. data collection 6. data processing 7. data analysis & interpretation |
|
data prep
|
cleaning & organizing data for analysis; checking raw data for accuracy; codebook (defines meaning of numbers & codes; age, homosex, marhappy, suicide); data transformations (missing values; re-codes)
|
|
univariate analysis
|
examination across cases of 1 variable at a time; descriptive stats (describe basic features of data)
|
|
3 major characteristics of data to examine (univariate analysis)
|
1. distribution: how many times various attributes of a variable are observed in a sample
2. central tendency: average or typical value 3. dispersion: how values are distributed around central value (S.D. -amt of variability in data) |
|
frequency tables
|
summarizes the distribution: the share of the sample associated w/each attribute; doesn't work well w/ratio variables
|
|
histograms
|
displays frequency distribution graphically; useful for assess whether data are normally distributed; can be used for any level of measurement; for variables w/many categories, may want to collapse the variable to create fewer categories
|
|
continuous variable
|
increases steadily in tiny fractions (ratio variables like age)
|
|
discrete variable
|
jumps from category to category w/o intervening (gender, military rank, etc.)
|
|
standard deviation
|
tells us about the spread of values around the mean; calculated as the square root of the mean of squared deviation scores; for normal distributions, the most stable measure dispersion
|
|
bivariate analysis
|
testing 2 variables together; cross tab & chi-square, t-test, one way ANOVA, correlation; level of measurement of variables determines which to use
|
|
assessing potential for harm
|
methodologies: use of public info vs. gathering new data; topics: public life vs. private matters; beliefs, attitudes, & values vs. behavior; stigmatizing or illegal behavior
|
|
IRB
|
established by fed gov; goal is to protect rights & welfare of human subjects participating in research
|