Term
|
Definition
| a conjectural statement, thought to be true but which can never be (dis)proven, of the relationship between two variables and logically implied by a proposition. |
|
|
Term
|
Definition
| a set of interrelated propositions or empirical generalizations that link two concepts together to help explain phenomena, organize knowledge and derive hypotheses. |
|
|
Term
|
Definition
1. simplicity/parsimony
2. internal consistency
3. testability/falsifiability
4. predictive accuracy
5. generality |
|
|
Term
|
Definition
observations -> empirical generalizations -> theory
challenge existing knowledge but can lead to false conclusions through assumptions of time order and causality |
|
|
Term
|
Definition
axiom/assumption -> proposition -> hypothesis
if axioms are accurate, allow from strong causal inferences and are rarely false conclusions, but axioms can be restrictive |
|
|
Term
|
Definition
| acquisition of knowledge through the application of the scientific principles of empiricism, intersubjectivity, explanation and determinism (EDIE) - always falsifiable |
|
|
Term
| hallmarks of the scientific method |
|
Definition
| EDIE - empiricism, determinism, intersubjectivity and explanation |
|
|
Term
|
Definition
| knowledge claim be based on systemic observation - assuming most accurate/reliable information. attaining information through senses helps guard against bias |
|
|
Term
|
Definition
replicability: clear enough that someone can redo and get same results
transmissible: people see how we arrive at our conclusion. Guard against bias **not objectivity** |
|
|
Term
| explanation (scientific method) |
|
Definition
| goal of scientific method - political phenomena are explained by how they relate to something else. looks at explaining recurring patterns. |
|
|
Term
|
Definition
| implies that there are recurring regularities in political behaviour. assumption that cannot be (dis)proven and is valid insofar as research withstands empirical testing |
|
|
Term
| compare logic to scientific method |
|
Definition
make observations -> make criteria in advance through observation
make a conclusion -> tests for plausible alternatives
ignore confounding evidence -> revises explanations |
|
|
Term
|
Definition
| refers to anything that is directly or indirectly observable and is given meaning by people. provide the basis for classification, comparison and quantification |
|
|
Term
|
Definition
| sort phenomena into categories that are exhaustive, mutually exclusive and preferably have a comparative concept |
|
|
Term
|
Definition
| allows you to order phenomena - more or less of a property |
|
|
Term
|
Definition
| the ability to measure how much property is present |
|
|
Term
|
Definition
| provide the empirical grounds for testing a concept - operationalization. How we move from theoretical to empirical realms. |
|
|
Term
|
Definition
| the operationalization of a variable. |
|
|
Term
|
Definition
comparison, manipulation, control: covariation, time order, spuriousness
** correlation is not causation** |
|
|
Term
| Internal threats to validity |
|
Definition
extrinsic: selection bias (not generalizable, sacrificed for intrinsic)
instrinsic: History, Regression, Reactivity, Instrumentation, Mortality, and Maturation |
|
|
Term
| external threats to validity |
|
Definition
| extrinsic: unrepresentative cases, reactivity and artificiality of the research setting. |
|
|
Term
|
Definition
| TG(PoT - PreT)- CG(PoT - PreT) |
|
|
Term
|
Definition
| does the measurement instrument perform consistently? MANDATORY FOR VALIDITY |
|
|
Term
|
Definition
| are we measuring what we think we are measuring? |
|
|
Term
|
Definition
| values given to observations are attributed to flaws in the measurement process |
|
|
Term
| systemic measurement error |
|
Definition
| a form of bias, can be controlled for later. jeopardized validity. |
|
|
Term
|
Definition
| occurs something transient, jeopardized validity and reliability. cancel out. |
|
|
Term
|
Definition
| focuses on content, how appropriate (face validity)are your indicators, and how completely (sampling validity) do they encompass the full meaning of the concept |
|
|
Term
|
Definition
| set up auxiliary hypothesis with different variables and indicators representing the same concepts. You should get the same result. |
|
|
Term
| probability (random) sampling |
|
Definition
| every individual in the population has an equal chance of being included in the sample |
|
|
Term
|
Definition
| the researcher chooses who is included and there is no guarantee that every member has an equal chance (or even a chance) of being included |
|
|
Term
|
Definition
| every member of a population has an equal chance of being included in the sample: avoids bias and is generalizable. need population list, can use resources and produce extreme samples |
|
|
Term
| multi-stage random cluster sampling |
|
Definition
| random selections of groupings of a population from which you randomly select members. reduce cost of sampling dispersed populations, obviates need for population list. Increases risk at every stage of selection. either increase population size or stratify within clusters. |
|
|
Term
|
Definition
the wording of the response is not fixed
4/5 |
|
|
Term
|
Definition
| the wording of the response is fixed 7/5 |
|
|
Term
| open-ended advantages/disadvantages |
|
Definition
A: control, rapport, new ideas, no putting ideas into respondents heads, rich contextual understanding.
D: time consuming, interviewer bias, irrelevant answers, comparability, categorization might lose meaning, replicability. |
|
|
Term
| closed ended advantages/disadvantages |
|
Definition
A: comparability, time, relevant, replicability, no interviewer bias, control, shy can speak.
D: put ideas into people's heads, answer without true opinion, not include all relevant categories, loss of rapport, limited contextual understanding. |
|
|
Term
| ethical principles of research |
|
Definition
| no deception, no harm, voluntary and informed consent. DHVI |
|
|
Term
|
Definition
| competence, voluntarism, full information and comprehension. CIVC. |
|
|
Term
|
Definition
| direct observation of political behaviour in its natural setting - data collecting and theory building |
|
|
Term
|
Definition
A: flexible, feasible, cost, understanding, external validity, immediacy.
D: reactivity and privacy |
|
|
Term
|
Definition
A: no reactivity
D: ethics |
|
|
Term
|
Definition
A: no reactivity
D: ethics, and people might react to someone not participating. |
|
|
Term
|
Definition
| attempt to mimic the logic of statistical control when sufficient data is not available. method of agreement, and method of difference |
|
|
Term
|
Definition
| pick cases similar on the IV and DV, but different in potentially confounding variables (controls) |
|
|
Term
|
Definition
| pick cases that have the potential causal factor and response (IV and DV), to cases that do not have the predicted causal factor or response, but are similar on potentially confounding variables. |
|
|
Term
|
Definition
lieberson:
1. deterministic logic, 2. assuming no measurement errors, 3. existence of 1 cause 4. absence of interaction effects, 5. don't always have these cases, 6. external validity jeapardized by small N, 7. too many variables - (deterministic measurement causes real interaction of valid variables) |
|
|
Term
|
Definition
| lijphart: 1. powerful to test theories, 2. problems of measurement reliability and validity are reduced, 3. might be only only way to validate causal hypotheses about macro-phenomena. |
|
|
Term
| case studies: combat critiques |
|
Definition
| increase N: geographic proximity, diachronically study one country, sub-national groups within a nation or combine similar variables or focus on only the key ones. |
|
|
Term
|
Definition
| is a technique for making inferences by systematically and objectively identifying certain aspects of communications that are available and recorded. who, said what, to whom and to what effect? |
|
|
Term
| systematically (content analysis) |
|
Definition
| all content is included (or not) based on a set of consistently applied criteria |
|
|
Term
| objectively (content analysis) |
|
Definition
| or intersubjectivity is the identification of content according to specific rules - different people will get the same results. |
|
|
Term
|
Definition
A: cheap, insightful, generalizable and no reactivity, look at history or inaccessible political figures, few ethical concerns.
D: criteria is subjective, so there can be problems with validity - are they representing the full concept? if you take a sample, there will be a relative degree of error. |
|
|
Term
| types of content analysis |
|
Definition
| latent (leaning), substantive (what was said), coding manifest (# of words) a structural (space, size, columns). (LSSC) |
|
|
Term
| statistically testing a hypothesis |
|
Definition
| statistical significance (chi-square), the co-variation (measure of association) |
|
|
Term
|
Definition
| variables vary in a consistent and patterned way |
|
|
Term
|
Definition
| show how strongly our variable vary - how much we reduce our error when guessing the DV, if we know values on the IV. Oridnal - gamma or tau, nominal - lambda or cramers V |
|
|
Term
| proportional reduction in error |
|
Definition
the amount we reduce error when guessing values on the DV by knowing values on the IV
errors without knowing IV - errors with knowing IV/ error without knowing IV |
|
|
Term
|
Definition
| measure of association for nominal level statistics - using modal category, guess number of mistakes you would make, then subtract number of mistakes you would make knowing IV. It will be 0 if modal categories are the same for IV - Cramers V |
|
|
Term
| measure of central tendency |
|
Definition
| helps to determine the most common/typical value of your data. how typical is the value? do the variables vary? is there a lot of dispersion? |
|
|
Term
|
Definition
| the amount to which variables vary in a data set - the extent to which variables are concentrated in a few categories or dispersed (tells accuracy of central tendency) |
|
|
Term
| central tendency for levels of measurement |
|
Definition
nominal - mode (variational ratio equation V = 1 - fmodal/N
ordinal - median (range or interquartile range, middle 50th percetile)
ratio/interval - mean (sensitive to outliers, so compare to median which looks at relative position - less affected by outliers) |
|
|
Term
|
Definition
N = ZCI2S2/E2
must be random, more than 30+, S = level of variation in population **size of pop doesn't matter unless more than 5% of total population** |
|
|
Term
| confidence interval equation |
|
Definition
| % = (sample mean) +/- (ZCI)(standard deviation of sample which is caluclated by √s/n) |
|
|
Term
|
Definition
| a range of values within which we have a predetermined level of confidence that the true population value lies within. Sacrifice accuracy for confidence. |
|
|
Term
|
Definition
| observed relationship within a sample is not generalizable onto the population (use inferential stat to find this out) |
|
|
Term
|
Definition
| there is a relationship in the population that was not in our sample. |
|
|
Term
|
Definition
1. create null hypothesis
2. calculate expected frequencies
3. compare expected to observed
4. partial adjustment for sample size (larger sample, less risk of type 1 error)
5. calculate degrees of freedom - more cells more opportunity to deviate from expected frequencies (#columns -1)(#rows-1)
6. consult chi-square distribution chart |
|
|
