RESEARCH METHODS MID-TERM
Terms
undefined, object
copy deck
- What are the 4 criteria for Causality?
-
1. Temporal Order
2. Correlation
3. Non-Spuriousness
4. Statistical Significance - Something that acts through another variable
- Indirect causation
- What is a Null Hypothesis?
- "There is no relationship between X and Y."
- Turning a hypothesis into something we can measure
- Operationalization
- Each single unit that is being observed
- Case
- Type of group of the case we are observing (ex: individuals, states, etc)
- Unit of Analysis
- What are the 3 Levels of Analysis? Describe.
-
1. Nominal- creating categories
2. Ordinal- Order of the values in the cases
3. Interval- Most level of info; standardized unit/# = most specific - The 2 types of Error are Random Error & Systematic. Describe them.
-
Random- truly random; not systematic; cancel each other out
Systematic- a huge problem; effects cases in same way - What is validity?
- How close you come to true value a case holds.
- When the values of multiple tests are the same. Get the true score of value several times. If measurement is unreliable = not valid. Can be reliable without being valid (hit dart in same area several times, but sight/aim is off)
- Reliability
- Get control without having control groups. (Ex post facto: can be gathered after the fact or event)
- Quazi-Experimental Design
- What we do when we have way too many cases we could possibly study. *A subgroup of the cases in a population (all the cases you can study)
- Sampling
- Select a sample so that you can be _____ of all the cases in a population. If no _____, it will only apply to a sample, not the population.
- Representative
- If sample drawn randomly, 2 main things have to be met:
-
1. Each case in pop has same chance to be selected into sample.
2. Each set or combo of cases has same chance to be selected - Oversample based on some perimeter.
- Weighted sampling
- How do you know how large a sample to use in a population?
-
-How like a population is
-Amt of sampling error or margin of error that you're willing to take - *95 times out of 100 that estimates will be within the margin of error (& will not be a biased sample) - How you feel that the sample you got from the population is representative of the population.
- Confidence Level
- What are the 3 types of Research?
-
1. Descriptive
2. Exploratory
3. Explanatory - Seek an explanation for some phenomenon (type of research)
- Explanatory
- Describing different things you have (research)
- Descriptive
- You have a question, but no info about it(research)
- Exploratory
- Steps of the Scientific Method
-
1. Identify research ?
2. Make observation
3. Formulate tentative model
4. Do literature review
5. Finalize model
6. Formulate specific hyp.
7. Operationalization
8. Research Design
9. Gather data
10. Analyze data
11. Test theory
12. Refine theory - Framework for how you view the world.
- Theory
- Concept that is specific in nature and it varies
- Variable
- Measurements we can make of variables
- Indicators
- 2 variables that increase together (%voting & age)
- Positive relationship hyp
- As age increases, voting decreases
- Negative relationship hyp
- Variables that change together; necessary but not sufficient
- Correlation
- Variable A causes Variable B
- Causation
- Force, event, change in what cause dependent variable to move
- Independent Variable
- Variable being acted upon
- Dependent variable
- Applying a measurement in order to assign values to the different variables
- observation
- Characterized according to different levels of analysis
- Values
- How specific your values are when charaterized - how they are measured
- Levels of analysis
- Give test to different groups, get same responses
- Subsample method
-
1. purpose
2.statement of hyp
3. specific ways variable will be measured
4. detailed statement of how observations are collected - Research Design
- Used to rule out spuriousness
- Rival hypothesis
- How close you come to the value
- Validity
-
Asks "Are you sure your research design provides good evidence for your hypothesis?"
**Do the variables really measure the concepts?** - Internal validity
- Asks "How generalizable are your results? (significant relationship)"
- External validity
- People's behavior change, react to stimuli
- Reactivity
- Control groups need to be as similiar as possible to experiment groups
- Crucial point
- Indicator for independent variable
- Perimeter
- Systematic Random Sampling =
- Phonebook
- Lee County, City of AU, Dean Rd., house on Dean Rd.
- Cluster sampling
- Intentional selection of cases
- Judgemental sample
- What are the 3 kinds of Surveys?
-
1. Mail
2. Phone
3. Interview - What survey has the lowest response rate?
- This type of question lets the respondent create own answer (used in interview-to clarify ?s)
- Open-ended ?s
- This type of question forces respondent to make a choice
- Close-ended ?s
- What are the 4 types of info gathered from surveys?
-
1.attitudes
2.beliefs
3.behavior
4.attributes - What survey takes the longest?
- Interviews
- What survey takes the shortest time?
- A learning experience from the pretest that teaches you something that helps you score better on the post test.
- Test effect