Research Midterm Week Adam Eckered
Research Midterm Week Adam Eckered Pols 201
Popular in Research Methods in Political Science
Popular in Political Science
This 9 page Class Notes was uploaded by Lauren Jones on Sunday March 13, 2016. The Class Notes belongs to Pols 201 at University of Tennessee - Knoxville taught by Adam Eckerd in Winter 2016. Since its upload, it has received 36 views. For similar materials see Research Methods in Political Science in Political Science at University of Tennessee - Knoxville.
Reviews for Research Midterm Week Adam Eckered
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 03/13/16
Memo 3 Data Collection Strategies o Are international NGO’s are effective at providing assistance in poor countries? Three different approaches Causal design o Secondary data from a variety of different sources (WHAT WE ARE DOING) Quasi experimental o Natural experiment with different approaches Case study o In depth process Theories o Commitment and experience The longer aid agencies work in an area, the more effective the assistance is o Direct or indirect o Collaborative economic development Actually working closely and collaboratively with recipeients, aid can be provided in a way that best suits each person Literature Review o What does it mean to be effective No clear way to measure that What is ineffective and effective Key concepts and indicators o Ngo behavior Broaden in efficiency Strategy/change theory Reach Who helping, how they are helping them Collaborative effort The extent to which they are collaborating with, or are they swooping in o Assistance Look at short term reach Medium term sustainability o Select ways to think about these projects (WHAT WE ARE DOING) Causal research design o When x does something, y should happen Write unit of anlysis: DV: economic development IV: length of NGO commitment and amount of NGO resources committed o Research design together Tells you what you need to do Need to determine how to get data from independent variables and depednet variabels o Collect secondary data at the national elvel Set it up as a panel data/time series design One country/ one year Numerous countries, numerous years You can do this because we have the data that exists that happen, here we can o Measures- one way to measure at this level Yearly gdp change Economic diversity Economic equity as indicated by the ratio of the economic gains for disadvantaged groups Breakdowns of measures o How to narrow o Because of data and availability Measures o DV: stustainability of economic conditions o Only way to get at this is if they have received assistance, what there current income is However define standard of living Form expectations or hypothesis Case Design o Still would like to do a comparison, but instad find where collaboration is actually occurring Not really controlling things o Ask people directly, how they use that Investigate that in further detail. Selecting on a dependent variable o Result in better outcomes Selective for a community that has collaboration, see how it takes place o Measures All after you’ve collected data Go through trasncripts, coding, key words, key phrases o For this to work, have others to identify the same things Know a little for a lot of places or a lot for a little places Causal design o Strengths in generalizing Quasi o Generalizability, but a little more indepth All provide a little piece of a lot. o Together they are more effective. See how they are layed out Midterm o One o 15 or 20 guess questions o 10 short answer Suggest short answer, multiple guess Time o Using both in the book and in the lecture, review it Short answer are applied there are some defend or efute this statement, usually do either one. Chapter 8: Experimental Method More common in biological physical sciences o Define key terms o Present the logic o Outline the various types of experiments o Discuss some of limitations 8.1 Attributes and Variables o Attribute A characteristic or concept or construct that expresses the qualities possessed by a certain objects. Quantity that expresses a quality in numbers Qualitative research o Focused on the subjective attributes Quantitative researchWha o Focused on the measurement of objective variables o Variable types Independent Constitute as the cause Dependent The effect Background Antecedents that affect the situation prior to the study Intervening Events between the treatment and posttest measurement Extraneous Observed which might affect outcomes, but can’t be controlled Alternative independent Causes different from existing independent variable o Variables have precise definitions Univariate Study a single variable Bivariate Two variable study Multivariate Three or more variable study o Each variable has to be unidimensional A way to check variables If adjectives of variables are not polar opposites your variables overlap, which isn’t great 8.2 o Cause and effect A demonstration of cause and effect, four types Statically significant relationship between independent and dependent The independent precede dependent in variable Experimental group exposed to the independent variable stayed the same Alternative independent variables did not determine the results o One limitations of experimental approach Unicausal approach Three possibilities o Another true cause o Many causes have one effect o One cause has many effects 8.3 o Key element Management of variables so that effect can be measured, held stasitically Important part of experimental design Problem can be dealt with through randomization o Three variables might be external to the experiement, three within experiment might need to be controlled Knowledge of subjects Hawthorne effect Researchers know about experimental protocols o Testing process 8.4 Types of experimental design o Require pre and post testing of randomized control and experimental groups Treatment introduced to experimental group not to control Both are tested to measure any changes The experimental group does, cause and effect is established because the exposure of experimental group to the independent variable has changed o Type of design One shot case study Single group studied once, provides a treatment, measures an effect o Lacks control o Has only two right hand elements o Does not measure what performance was previously and therefore, whether it has actually changed One group- pre test, post test Measures performance before as well after introduction of the treatment to a single group Pre test and post test Basic true experimental design Correlation studies o Surveys measuring associations between variables The best we can achieve in survey research? o Through quasi experimental design Non-equivalent control group pre test post test Time series Control group time series Equivalent time samples Ex post facto Correlational study 8.5 Quasi Experimental and ex post facto research o Not true experiments, apply logic to attempt to control variables o Ex post facto designs Revese approach by searching backwards Correlation studies o Surveys measure associations between variables 8.6 Validity of experiments o Confined to laboratories, or classrooms Environments closely control variables Experimental research lower validity, conditions are limiting to world o Internal validity o Corrections of data collected, whether research made the difference observed o External validity Extent to which an experiment can be generalized to other situations Ecological validity o Reliable generalization from lab Just Plain Data Analysis Social indicators o The standards by which people measure performance, and hold people accountable o Si Time series comparisons Crucial evidence employed in arguments about politics and public policy Measurement Validity and Reliability o Indicator is valid if? It measures the concept it is trying to measure o Is consistent if? The measurements are consistent o Validity Consider the context in which it is used, and choices made in determining counts and divisors Social Capital Index by Putnam o Gives weighted average of fourteen measures of community life, engagement in public affairs, volunteerism, and social trust o Reliability Unreliability Caused by errors in data collection Cross national data is subject to inconsistent and collection Be wary of the use of sampling error alone as a measure of the reliability Reliability and Validity of Crime Statistics o Crime hasn’t declined as dramatically as thought, even though some things say that it has Evaluating Causal Relationships Relationships based on social indicator comparisons across time, groups, etc. If we find a observed relationship between two variables X and y o E can conclude that x causes y if it is not the case There is no real relationship Y causes x Something else related to x, causes y In the absence of a experimental setting The reasonable method to see if a relationship is true o Is to see If additional relationships can be drawn Controlling for Spurious Relationships o Control for various factors that might for account for and elaborate on a certain relationship Consider all factors o Simpson’s Paradox Occurs when a relationship exists for all subgroups of a population disappears when data are aggregated for the whole population Or the reverse o When there is no relationship among subgroups o Ecological Fallacy Drawing erroneous inferences about an individual behavior from relationship based on aggregated geographical data. Cherry picking o Selective culling of evidence to support a claim is a common problem Selective picking evidence is done intentionally Rate of Change Fallacy o Look at both sides of the story Depends on what you mean by climbing at faster rates depends on definitions
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'