Research, Pols 201,
Research, Pols 201, 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 February 14, 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 33 views. For similar materials see Research Methods in Political Science in Political Science at University of Tennessee - Knoxville.
Reviews for Research, Pols 201,
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: 02/14/16
Research Methodology: Chapter 4 4.1 Research Hypotheses o Clearly identifies the type of data you have to collect o Deductive hypothesis Research hypotheses bridge literature review and methodology Deductive hypothesis o Derives beforehand from previous theory Formal Hypothesis Should provide possible explanations State relationships between variables Be testable o Be consistent o Be simple o Grounded Research On experiences, not biased by academic preconceptions Review data, see possible pattern Theories are Rejected, refuted, falsified o Nothing is ever proved correct All findings are probabilities 4.2 Objective and Subjective o Objective Treats physical and social world that we can sense directly o Subjective Considers what people think, more based on the mental constructs Inferences o Metaphysics becomes a part of the argument World is an illusion vs. world is our reality Positivism and post-positivism o Positivist Found in experimental psychology, quantitative research, and physical/natural sciences World studied as objective things Data is independent Accepted data as evidence only if conducted in direct observation under strict rules Isolate elements of cause and effect Scientific Method is considered objective Low level of validity o Post-positivist Knowledge is subjective, with value Data is dependent on relationship between knower and known Naturalistic non experimental research Knowledge is subjective and holistic Scientific methods are social constructs Qualitative studies offer very little. o Positivistic use of formal theory can raise reliability But lower relevance o Post positivize can improve relevance, but not reliability 4.4 Commonsense and Pragmatism o Popper Science doesn’t require metaphysical basis Realism o Accepting that the real world exists Commonsense Phenomenology o All researchers believe in integral systems o Positivism and post positivism are both social constructs Pragmatism Knowledge useful in practical effect 4.5 Mixed Methods o Combining qualitative and quantitative techniques 4.6 Triangulation o Similar findings from different sources Bringing together multiple data types on one problem Meta analysis o Analysis of large numbers of similar studies o Methodology o A general approach to learning about that which we can know In general a balance of goals A theory that is very generalizabile and reliable o Validity and relevance Aggregate is something that gives you less personal attributes Very little about the broader relationship o Methodological approaches Do we look on a broad scale, do we look at a smaller scale Inference and probability o We sill study a subset of a population to learn about something larger in the population Could get a sample Average of the averages give you a good parameter We don’t know what the population parameter o When they did the poll, they have no idea what the result is going to be. Samples and populations o No capacity to measure an entire population Impractical to think that we can calculate o Charachteristic we have calculated from some sample We use Roman letters do denote things about this sample How much money we make Parameters are characteristics of a population o We use greek letters to denote them o Parameters are usually hypothetical American, sample, greek, population Calculate the sample statstic, use it to estimate o We will calculate the mean of a sample, as x bar Iuse x bar to estimate mu Use s to estimate o Mean Figure out how good of an estimate you have o You will have some number below three Probability and the Central Limit Theorem o Our sample statistic is unlikely to ever match the population parameter If we took a mean of the means we would be close to the population average Probaility o The relative likelihood of occurrence of any given outcome or event o The difference between sample statistics Is sampling error The whole point is to estimate with as little sampling error as possible. How to minimize sampling error o To be a valid representation of population, must be selected Every item in the population has an equal chance of being selected Make it as random as possible o Sampling error can be positive or negative Mean K of the sample means o Much better estimate than any particular sample is. More samples taken The larger your sample size is in one sample, the better the population parameter Guy who figured this out o Worked at Guinness brewery The closer they get to that number, the better the estimate We know sample isn’t exact replication o We want to use this to represent and say something about the population Assess how well We can get an exact number that says we would have seen o We can’t afford it if it doesn’t pay for itself Constantly trying to gauge the benefit of a larger sample size o There is till gonna be 50 people who just don’t fill it out Parameter estimates o We don’t know how good it is o Does the sample mean Hypothesis o Research hypothesis states the theoretically expected outcome Null hypothesis states the inverse of the research hypotheses o Hypothesis testing Collect relevant or reject or fail to reject to nul hypothesis and discuss implications and limitations Significance o Two types of significance Statistical significance Only meaningful if we are using a sample to infer to a population o If you have population there is no need to asses significance, any change is meaningful 40 minutes better, than 30 Higher the level, more comfortable I will feel o Figure out how many days worth you need Stistcal tells you an estimate Practical Does any change actually make any real world difference Samples o When to use a population We will use the hypothesis in assessing some level of uncertainty Within or beyond a particular stage o Testing if wheter or not something is ineffective s Not testing on the entire population Your population is humanity The meaning of the null hypothesis o Any observed difference is the result of sampling error alone We want to know If it is likely, if it is how unlikely that would occur. Chapter 2: Research Ethics Before research, it’s important to consider ethics 2.1 Codes of Ethics o Standards of behavior Guide how to act with integrity, and how professionals should act. Professional competence Integrity Professional, scientific responsibility Respect of people Social responsibility o Adhere to technical standards o Ensure research is competent o Correct expertise o No harassment o Avoid conflicts of interest o Confidentiality o Informed consent o Avoid plagiarism 2.2 Permissions to Research o Authority to carry the project out and also to obtain consent from participants If approved, might need subsequent agreement Prevents parts for being overloaded Makes sure credentials are appropriate Ethical Not disruptive o Often Bureaucratic delays o Tell participants Purpose of research What you do with results Questions about research answered Permission to continue Right to refuse is respected Right to withdraw at any stage 2.3 Responsibilities o Respect for culture Don’t get involved Don’t break law Never invent information Don’t misrepresent 2.4 Confidentiality o Interviewers have the right to only answer some questions Only a code number to identify 2.5 Feedback o Community feedback from ethical and practical views Approaches Summary of findings on community or work boards Arrange public meetings Deliver summaries Arrange interviews Write articles Call meetings Hold workshops 2.6 Participatory Research o Some people may feel exploited by research Research should be a process of reciprocal social action Researchers and participants are on equal footing in participatory research Requires full, active involvement Flexibility 2.11 The meaning of the null hypothesis Any observed difference is the result of sampling error alone o The observed difference is random, not meaningful Distribution o The variable tells us what values the variable takes and how often each value occurs We have some number of total observations, n o Population or sample, doesn’t matter o Each observation has some value or score for each o The distribution tells us the frequency with which each particular variable is found There is going to be variability There will be some distribution o There is some amount of time that is the lowest, and highest o Everyone in between Frequency distribution o Watch and wait, tabulate N=51, after gauges Response to long line here is a variable o Categorized into four responses o Tabulated frequency When ordinal, make it make sense o Help us make sense in pattern and data Creating a set to intervals Every five points o Bell curves Histogram Most in the middle, l Standard deviation is spread about our observations are o 68% are in the one standard deviation of the mean Taking the sample o The sample will not necessarily be distributed normally o If we have the guess true population, but we get a sample mean from 75, it is extremely unlikely we would get a 75, if the true pop is 50. Should this sample result reasonly have occurred if ho is true o We require strong evidence aginst ho to say no Possibility one o Look at sample mean and figure out if it could have come How likely is it that the sample is pulled out Possibility Two o If you know the mean and standard deviation, all the time, observations should be within three standard deviatinos, then its very likely Students t o Allows us to evaluate whether the sample mean is statistically the same as a hypothesized population mean o Either our null is true, or fail to reject Type II o Our sample mean of 35 could have come rom a sample mean of thrity or less If the n o Our population mean is less than thirty It functions, makes money, doing great. In our projects o This is the process you would go through, collecting data, run analysis Ascertain the extent to which the likelihood that it is reflective of the hypothetical Type one errors are more costly than type twos
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'