Week of notes 4
Week of notes 4 400
Popular in Research in Political Science
Popular in Public Relations
This 4 page Class Notes was uploaded by Dora Notetaker on Thursday September 17, 2015. The Class Notes belongs to 400 at University of Alabama at Birmingham taught by John E McNulty in Summer 2015. Since its upload, it has received 61 views. For similar materials see Research in Political Science in Public Relations at University of Alabama at Birmingham.
Reviews for Week of notes 4
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: 09/17/15
PSC 4002D Research in Political Science Dr McNulty Set of Notes 4 Week of Sept 14 TueSept15 Curves o A normal curve is an average around a distribution 0 Example average height 0 Most people fall in the middle so the curve graph is biggest at the middle 0 But some people may be below or above so they go on what is called the tails of the curve Standard Deviation o It is the estimated difference between the mean and the part of the curve where it starts to slope down o It is represented with o o For example If a test average is 90 and standard deviation is 5 then 23rds of the class got between an 80 and a 95 o 3 standard deviations 997 of your study 0 And 95 confidence in your study means that it is considered valid accepted Mean 0 Represented with the Greek Mu M 0 Middle of your graph Not all distributions are continuous 0 So sometimes there may be a lot of really tall and really short people 0 So your graph would look something like two small humps with a larger hump in the middle Logistic curve 0 This is just a curve that represents a logarithmic relationship not a linear one o Graphed with an S curve Some date will have ceilings or floors 0 Which are places where there is just no more data 0 Like test scores 0 The graph hits a wall at 100 because you can t make more than a 100 on a test 0 These are called truncated curves 0 They are also referred to as censored right or left censored 0 Medical studies are often right censored because they will test a study for 5 years and then will want to publish their findings Accuracy and Precision These are the 2 types of measurement error Accuracy How well a measurement reflects a concept lnaccuracy is random error Bias is error that clusters in a place not random 80 image a target with many spots hitting the center and some are off 0 The ones hitting the center are more accurate Precision o The degree to which your measure reflects the concept 0 So when you are not measuring what you need to you are imprecise 0 Like measuring body mass with a yard stick 0 Or measuring democracy based on whether there are elections 0 Imagine a target but all the shots are clustered outside of it 0 So the shots are accurate but imprecise Levels of Measurement Categorical 0 Deal with things that aren t numeric 0 Types 0 Nominal I Things you can t put in order I Citizenship race gender etc o Ordinal I You can put them in order I Education approval of the president etc Numeric 0 Measured in numbers 0 Example income age height cholesterol 0 Types 0 Interval I Means they don t have a true 0 point I Like temperature even when its 0 degrees it doesn t mean that there is an absence of heat 0 Ratio I Means they have a true 0 point I Like no income 0 income 0 IntervalRatio Thurs Sept 17 2015 Continuing with levels of measurement 0 Exercise What would you call a ranking of the performances of the GOP Presidential candidates in terms of excellent good fair poor 0 Categorical and Ordinal o What if we change excellent good fair and poor to 1234 o ltwould still be ordinal o The numbers are just placeholders but the level of measurement remains the same 0 People will use numbers to stand for categorical data but that doesn t make them numeric and you mustn t treat them as such 0 There are some numeric numbers when it comes to presidential debates o For example how much money they raised o It is considered numeric and a ratio o It is a ratio because there is a true 0 point I So if a candidate raises 0 for his campaign Criteria for Causality 0 These are conditions that must be met for a causal claim to be empirically valid 0 Types 0 1 Correlation I But remember CORRELATION ALONE DOES NOT IMPLY CAUSATION I There can be a great number of reasons that 2 variables can covary I It can be random like which team winning the Superbowl correlating with what the economy is going to be like the following year 0 2 Time Order Clock I Refers to the direction of causal order I Ex If someone says high income causes you to be welleducated they would be wrong since most of the time you get an education first and then go on to maximize your income later in life 0 3 Causal Mechanism I A logical explanation why the causality is happening I The process by which a change in 1 variable causes another to change 0 4 Controlling of Confounds I Confounds are other variables that might change independent and dependent variables so rather than x changing y 2 changes both x and y It can be a variable behind it all and researchers can t ignore it Ex lf good education does not cause higher incomes than what does maybe natural talent etc o All of these are necessary meaning that you need it but it is not enough by itself but not sufficient by itself
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'