Schar School of Policy and Government - Stats Screening Exam Prep

Size: px

Start display at page:

Download "Schar School of Policy and Government - Stats Screening Exam Prep"

Francis Brooks
5 years ago
Views:

1 Schar School of Policy ad Govermet - Stat Screeig Exam Prep Welcome SSPG Grad Studet! Ue thi guide for preparatio for the tat creeig exam. Part 1 Itroductio A you kow, the itroductory graduate method coure PUBP/GOVT/PUAD 511 ha a creeig examiatio. Thi i to eure that all tudet are prepared to begi o day 1 without havig to reteach exteively may of the topic covered i a typical udergraduate tat or ocial ciece method requiremet coure. I order to help tudet prepare for the creeig exam, the Schar School of Policy ad Govermet offer two path to help you pa the exam: 1. Ue thi brief guide to help you tudy for the exam itelf, ad write the exam o oe of the day available to you durig the emeter prior to your itetio to regiter i your deired ectio. (ote: ucceful completio of the creeig exam doe ot guaratee you a pot, o the earlier you take the exam uccefully, the greater the likelihood that you will be able to regiter i your deired ectio.) Thi path i recommeded for thoe tudet who feel cofidet i their ability ad kowledge that they have retaied from their udergraduate career courework.. Atted oe of the 5 cheduled tat workhop held throughout the year ad write the creeig exam at the cocluio of the workhop. The workhop i more tha a refreher, it aume o prior tatitical traiig whatoever, o thi path i recommeded for thoe who have ot take tatitic ad/or a ocial ciece reearch method cla i a log time or ot at all, ad thoe who may wih to refreh themelve before udertakig more challegig material i GOVT/PUAD/PUBP 511. Part Geeral Priciple What follow i a brief lit of term, cocept, idea, calculatio, etc. for which you will be expected to have kowledge of to uccefully pa the creeig exam. Reearch Deig Depedet Variable Idepedet Variable Reearch Motivatio Idetifyig Audiece Hypothee Developmet Graphic Hitogram Box-Plot Bar Graph Pie Chart Scatter Plot Frequecy Ditributio Table Percet Proportio Cumulative Table Uivariate Meaure Level of Meauremet Cetral Tedecy Diperio Z-Score & Probability Outlier Statitical Iferece The Law of Large umber Cetral Limit Theorem Stadard Error (Theoretical) Stadard Error of Ratio ad Iterval level data Stadard Error of Proportio Cofidece Iterval Margi of Error Iterval for Iterval/Ratio Level Data Iterval for Proportio Baic Sample Size Etimatio The Igorace Aumptio -1-

2 Schar School of Policy ad Govermet - Stat Screeig Exam Prep Part 3 Some Example Screeig Exam Quetio Digetig reearch read the abtract below. Idetify the followig ad jutify your repoe: a) Who you believe the audiece to be (ector/motivatio, etc.) b) Who the producer of the argumet are (ector/motivatio, etc.) c) idepedet variable d) depedet variable e) Idetify the data the author are uig f) hypothee the author may be tetig, give the abtract Stephaie Riegg Cellii The George Wahigto Uiverity Fiacial Aid ad For-Profit College: Doe Aid Ecourage Etry? Joural of Policy Aalyi ad Maagemet. Summer 010. Vol. 9 o. 3 pp Cocer over riig college tuitio ad low ecoomic growth have brought reewed attetio to the role of federal ad tate fiacial aid program i opeig acce to educatio. Depite a large body of literature examiig the effect of grat aid o four-year ad public two-year college erollmet, for-profit college particularly the vat majority that offer two-year degree ad certificate have largely bee igored. Uig pael data method ad a ew admiitrative data et of for-profit college operatig i Califoria betwee 1989 ad 003, I ae the impact of the federal Pell Grat program, the G.I. Bill, ad Califoria' Cal Grat program o the et umber of for-profit college per couty. The reult ugget that for both Pell ad Cal Grat, icreae i the per-tudet maximum award ecourage for-profit etry. Thi relatiohip i particularly trog i coutie with high adult poverty level, where more tudet are eligible for aid. Further, thee gai i the private ector do ot appear to come at the expee of the public ector. Rather, public commuity college alo experiece erollmet gai a the geeroity of Pell ad Cal Grat icreae, although thi reactio appear to be weaker tha the reactio of for-profit. Arijit Chatterlee & Doald C. Hambrick Peylvaia State Uiverity It all about me: arciitic Chief Executive Officer ad their effort o Compay Strategy ad Performace Admiitrative Sciece Quarterly. September 007. Vol. 5 o. 3 pp Thi tudy ue uobtruive meaure of the arciim of chief executive officer (CEO) the promiece of the CEO' photograph i aual report, the CEO' promiece i pre releae, the CEO' ue of firt-pero igular proou i iterview, ad compeatio relative to the ecodhighet-paid firm executive to examie the effect of CEO arciim o a firm' trategy ad performace. Reult of a empirical tudy of 111 CEO i the computer hardware ad oftware idutrie i how that arciim i CEO i poitively related to trategic dyamim ad gradioity, a well a the umber ad ize of acquiitio, ad it egeder extreme ad fluctuatig orgaizatioal performace. The reult ugget that arciitic CEO favor bold actio that attract attetio, reultig i big wi or big loe, but that, i thee idutrie, their firm' performace i geerally o better or wore tha firm with o-arciitic CEO. --

3 Schar School of Policy ad Govermet - Stat Screeig Exam Prep The quetio below refer to the Box Plot to the right. a. What ca you ifer about fuel coumptio from thi graphic compario betwee dometic ad foreig vehicle? Dicu both locatio ad variability i your awer. b. What i the level of meauremet for both the idepedet ad depedet variable? c. Etimate a 5 umber ummary for dometic automobile fuel coumptio, give what you ca ee i the graph. Mileage (mpg) Box Plot Comparig Forieg to Dometic Productio Vehicle Fuel Ecoomy (1981 Data o 74 top ellig peroal automobile) Dometic Foreig The followig quetio refer to the Box Plot (o the right) take from the 008 Geeral Social Survey (GSS). a. What ca you ifer about the relatiohip betwee the year of choolig that a adult ha ad the umber of correct repoe to a geeral ciece quiz a admiitered a a compoet of the 008 GSS? Dicu both locatio ad variability i your awer. b. What i the level of meauremet for both the idepedet ad depedet variable? c. Etimate a 5 umber ummary for the ciece quiz amog the category of urvey repodet with the greater umber of year of formal educatio, give what you ca ee i the graph. Year of Formal Schoolig Predictig umber correct o ciece quiz Sciece Quiz Reult 0-1 yr 13 plu yr 008 Geeral Social Survey (Adult aged 18-90) Awer the followig quetio i hort eay form: Why doe probability amplig allow u to ue tatitic to make etimate of populatio parameter? Explai the beefit of the cietific method a a kowledge acquiitio ytem. What doe it mea to operatioalize cocept? How doe our choice of itrumetatio cotrai our ability to ru tatitical tet? Explai what a cofidece iterval i. Why ca we ue a igorace aumptio whe etimatig ample ize from proportio? -3-

4 Schar School of Policy ad Govermet - Stat Screeig Exam Prep Below are lited a radom ample of jury deliberatio time (the time it take a jury to determie a upect guilt or iocece) take from a ample of trial coducted from May to July of 01 i the Fairfax Couty Circuit Court. Complete the quetio that follow about thi ditributio: Fairfax Couty Circuit Court - Jury Deliberatio Time (I Hour) a. Build a appropriate cumulative percet frequecy table for jury deliberatio time i Fairfax Couty. Fid each of the followig: b. Mode c. Media d. Mea e. Variatio ratio f. Q1 g. Q3 h. IQR i. Rage j. High value k. Low value l. Variace m. Stadard deviatio. Build a +/- iterval aroud the mea ad make a claim about outlier. o. The Z-core of the highet value p. Probability of fidig a core a high a the highet value or higher. Cofidece Iterval. a. Below are ome ummary tatitic for the Feelig Thermometer core repodet to the 008 atioal Electio Study gave to the US Supreme Court. The thermometer core rage from 0 Very Cold to 100 Very Warm. Calculate the tadard error of the etimate of the mea. Variable Mea Std. Dev. Mi Max Supreme Court b. Uig the tadard error you calculate above, create a 90% cofidece iterval aroud the etimate ad make a claim about the locatio of the populatio parameter. c. A recet poll of 61 govermet admiitrator foud that 44% worked ucompeated overtime durig July of 01. Calculate the tadard error of the etimate of the proportio. d. Uig the tadard error of the proportio you calculated above, build a 95% cofidece iterval aroud the etimate ad make a claim about the locatio of the populatio proportio. e. Uig the origial proportio ued i quetio c, determie how much ample you would eed to make a etimate of the proportio withi 3% with 90% cofidece. -4-

5 Backgroud to the GOVT / PUAD 511 Screeig Exam Part 4 Formula to kow M ( ) ( 1)*0.5 Media poitio v ratio f (omodal) IQR Q3 Q1 Populatio Sample Cofidece Iterval X i x i x x x z ( X ) i ( x ) i x 1 MOE Z or x z x ( X i ) ( x x) i 1 Z MOE CL p z / p(1 p) z X z x x p p pq pq 90% CI X % CI X % CI X.575 X X X -5-

6 Backgroud to the GOVT / PUAD 511 Screeig Exam Part 5 The ormal Ditributio Table ormal Ditributio Probability of ot beig i the RH Tail z-core + colum (X.X) Z

Data Analysis and Statistical Methods Statistics 651

Data Analysis and Statistical Methods Statistics 651 Data Aalyi ad Statitical Method Statitic 65 http://www.tat.tamu.edu/~uhaii/teachig.html Lecture 9 Suhaii Subba Rao Tetig o far We have looked at oe ample hypothei tet of the form H 0 : µ = µ 0 agait H