Скачать презентацию Data in the Classroom CSU Fresno November 1 Скачать презентацию Data in the Classroom CSU Fresno November 1

7426189bfc4f725f07817e09a7ae64f7.ppt

  • Количество слайдов: 71

Data in the Classroom CSU Fresno November 1, 2010 3/16/2018 1 Data in the Classroom CSU Fresno November 1, 2010 3/16/2018 1

Presenters John Korey l Cal Poly Pomona, Political Science l jlkorey@csupomona. edu Ø Ed Presenters John Korey l Cal Poly Pomona, Political Science l [email protected] edu Ø Ed Nelson l CSU Fresno, Sociology l [email protected] edu Ø 3/16/2018 2

Workshop Agenda Ø Ø Ø Ø 3/16/2018 Introductions (Ed Nelson) SSRIC (Ed) Data for Workshop Agenda Ø Ø Ø Ø 3/16/2018 Introductions (Ed Nelson) SSRIC (Ed) Data for this workshop (John Korey) Issues and examples l Experimental design (John) l Sampling and Statistical Inference (Ed) l Causality and contingency tables (Ed and John) l Fun with graphics (John) l Change over time (John) Where can we get the data? (John) What are we doing this year at Fresno State? (Ed) Evaluations 3

SSRIC Social Science Research & Instructional Council http: //www. ssric. org 3/16/2018 4 SSRIC Social Science Research & Instructional Council http: //www. ssric. org 3/16/2018 4

The Council Ø Ø Oldest CSU affinity group -- founded in 1972 Each campus The Council Ø Ø Oldest CSU affinity group -- founded in 1972 Each campus has a representative Works to provide access to data Promotes use of data analysis in research and teaching 3/16/2018 5

The Council Annual student research conference on April 29 at San Jose State University The Council Annual student research conference on April 29 at San Jose State University Ø Sponsors attendance at the ICPSR summer workshops in Ann Arbor, Michigan l http: //www. ssric. org/participate/icpsr_summ er Ø Works with the Field Institute -- selects faculty fellow (12 questions) – proposal due April 15 Ø 3/16/2018 6

Datasets for This Workshop Based on SPSS for Windows 16. 0: A Basic Tutorial Datasets for This Workshop Based on SPSS for Windows 16. 0: A Basic Tutorial (http: //www. ssric. org/trd/spss 16) l General Social Survey (GSS) 2006 Subset Ø Based on Introduction to Research Methods (http: //www. csupomona. edu/~jlkorey/POWERMUTT/i ndex. html) l American National Election Study (ANES) 2004 Subset l GSS Cumulative File Subset l ANES 2000 -2002 -2004 Panel Study Subset l U. S. Senate Ø 3/16/2018 7

Issues and Examples l l l Experimental design Sampling and statistical inference Causality and Issues and Examples l l l Experimental design Sampling and statistical inference Causality and contingency tables Fun with graphics Change over time 3/16/2018 8

Experimental Design 3/16/2018 9 Experimental Design 3/16/2018 9

Design Requirements Experiments l Random assignment to groups l Manipulation by experimenter of independent Design Requirements Experiments l Random assignment to groups l Manipulation by experimenter of independent (predictor) variable Ø Quasi-experiments Ø 3/16/2018 10

Types of Experiments Laboratory Ø Field Ø 3/16/2018 11 Types of Experiments Laboratory Ø Field Ø 3/16/2018 11

Laboratory Experiment: Prisoner’s Dilemma HOMICIDE DIVISION INTERROGATION ROOM A HOMICIDE DIVISION INTERROGATION ROOM B Laboratory Experiment: Prisoner’s Dilemma HOMICIDE DIVISION INTERROGATION ROOM A HOMICIDE DIVISION INTERROGATION ROOM B 3/16/2018 12

Laboratory Experiment: Prisoner’s Dilemma INTERROGATION IN PROGRESS DO NOT ENTER 3/16/2018 13 Laboratory Experiment: Prisoner’s Dilemma INTERROGATION IN PROGRESS DO NOT ENTER 3/16/2018 13

Laboratory Experiment: Prisoner’s Dilemma JACK’S BAIL BONDS “I’ll get you out if it takes Laboratory Experiment: Prisoner’s Dilemma JACK’S BAIL BONDS “I’ll get you out if it takes 20 years. ” 909/869 -4619 24/7 3/16/2018 14

Laboratory Experiment: Prisoner’s Dilemma Outcomes KEY: A'S OUTCOME B'S OUTCOME A TALKS A DOESN'T Laboratory Experiment: Prisoner’s Dilemma Outcomes KEY: A'S OUTCOME B'S OUTCOME A TALKS A DOESN'T TALK B TALKS 10 YEARS DEATH 1 YEAR B DOESN’T TALK 3/16/2018 1 YEAR DEATH WALK 15

Field Experiments Gosnell (1927) 3/16/2018 Gerber and Green (2000) 16 Field Experiments Gosnell (1927) 3/16/2018 Gerber and Green (2000) 16

Resources Ø The Center for Experimental Social Science 3/16/2018 17 Resources Ø The Center for Experimental Social Science 3/16/2018 17

Experimental Design in Survey Research Telephone vs. face to face (2000 ANES) Ø Question Experimental Design in Survey Research Telephone vs. face to face (2000 ANES) Ø Question wording: l Do you favor or oppose doing away with the DEATH tax? l Do you favor or oppose doing away with the ESTATE tax? Ø 3/16/2018 18

House 3/16/2018 19 House 3/16/2018 19

Estate http: //en. wikipedia. org/wiki/File: Ashford_castle. jpg 3/16/2018 20 Estate http: //en. wikipedia. org/wiki/File: Ashford_castle. jpg 3/16/2018 20

Results (2002 ANES) Favor abolishing “death tax”: 74. 3% Ø Favor abolishing “estate tax”: Results (2002 ANES) Favor abolishing “death tax”: 74. 3% Ø Favor abolishing “estate tax”: 71. 5% p = n. s. Ø 3/16/2018 21

Sampling and Statistical Inference 3/16/2018 22 Sampling and Statistical Inference 3/16/2018 22

What do we want to make sure our students understand? Ø Ø Ø Populations What do we want to make sure our students understand? Ø Ø Ø Populations and samples Parameters and statistics Sampling variability Margin of error Confidence intervals and confidence levels 3/16/2018 23

Basic principle Samples vary Ø What factors influence sampling variability? l Size of sample Basic principle Samples vary Ø What factors influence sampling variability? l Size of sample l Population variability l How sample was selected Ø 3/16/2018 24

Using Simulations to Teach Statistical Inference Draw repeated random samples Ø Compute sample statistic Using Simulations to Teach Statistical Inference Draw repeated random samples Ø Compute sample statistic Ø Construct chart showing the distribution of these sample statistics Ø Demonstration – see http: //constats. atech. tufts. edu Ø 3/16/2018 25

Estimators and Estimates An estimator is the method an estimate is the numerical result Estimators and Estimates An estimator is the method an estimate is the numerical result Ø Demonstration – see http: //inspire. stat. ucla. edu/unit_09/teaching_t ips. php Ø 3/16/2018 26

Resources -- Exercises Rolling dice and flipping coins – see http: //www. causeweb. org/repository/Star. Resources -- Exercises Rolling dice and flipping coins – see http: //www. causeweb. org/repository/Star. Libr ary/activities/andrews_2003/ Ø M&M’s – see http: //www. ropercenter. uconn. edu/education/ assignments/polling_basics. pdf Ø Drawing cards (Aces to Kings) – Xuanning Fu (CSU Fresno) Ø 3/16/2018 27

Resources – Web Sites Roper Center -- Fundamentals of polling: http: //www. ropercenter. uconn. Resources – Web Sites Roper Center -- Fundamentals of polling: http: //www. ropercenter. uconn. edu/education/ polling_fundamentals. html Ø American Association for Public Opinion Research – more on polling -- http: //www. aapor. org/Poll_andamp_Survey_F AQs. htm Ø Sample size calculator -- http: //www. surveysystem. com/sscalc. htm Ø 3/16/2018 28

Causality and Contingency Tables 3/16/2018 29 Causality and Contingency Tables 3/16/2018 29

What do we need to do to establish cause and effect? Statistical relationship Ø What do we need to do to establish cause and effect? Statistical relationship Ø Causal ordering Ø Eliminate alternative explanations Ø 3/16/2018 30

Example Ø Religiosity and how to regulate the distribution of pornography – data set Example Ø Religiosity and how to regulate the distribution of pornography – data set – gss 06_subset_for_classes_modified 2. sav l RELITEN – how religious the respondent is l PORNLAW – how the respondent feels about regulating the distribution of pornography 3/16/2018 31

Spuriousness Ø Ø Are there any alternative explanations (other than the causal one) for Spuriousness Ø Ø Are there any alternative explanations (other than the causal one) for the relationship? Can we think of any alternative explanations for RELITEN and PORNLAW? Gender might account for this relationship. Women are more religious than men and also more likely to want to restrict the distribution of pornography In other words, the relationship between X and Y might be spurious. So what we need to do is to test for spuriousness 3/16/2018 32

Testing for Spuriousness Independent variable (X) is RELITEN Ø Dependent variable (Y) is PORNLAW Testing for Spuriousness Independent variable (X) is RELITEN Ø Dependent variable (Y) is PORNLAW Ø Control variable (C) is SEX Ø 3/16/2018 33

Conclusions We found out that the relationship of RELITEN and PORNLAW was not spurious Conclusions We found out that the relationship of RELITEN and PORNLAW was not spurious when we controlled for SEX Ø But does that mean that we can conclude that the relationship is never spurious? Ø What does this say about proving causality? Ø 3/16/2018 34

Applying this to the Classroom Start with examples that make sense to students Ø Applying this to the Classroom Start with examples that make sense to students Ø Move to examples with real data that students can run Ø Generalize to issues of testing causality l Can show that a relationship is not causal (i. e. , it’s spurious) l Can never prove that a relationship is causal. Ø 3/16/2018 35

Example: Specification Open General Social Survey Subset Ø Does level of education influence the Example: Specification Open General Social Survey Subset Ø Does level of education influence the relationship between political views and party identification? Ø 3/16/2018 36

Specification (continued) Ø From Menu bar, go to: Analyze Descriptive Statistics Crosstabs Dependent variable Specification (continued) Ø From Menu bar, go to: Analyze Descriptive Statistics Crosstabs Dependent variable (first box): partyid Independent variable (second box): polviews Control variable: (third box): degree Statistics: Kendall’s taub Cells: Column percentages 3/16/2018 37

Specification (continued) Ø Look at pattern of Kendall’s taub statistics 3/16/2018 38 Specification (continued) Ø Look at pattern of Kendall’s taub statistics 3/16/2018 38

Example: Reactivity We know that the race of the interviewer in face-to-face interviews affects Example: Reactivity We know that the race of the interviewer in face-to-face interviews affects what people tell us about race Ø We know that the perceived race of the interviewer in telephone interviews also influences what people tell us Ø What about the gender of the interviewer in face-to-face interviews? Ø 3/16/2018 39

ANES Example Open anes 04 s Ø We’ll going to use three variables l ANES Example Open anes 04 s Ø We’ll going to use three variables l GENDER – gender of respondent l INTGENPO – gender of interviewer l WORKMOM – do you agree or disagree [that a] working mother can establish just as warm and secure a relationship with her children as a mother who does not work? Ø Let’s start by using the gender of the interviewer (INTGENPO) as our independent variable and WORKMOM as our dependent variable Ø 3/16/2018 40

ANES Example Continued What did we discover? Respondents interviewed by women are more likely ANES Example Continued What did we discover? Respondents interviewed by women are more likely to agree that working mothers can have a warm relationship with their children Ø Now let’s see if this is true for both male and female respondents. Let’s control for GENDER – gender of the respondent Ø We discover that it is true for both men and women. It appears that the gender of the interviewer does influence what people tell us about working mothers and their children Ø 3/16/2018 41

ANES Example Implications Since about 75% of the interviewers in this survey were women, ANES Example Implications Since about 75% of the interviewers in this survey were women, this has some serious implications. Ø This suggests that we will overestimate the percent of people that feel that working mothers can have a warm relationship with their children Ø 3/16/2018 42

Fun with Graphics 3/16/2018 43 Fun with Graphics 3/16/2018 43

Box and Whiskers Plots Open senate file (senate_mod. sav) Ø Compare acu and dwnom Box and Whiskers Plots Open senate file (senate_mod. sav) Ø Compare acu and dwnom scores 1. Graphs Legacy Dialogs Boxplots Clustered Summarize by Separate Variables Define 2. 1 st box: acu, dwnom; 2 nd box: party; 3 rd box: name; OK Ø 3/16/2018 44

Box and Whiskers Plots (continued) Ø Convert acu and dwnom to Z scores 1. Box and Whiskers Plots (continued) Ø Convert acu and dwnom to Z scores 1. Analyze Descriptive Statistics Descriptives 2. Move acu and dwnom to right window 3. Check Save standardized values as variables 3/16/2018 45

Box and Whiskers Plots (continued) Ø Compare Zacu and Zdwnom scores 1. Graphs Legacy Box and Whiskers Plots (continued) Ø Compare Zacu and Zdwnom scores 1. Graphs Legacy Dialogs Boxplots Clustered Summarize by Separate Variables Define 2. 1 st box: Zacu, Zdwnom; 2 nd and 3 rd boxes remain the same; OK 3/16/2018 46

Sample Size and the “Margin of (Sampling) Error” 3/16/2018 http: //www. surveysystem. com/sscalc. htm Sample Size and the “Margin of (Sampling) Error” 3/16/2018 http: //www. surveysystem. com/sscalc. htm 47

Just the Facts http: //pollingreport. com/guns. htm 3/16/2018 48 Just the Facts http: //pollingreport. com/guns. htm 3/16/2018 48

Poll Aggregators 3/16/2018 http: //www. pollster. com/polls/ 49 Poll Aggregators 3/16/2018 http: //www. pollster. com/polls/ 49

Do It Yourself Prognostication http: //uselectionatlas. org/PRED/ 3/16/2018 50 Do It Yourself Prognostication http: //uselectionatlas. org/PRED/ 3/16/2018 50

Resources Examples of Assignments (Roper Center) Ø Polling 101: Fundamentals of Polling (Roper Center) Resources Examples of Assignments (Roper Center) Ø Polling 101: Fundamentals of Polling (Roper Center) Ø Polling 201: Analyzing Surveys (Roper Center) Ø Polling for Dummies Ø Sample size calculator (Creative Research Systems) Ø Sampling Distributions (Tufts) Ø Polling and Survey FAQs (AAPOR) Ø 3/16/2018 51

Change Over Time 3/16/2018 52 Change Over Time 3/16/2018 52

Objectives To explain: ØTrend and cohort analysis (gsscums. sav) ØPanel studies (anespanl. sav) 3/16/2018 Objectives To explain: ØTrend and cohort analysis (gsscums. sav) ØPanel studies (anespanl. sav) 3/16/2018 53

Age Cohorts § § § GI Generation (born 1927 or earlier) Silent Generation (1928 Age Cohorts § § § GI Generation (born 1927 or earlier) Silent Generation (1928 -1945) Baby Boomers (1946 -1964) Generation X (1965 -1981) Generation Y (1982 or later) 3/16/2018 54

Procedure Ø SPSS line charts 3/16/2018 55 Procedure Ø SPSS line charts 3/16/2018 55

Dependent Variables Ø Values recoded into two categories (0 and 100) as nearly equal Dependent Variables Ø Values recoded into two categories (0 and 100) as nearly equal in size as possible. ØExample: Confidence in press is recoded as 100 (a lot or only some) and 0 (hardly any or none). ØThe resulting line graph can be interpreted as the percent of respondents coded as 100, that is, having at least some confidence in the press. 3/16/2018 56

Trend Analysis: Daily Newspaper Readership (Commands) Open gsscums. sav Ø Click on Graphs -> Trend Analysis: Daily Newspaper Readership (Commands) Open gsscums. sav Ø Click on Graphs -> Legacy Dialogs -> Interactive -> Line Ø Move NEWS to first window on right, and YEAR to second window. Click on OK Ø 3/16/2018 57

Trend Analysis: Daily Newspaper Readership (Results) 3/16/2018 58 Trend Analysis: Daily Newspaper Readership (Results) 3/16/2018 58

Cohort Analysis Ø To illustrate: l Generational replacement l Life cycle patterns l Across Cohort Analysis Ø To illustrate: l Generational replacement l Life cycle patterns l Across the board change 3/16/2018 59

Cohort Analysis: Daily Newspaper Readership (Commands) Open gsscums. sav Ø Click on Graphs -> Cohort Analysis: Daily Newspaper Readership (Commands) Open gsscums. sav Ø Click on Graphs -> Legacy Dialogs -> Interactive -> Line Ø Move NEWS to first window on right, YEAR to second window, and COHORT to third window. Click on OK Ø 3/16/2018 60

Cohort Analysis: Daily Newspaper Readership (Results) 3/16/2018 61 Cohort Analysis: Daily Newspaper Readership (Results) 3/16/2018 61

More Cohort Analysis Repeat above commands (first without, then with, COHORT), but instead of More Cohort Analysis Repeat above commands (first without, then with, COHORT), but instead of NEWS, use TVHOURS (over 2 hours per day watching TV), then CONPRESS (at least some confidence in the press) 3/16/2018 62

Even More Cohort Analysis Repeat above, but try the following: Ø GRASS (favor legalization Even More Cohort Analysis Repeat above, but try the following: Ø GRASS (favor legalization of marijuana) Ø RACMAR (oppose interracial marriage) Ø TRUST (think most people can be trusted) 3/16/2018 63

Panel Studies Open anespanl. sav Ø Did respondents in 2004 recall their 2000 vote Panel Studies Open anespanl. sav Ø Did respondents in 2004 recall their 2000 vote differently than they had in 2000? Ø Click on Analyze -> Descriptive Statistics -> Frequencies Ø Obtain frequency distributions for P 200004 and P 200000. Ø 3/16/2018 64

Panel Studies Did the relationship between party identification and feelings about Ralph Nader change Panel Studies Did the relationship between party identification and feelings about Ralph Nader change between 2000 (pre-election) and 2004? Ø Click on Analyze -> Compare Means -> Means. Ø Move NADR 00 PR and NADR 04 to first window on right, and PTYID 300 to second window. Click on OK. Ø 3/16/2018 65

Where Can We Get Data? Data resources on or linked from the SSRIC website: Where Can We Get Data? Data resources on or linked from the SSRIC website: http: //www. ssric. org/data 3/16/2018 66

Social Science Databases The California State University subscribes to three data bases to support Social Science Databases The California State University subscribes to three data bases to support teaching and research Ø Data bases l Inter-university Consortium for Political and Social Research (ICPSR) at the University of Michigan l Field Poll in San Francisco l Roper Center for Public Opinion Research at the University of Connecticut l General Social Survey and American National Election Studies are available through these databases l These are available to campuses by annual subscription Ø 3/16/2018 67

Proxy Servers On-campus access to data bases is IP authenticated Ø Off-campus access to Proxy Servers On-campus access to data bases is IP authenticated Ø Off-campus access to ICPSR and Roper through your campus’ proxy server Ø For ICPSR, account only needs to be authenticated from on campus or via proxy server every six months; otherwise, can be accessed from anywhere. Ø Off-campus access not available for Field data Ø Another alternative: set up a VPN on your home computer Ø 3/16/2018 68

Where Do We Get the Data? • SSRIC: http: //www. ssric. org/data • Pew: Where Do We Get the Data? • SSRIC: http: //www. ssric. org/data • Pew: http: //people-press. org/dataarchive/ • PPIC: http: //www. ppic. org/main/datadepot. asp • Berkeley’s SDA archive: http: //sda. berkeley. edu/archive. htm • ICPSR: http: //www. icpsr. org • Roper: http: //www. ropercenter. uconn. edu • Field Public : ftp: //128. 32. 165. 222: 2121/ (download spss files) CSU and UC only ( analyze online): http: //ucdata. berkeley. edu/data_record. p hp? recid=3#analyze 3/16/2018 69

What are we doing this year at Fresno State? Ø Ø Ø Workshops for What are we doing this year at Fresno State? Ø Ø Ø Workshops for faculty and staff l Teaching with Data (September 23) l Data in the classroom (November 1 with special guest presenter John Korey, Political Science, CSU Pomona) l Online statistical packages (SDA) (early spring) l SPSS (introductory and intermediate) (late spring) Encourage students to present their research at student research conferences (SRC) l SSRIC’s SRC in San Jose on April 29 l Santa Clara University’s Anthropology and Sociology SRC in April l CSU’s Student Research Competition in Fresno on May 6 -7 Presentations at the department level One-on-one consultations with faculty Surveys to get faculty’s input and feelings 3/16/2018 70

Evaluations 3/16/2018 71 Evaluations 3/16/2018 71