This is Political Science Problem set Based on the code book, R data and Article that I attach, I need to analyze the data and write up my summary I attach the introduction as well. Also I need to send you data file which is associated with R by e-mail

Problem Set 2: Analysis of 2018 Framing and Border Death Awareness Study Start this early. It will be due IN CLASS in HARDCOPY FORM on Nov. 13. In addition to turning in the write up, you will need to email your R code DIRECTLY to me AND your TAs (do not send it through Canvas). We will run the code to ensure it reproduces the results. You may work together on coding but the write-up should be entirely your own. The WILDCARD must be your own work. In your write-up, speak in plain language. These are real data involving real world issues. DO NOT WAIT UNTIL THE LAST MINUTE TO START THIS. Assignments turned in after the first 10 minutes of class on Nov. 13 will be docked 10 percent. Each subsequent day it is turned in late, it will be docked 10 percent. E-MAILED copies of the write-up will NOT be graded. Grading will be done as follows: Overall plot quality: 100 points. Plot quality will be determined by proper labeling, readability, orientation on the page (gigantic plots will be downgraded), and correctness of code. We will evaluate the plots in their totality. Pro-tip: if you colorize your plots but print them in grayscale, I will NOT be able to determine the colors you reference. Either print them in color or use grayscale variants in the plots. Overall analysis: 200 points. Analysis grade is based on the overall clarity of writing, precision in language, ability to address the question/task that is asked. Overly short answers (a few sentences) will be downgraded. You are meant to do the write-up on your own. Evidence of equivalent text on two or more papers will be viewed as possible evidence of plagiarism and students involved will be reported to Student Judicial Affairs. R Code: 20 points. Code will be based on accuracy of the code and the ability for us to reproduce the analysis. Failure to reproduce the analysis will result in a downgrade. You are welcome to work together on coding. As such, similar code across students will NOT be viewed as possible plagiarism. Indeed, I encourage you to work together. Here are details of the data set upon which you are working: This study, which I will discuss in detail in class (10/30), was administered to several hundred undergraduate students during the Winter Quarter of 2018. You have access to the recoded data. It is saved as a .dta file on Canvas. You can read this file into R using the “foreign” library. Do not try to open this file as you’re not likely to have the app to open a .dta file (it requires a software program called Stata).

The dataset has several variables of interest. I have uploaded the study codebook to Canvas, but I’ve also recoded several variables (you will need to consult the code book to get the exact wording for some of the items). Recoded Variables: Demographics and political orientation: sex: 1=Female; 0=Male citizen: 1=naturalized citizen; 2=native born citizen; 3=noncitizen parent_born: (see code book; category 4=parents born in U.S.) pid7: 7-point party affiliation scale anchored with 1=Strong Democrat; 7=Strong Republican. ideo9: 9-point ideology scale ranging from 1=extreme liberal to 9=extreme conservative trump_support: 1 means the respondent either voted for President Trump or would have voted for him had he/she been eligible to vote in 2016 TAMU: coded 1 if the respondent is a student at Texas A&M university. 0 otherwise. ucd: coded 1 if the respondent is a student at UC Davis. 0 otherwise. Psychological scales: social_dominance: Social dominance orientation scale (high scores=greater beliefs on social dominance) system_justification: System justification beliefs (high scores=higher system justification beliefs). For both items, see the code book for specific wording. If you want further background, you can skim the article I posted on Canvas by Jost. Race variables to be used in this homework: R_onlywhite: 1 if the respondent indicates they are white (and does not indicate any other race) and 0 if otherwise. R_latino: 1 if the respondent identifies as being Latina/o; 0 otherwise.

(There are other racial/ethnic categories but there are not sufficient numbers of them to obtain statistical power). Experimental conditions (discussed in class): Several contrasts have been coded. Refer to your class notes from 10/30 for further information. crime_control: 1 if respondent was in the crime statistic condition and 0 if in the neutral condition. death_control: 1 if respondent was in the death statistic condition and 0 if in the neutral condition. crimestory_control: 1 if respondent was in the personalized crime condition and 0 if in the neutral condition. deathstory_control: 1 if respondent was in the personalized death condition and 0 if in the netural condition. death_crime: 1 if respondent was in death statistic condition and 0 if in the crime statistic condition deathstory_crimestory: 1 if respondent was in the personalized death condition and 0 if in the personalized crime condition. Note that any contrast can be created since respondents were randomly assigned to one of these conditions. Policy Outcomes Four variables are coded to measure support/opposition for: a border wall, increasing number of U.S. Border Patrol agents, expedited deportation, and making it illegal to give assistance to immigrants in distress. These variables are: support_wall: 1=Strongly oppose; 7=strongly support. support_bp: 1=Strongly oppose; 7=strongly support. support_expedited: 1=Strongly oppose; 7=strongly support. support_aid: 1=Strongly oppose; 7=strongly support. There is a summary scale of these four items called: restriction. It is an additive scale.

Migrant Death items estimate_dead: About how many illegal immigrants would you guess have died on the U.S. side of the border while crossing from Mexico in the past 20 years? Please give your best guess without looking up the answer 0 to 500 (1) 501 to 1,000 (2) 1,001 to 2,000 (3) 2,001 to 3,000 (4) 3,001 to 4,000 (5) 4,001 to 5,000 (6) 5,001 to 6,000 (7) 6,001 to 7,000 (8) 7,001 to 8,000 (9) 8,001 to 9,000 (10) 9,001 to 10,000 (11) 10,001 to 11,000 (12) More than 11,001 (13) personal_surprise: 1=Extremely surprised; 5=Not at all surprised (see codebook) americans_surprise: 1=Extremely surprised; 5= Not at all surprised (see codebook) Blame Attribution Items: Respondents were randomly assigned to one of two conditions. In the first condition, respondents answered six generalized blame questions about migrant deaths. These are scaled in a variable called: general_blame See the codebook for question wordings. In the second condition, respondents were asked 9 questions about immigration policy and asked to assess their beliefs about how much the policy was to blame. These items are scaled as: policy_blame. Both scales are scored such that higher scores are associated with dispositional blame and lower scores are associated with situational blame (recall class discussion on this on 2/28). blame_post: this is a 5-category unordered variable about which factor is most to blame for immigrant deaths. 1=US immigration policy; 2=poor Mexican economy; 3=US labor demand; 4=lack of deterrence in Mexico; 5=immigrants themselves.

From this variable, there is a dummy variable called: blame_uspolicy (1 if respondent indicates US immigration policy is to blame and 0 otherwise) and blame_immigrants (1 if respondent blames immigrants themselves and 0 otherwise). Immigrant Numeracy Items undocumented_size: gives the estimated size of the undocumented Hispanic population. undocumented_size_metabeliefs: ASKED ONLY of Latina/o respondents. It is their estimate of how they think non-Latina/os would estimate the size of the undocumented population. Both variables are integers from 0 to 100. Chain Migration Respondents were randomly assigned to either a question referencing “chain migration” or “family reunification”. These responses are saved in the variable call: family_policy This variable is scored such that higher scores imply the respondent is OPPOSED to the policy. There is a dummy variable called chain_family and is coded 1 if respondent got the chain migration question and 0 if the respondent got the family reunification item. chain Do you support or oppose a policy of chain migration that allows legal immigrants to the U.S. to sponsor family members for consideration for legal immigration status in order to reunify their families? Strongly support (1) Moderately support (2) Slightly support (3) Neither support nor oppose (4) Slightly oppose (5) Moderately oppose (6) Strongly oppose (7) reunification Do you support or oppose a policy of family reunification that allows legal immigrants to the U.S. to sponsor family members for consideration for legal immigration status in order to reunify their families? (This has same response options as above). Using these data, please answer the following questions/do the following tasks: 1. Is there a difference in social dominance orientation for Republican identifiers compared to Democrat identifiers? How do you interpret this difference? What do you think it means substantively? To answer this, you may consider a side-by-side boxplot as well as a t-test.

2. Is there a significant difference in system justification for Republican identifiers compared to Democrat identifiers? What do you think it means substantively? To answer this, you may consider a side-by-side boxplot as well as a t-test. 3. Create a scatterplot between social dominance and system justification for all respondents, for Democrats only, and for Republicans only. Provide an interpretation of these plots. What do you learn from them? What does the relationship suggest? 4. Using the variable named “restriction,” is there a difference in policy preferences for individuals exposed to the migrant death condition compared to the crime condition? Specify what the null and alternative hypotheses are. Feel free to draw on my discussion of the research project from 10/30 lecture to inform you as to the hypotheses. To answer this, you may consider a side-by-side boxplot as well as a t-test. 5. THE AGGIE CHALLENGE! UC Davis is the Home of the Aggies. But Texas A&M calls themselves the Aggies too. Let’s take the Aggie Challenge. Create a dummy variable in R coded 1 if a respondent is from UC Davis and 0 if the respondent is from Texas A&M. For the following variables, assess whether or not there is a significant difference between the two groups in (to answer this, you may consider a side-by-side boxplot as well as a t-test): a. social dominance orientation b. system justification c. left-right ideology d. support for restrictionist immigration policy e. estimated size of the undocumented Hispanic population Based on your analysis, what can you say about UCD respondents compared to Texas A&M respondents who participated in this study? 6. Is there a significant difference between support for family reunification policy when “chain migration” is used compared to the phrase “family reunification”? To answer this, you may consider a side-by-side boxplot as well as a t-test. WILDCARD 7. Using the data provided, test three well specified and specific hypotheses that may be of interest to you or to someone generally. (For example, are there gender differences or racial differences in some of the outcomes).