The article covers FiveThirtyEight’s democratic primary polling averages method in the 2020 election, and why they think they’ve created a better method. They adjust state polls based on trends in national polls, pollster bias and included pollster ratings weights, and finally putting more weight on polling changes after a major news event.
Loaded the csv from github raw into data frame in r:
polling<-read.csv("https://raw.githubusercontent.com/GuillermoCharlesSchneider/HW1/main/pres_primary_avgs_1980-2016.csv", header= TRUE, sep=",")renamed columns:
colnames(polling) <- c("RaceYearAndParty", "State", "ModelDate","CandidateName","Candidate_ID","Estimate","Trend_Adjusted", "Timestamp", "OneDayTest", "Contest_Date")subset of just iowa polling:
IowaPolling <- subset(polling, State == "Iowa")I successfully pulled in the data