top of page

Where the polls went wrong

Click the above picture to listen

 

THE national presidential polls were actually not too far off. RealClearPolitics, a polling aggregator, had Hillary Clinton winning the popular vote by three percentage points. With votes still being tallied, she currently looks poised to win by roughly one. Unfortunately for her, a two-percentage point polling miss can make all the difference in a tight race. Polling errors in states tend to be correlated with misses by demographic, meaning that if a state is off in one direction, similar states will follow. Mrs Clinton’s two-point polling slip cost her dearly in the “Rust Belt”, a collection of states surrounding the Great Lakes characterised by manufacturing-oriented economies that have been decimated by globalisation.


For months, Donald Trump has claimed that he would excel in the region. However, pundits had little reason to believe him as polls had him behind in most of the area. On election night, Mr Trump’s thesis was validated as he went on to win the states of Wisconsin and Pennsylvania, despite polling that claimed otherwise. He also came close in Minnesota, and seems to poised to even take Michigan.


Polls underestimated Mr Trump’s success in 40 out of 50 states across the country. They were especially wrong in deeply red states, perhaps in part because pollsters struggled to properly estimate turnout rates of the white-working class. Mr Trump has for months claimed that he would redraw the electoral map. He was right.

Read full article here.

aboutME

I am John Fan Zhang, a data analyst and finance researcher. I hold a PhD in finance, CFA charter and full membership of CFA New Zealand Society. I have fifteen-year experience in corporate investment and eight-year experience in advanced data analysis. My research focuses on the effect of social psychology (culture) on financial decisions. Finance research involves heaps of data analyses that lead me to the data field. I am a Microsoft Certified Solutions Expert (MCSE): Data Management and Analytics (Excel, Power BI, and SQL). Aside from Excel, Power BI and SQL, I am also familiar with econometric tools such as Stata, Eviews, and MATLAB. I use OX and Python for programming. I am an active data community event participant, volunteer, speaker, moderator, program reviewer, including PASS Marathon 2020, Global AI BootCamp Auckland 2019, SQL Saturday Auckland (2017, 2018, 2019), and Definity Conference (2018, 2019, 2020, Auckland, New Zealand).

Auckland, New Zealand

  • Google Site
  • Twitter
  • LinkedIn

©2016 BY JOHN'S DATA STORY

bottom of page