Did polls with larger samples perform better? Do you find a consistent answer across both time periods? Discuss what could be the reasons for potential differences (or consistency) between the two periods analyzed. * YouGov & Opinium are 2 polling companies.

print(paste("mean sample Opinium 2015", mean(p2015.op$sample)))

[1] "mean sample Opinium 2015 2018.88235294118"

> print(paste("mean sample Opinium 2017", mean(p2017.op$sample)))

[1] "mean sample Opinium 2017 2081.23076923077"

> print(paste("mean sample YouGov 2015", mean(p2015.yo$sample)))

[1] "mean sample YouGov 2015 1752.95833333333"

> print(paste("mean sample YouGov 2017", mean(p2017.yo$sample)))

[1] "mean sample YouGov 2017 1803.77419354839"

42.4 42.8 40 40- 36.9 33.6 33.6 30.4 31 30 - election 2015 20 - 2017 10 - 7.9 7.4 7.3 7.9 0 - ob.res_con ob.res_lab ob.res_libdem pr.res_con pr.res_lab pr.res_libdem observed and predicted