ECON2209 Lecture 18: 13March_EconStats_Karagodsky
Document Summary
1. 1 food shares in south africa: h0 : = . 35, ha : 6= . 35. 100: if x is less than then tstat should be negative, as long as tstat falls within a range it is unsurprising. In this case the range is between -1. 96 and 1. 96 (which we get from the chart): reject if |tstat| < |z z. 1. 2 looking at p-value: p-value = p(z < -5 or z > . 5, if p-value < we reject, if p-value > we accept. 1. 3 one sided hypothesis testing: h0 : < . 35, ha : . 35. In these cases you just look at one side of the normal probability distribution. So nd where there is 5% to the right of z on the standard normal distribution table, then z = 1. 645. We know from our previous example that our tstat is less than this and so we fail to reject the null hypothesis.