In the CBI business confidence survey, employers are asked whether they expect order to increase in the next 12 months. In January 2009, 38 from a random sample of 120 employers reported they expected an increase in orders over the next 12 months. From an independent random of 210 employers asked in January 2010, 86 reported an expected increase in orders over the next 12 months. At the 5% significance level, calculate the power of a test of the null hypothesis that there is no difference in business confidence between the two dates, (against the 1-sided alternative that confidence had improved), given that the true proportion expecting an increase in orders was 0.35 in January 2009 and 0.45 in January 2010.

How do i find the variance for this and the apparently you have to calculate the pooled mean aswell? How do you find these?