Question: A market research firm wants to estimate the difference in the population proportions of households that have high speed internet access for two metropolitan areas. A confidence coefficient of 0.92 and an interval width of 0.18 is desired. How many households should be selected from each population, assuming that both sample sizes are the same?