Statistics Essay Example | Topics and Well Written Essays - 1000 words - 1

Statistics - Essay Example if electronics’ price was measured in pounds). Since we have the same units for all the goods, we should use the ratio of aggregate prices. (c) Since we have fixed number of observations (random sample size is 16) the sample distribution is binomial. The Central Limit Theorem allows us to use approximate binomial distribution with the normal distribution when operating with large values of N. We can approximate binomial distribution with the normal distribution assuming that the whole number of steel coils is much larger than 16 (then the finite population correction will be close to one). (b)Confidence intervals give us the information that we can say with 90% of accuracy that the change between proportion is between 0.003 and 0.059. From this we can conclude that the proportion has slightly increased. (c) Sampling error can impact on the results of this study. There may be serious differences from sample to sample because the sampling technique used was simple random sampling. Additionally, non-response error can also influence the results of the study, since not all the drivers within the sample will admit using the cell phone. If we are to assume that all the drivers answer honestly then this error will not influence the results. (ii) If the null hypothesis is not true, then we can state that the productivity of exercisers is higher on average than that of non-exercisers. However reasons of this difference are not clear and require additional research. (b) Cluster sampling is about clusters, into which the whole population is divided. After that all units within the selected clusters are measured. Like cluster sampling stratified sampling also implies the division of the population into subgroups. However in stratified sampling they do not overlap and simple random sampling is implemented within each strata (subgroups). Due to dividing into subgroups every important strata is presented in the final sample. Additionally cluster