Review of Mathematical Statistics Yulin Hou Department of Economics Florida International University Spring 2017 Yulin Hou, FIU Eco Econometrics Spring 2017 1 / 13 Random Sampling If Y1 , Y2 ,..., Yn are independent random variables with a common probability density function f (y ), then Y1 , Y2 ,..., Yn is said to be a random sample from f (y ). A random sample is a set of independent identically distributed (i.i.d) random variables. Yulin Hou, FIU Eco Econometrics Spring 2017 2 / 13 Estimators and Estimates Typically, we can not observe the full population, so we must inference based on estimates from a random sample. An estimator is just a mathematical formular for estimating a population parameter from a sample data. An estimate is the actual number the formular produces from the sample data. Yulin Hou, FIU Eco Econometrics Spring 2017 3 / 13 What make a good Estimator? Unbiasedness An estimator, W of θ, is an unbiased estimator if E(W)=θ (e.g), sample average is unbiased estimator of the population mean. Efficiency If W1 and W2 are two unbiased estimator of θ, W1 is efficient relative to W2 when Var(W1 ) < Var(W2 ). (e.g.) Yulin Hou, FIU Eco Econometrics Spring 2017 4 / 13 Consistency Law of large number refers to the consistency of sample averages as estimator for θ, that is, to the fact that: plim(W )=θ An unbiased estimator is not necessarily consistent. An biased estimator is not necessarily inconsistent. If W is an unbiased estimator of θ and Var(W) →0 as n → ∞, then plim(W )=θ. (e.g. sample average is a consistent estimator of the population mean ) Yulin Hou, FIU Eco Econometrics Spring 2017 5 / 13 Estimators as Random Variables Each of our sample statistics (e.g. the sample mean, sample variance, etc.) is a random variable. Each time we pull a random sample, we will get different sample statistics. If we pull lots and lots of samples, we will get a distribution of sample statistics. Yulin Hou, FIU Eco Econometrics Spring 2017 6 / 13 Central limit theorem The central limit theorem states that a random sample for any population (with finite variance) has an asymptotic normal distribution Y n ∼ N (µ, σ2 ) Yulin Hou, FIU Eco Econometrics Spring 2017 7 / 13 Hypothesis testing A premise or claim that we want to test H0 and H1 Null Hypothesis–H0 Alternative Hypothesis–H1 eg.It is believed that a candy machine makes chocolate bars that are on average 5g. A worker claims that the machine after maintenance on longer makes 5g bars. Write H0 and H1 . H0 : µ = 5g H1 : µ 6= 5g. Yulin Hou, FIU Eco Econometrics Spring 2017 8 / 13 Examples A company has stated that their straw machine makes straws that are 4 mm diameter. A worker believes the machine no longer makes straws of this size. Write H0 and H1 . Yulin Hou, FIU Eco Econometrics Spring 2017 9 / 13 Examples Doctors believe that the average teen sleeps on average no longer than 10 hours per day. A researcher believes that teens on average sleep longer. Write H0 and H1 . Yulin Hou, FIU Eco Econometrics Spring 2017 10 / 13 Examples The school board claims that at least 60 percent of students bring a phone to school. A teacher believes that this number is too high. Write H0 and H1 . Yulin Hou, FIU Eco Econometrics Spring 2017 11 / 13 Examples The school board claims that at least 60 percent of students bring a phone to school. A teacher believes that this number is too high. Write H0 and H1 . Yulin Hou, FIU Eco Econometrics Spring 2017 12 / 13 Continue Possible outcomes of hypothesis testing Reject null hypothesis Fail to reject null hypothesis Level of confidence: how confident are we in our decision? C=90%; 95%; 99% confidence interval Levels of significance α=0.10; 0.05; 0.01 Yulin Hou, FIU Eco Econometrics Spring 2017 13 / 13
© Copyright 2026 Paperzz