Online Participation Topic: Week 5
For week 5, the online participation topic is on confidence intervals. We basically want to find the mean of a population (µ) by taking a sample and inferring a range of µ from the sample’s properties. Let
N = the population size
n = the sample size
µ = the population mean
= the sample mean
σ = the population standard deviation
s = the sample standard deviation
The sample’s properties (n, , s) are known or can be computed. The population’s properties (N, µ, σ) are the unknowns. However, n << N (usually).
Sometimes σ is known. If this is the case, we can compute a z-score:
z = =
If σ is unknown, we then compute a t-score
t = =
k is a measure of the number of standard deviations from the mean. By choosing an appropriate value of z or t, we can construct a “confidence interval” about the mean
( – k) < µ < ( + k)
If we let z = 1.645 so that 90% of a normal distribution would fall between +/-1.645σ of the mean, then the above inequality would be a 90% confidence interval. This means that 90% of the time, the population µ will be inside the confidence interval constructed using the sample data.
For week 5:
-
Download the Excel spreadsheet OnlineWeek2.xlsx, or OnlineWeek2Data again from Week 2.
-
Reserve one the of the data set columns (A-V) different from the one you did for Week 2.
-
Compute , s using Excel, and df (degrees of freedom).
-
Find the 90% t-score using the t-distribution applet. We are not using the z-distribution because we do not know σ, the standard deviation of the population (stdev.p). Thus, we are using the t-distribution because we know only the standard deviation of the sample (stdev.s).
-
Compute the 90% “confidence interval”.
-
Write in words the meaning of your confidence interval.
-
Post your answer to “Week 5 Discussion Group Inputs”.
">