Sei sulla pagina 1di 7

Project 2 Submitted by Derek Harding

1) a. The population is all UT students. b. A survey of Stat 201 students is the sampling frame. c. Non-business students and upperclassmen business students are under-represented or absent in this sampling frame.

2) My sample size for the random sample dataset will be 70+the last digit of my student ID (4) = 74 The following JMP output is from Q32 Hours Prep Exam 1.

3) Confidence Interval for a proportion - Q02 Gender a. The following JMP output is from the full dataset showing p for males.

p=0.48028 b. The following is from the random sample dataset showing and a 94% confidence interval.

= 0.41892 94% Confidence Interval = (0.31717,0.528065) We say that we are 94% confident that the true proportion of males is within the interval. c. The true proportion of males (0.48028) is within the 94% confidence interval of (0.31717,0.528065). We expect 94% of classmates confidence intervals to contain the true proportion.

4) Confidence interval for a mean Q06 GPA a. The following JMP output is from the full dataset

Because there seems to be an implausible value in the full dataset, I will hide and exclude entry 374 of 0. The following is the JMP output without entry 374.

b. c. The following JMP output is from the random sample dataset.

Comment: The Nearly Normal Condition is met in this set of data, because the sample size is > 40 and the data is not extremely skewed. d. e. The following is JMP output from the random sample dataset showing a 96% confidence interval.

The 96% confidence interval for the population is (3.160616,4.410736) We say that we are 96% confident that the true mean GPA is in the interval (3.160616,4.410736). The true mean GPA ( ) is within the 96% confidence interval of (3.160616,4.410736).

5) Hypothesis test regarding the difference in means for independent samples Q06 GPA, Q08 Born in TN? Do People Born in TN have a different average GPA than People Not Born in TN? a. The following JMP output is from the random sample dataset and proves that both samples are normal enough to perform a 2-sample test.

For not born in Tennessee, n<40, but the data is relatively normal. Therefore we can conclude that it is normal enough to run a two-sided hypothesis test. For born in Tennessee, n>40 and the data is not extremely skewed. Therefore we can conclude that it is normal enough to run a two-sided hypothesis test. The normal enough condition passes. b. The following is JMP output for a two-sided hypothesis test about the difference in the population means.

Ho: Ha: ii) Difference in averages: ( Born TN- Not Born TN) = 3.30729-3.24577 =0.06152. This means that the average GPA of people born in Tennessee is 0.06 grade points higher than people not born in Tennessee. iii) P = 0.5791 iv) ( 0.5791 > .05 Fail to reject the Null Hypothesis v) Based upon the two-sided hypothesis test, there is no statistically significant difference in the average GPA of people born in TN vs. people not born in TN. c. Because the true difference in average GPA for the population is not equal to 0, there was a Type II error made (one fails to reject a false null hypothesis). To decrease the frequency of this error without changing , one has to increase the sample size.

i)

6) For this question I used the data from Q09 (Drivers texting or emailing) a. (Question from the Tennessee residents survey) Complete this sentence: I feel that drivers text messaging or emailing are __________ to my personal safety. Very serious threat Somewhat serious Minor threat Not a threat threat 88.5% 8.9% 2.0% 0.4% The following is JMP output from the random sample dataset regarding the same question asked to the stat 201 students.

*Note: For the Stat 201 survey question, there is no answer for not a threat. Also, the answer is worded very serious threat in the Tennessee residence survey and serious threat in the Stat 201 survey. Between the two differences in the way the answers were worded, there may be some bias. b. The following JMP output shows the confidence intervals for the answers in the Stat 201 survey question.

c. Comparing the confidence intervals from the Tennessee residence survey and the Stat 201 survey, none of the results from the Tennessee residence survey fall within the confidence intervals from the Stat 201 survey.(Note: I combined minor threat and not a threat in the Tennessee residence survey to compare it with minor threat in the Stat 201 survey) d. If the results from the survey of Tennessee residents were within the confidence intervals for the Stat 201 survey, it would mean that the Stat 201 survey accurately represents the opinions of Tennessee residents. e. Tennessee residents cover a broad range of people from different ages and amounts of education. The Stat 201 students tend to be of the same age and amount of education. Other factors such as state of origin, race mix, and political background also play a role in this data. This will mean that the answers to this question will differ as the demographics of the two groups differ.

Potrebbero piacerti anche