Presentation is loading. Please wait.

Presentation is loading. Please wait.

Lecture 14 Testing a Hypothesis about Two Independent Means.

Similar presentations


Presentation on theme: "Lecture 14 Testing a Hypothesis about Two Independent Means."— Presentation transcript:

1 Lecture 14 Testing a Hypothesis about Two Independent Means

2 Independence  No relationship between the people or objects in groups  Book example: People who use the internet daily and those who do not

3 Choosing a Hypothesis Test  Testing a single mean Comparing the mean health of 20-year olds to the mean of the general population  Population: general population  Test value: mean health of 20-year olds Testing the hypothesis that on average, the work week is 40 hours long  Population: all workers  Test value: 40 hours

4 Choosing a Hypothesis Test  Testing two independent means Testing the hypothesis that individuals in their 20’s and individuals in their 50’s watch the same amount of TV  People in their 20’s  People in their 50’s

5 Examples  Hypothesis: The average income of a 20 year old is the same as the average income for all Americans  What is the alternative hypothesis?  What type of test should we use?  What are the important variables?

6 Examples  Hypothesis: The mean income for 20 year olds and 50 year old in 2004 is the same  What is the alternative hypothesis?  What type of test should we use?  What are the important variables?

7 Difference Between Sample Means  Test to see if the difference is zero  Difference is normally distributed  Variance decreases as sample size increases

8 Standard Error of Mean Difference

9 Computing the T-Statistic Likeliness of an observed difference

10 SPSS Output

11 Confidence Interval  95% of the time, the confidence interval will contain the true difference  If the confidence interval contains zero, we cannot say the means are not the same

12 Variance  2 ways to estimate standard error: Assume variance of the populations is the same Assume variance of the populations is different Levine test determines whether there is a difference between variances If the significance is small, reject the hypothesis that the two variances are equal

13 Pooled Estimate of Variance

14 Hypothesis Testing  Ho: There is no differences between means  Ha: There is a difference between means  Conduct a t-test to determine if we can REJECT the null hypothesis

15 Example  Individuals making between $10,000 and $12,499 have the same number of years of education (educ) on average as individuals making between $40,000 and $49,999.

16 In-Class Activity  Conduct a hypothesis test to evaluate the following assertions:  Men and women (sex) have the same number of children on average (childs)  Men and women attend school for the same length of time (educ)  Men and women earn the same amount of money on average (rincome06)


Download ppt "Lecture 14 Testing a Hypothesis about Two Independent Means."

Similar presentations


Ads by Google