- By
**cili** - Follow User

- 130 Views
- Uploaded on

Download Presentation
## PowerPoint Slideshow about ' Section 10.1' - cili

**An Image/Link below is provided (as is) to download presentation**

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript

### Section 10.1

Fundamentals of Hypothesis Testing

With more helpful content added by D.R.S., University of Cordele

Null and Alternative Hypotheses

Null and Alternative Hypotheses

Hypothesis testing is a technique for testing a claim about a population parameter using statistical principles.

The alternative hypothesis, denoted by is a mathematical statement that describes a population parameter, and it is the hypothesis that the researcher is aiming to gather evidence in favor of; it is also referred to as the research hypothesis.

The subscript is a lowercase little letter “a”

Null and Alternative Hypotheses

Null and Alternative Hypotheses (cont.)

The null hypothesis, denoted by is the mathematical opposite of the alternative hypothesis; it will always include equality.

The subscript is the digit zero, 0.

This means the Null Hypothesis will

always involve the symbol “=“ or “≥” or “≤”.

And the Alternative Hypothesis will

always involve the symbol “≠” or “<“ or “>”.

Example 10.1: Determining the Null and Alternative Hypotheses

Determine the null and alternative hypotheses for the following scenario.

At Northwest Mississippi Community College, syllabi for online classes state that students should expect to spend 10 hours per week doing coursework for each three-credit-hour class. The director of eLearning for the community college is concerned that students are being required to spend more than 10 hours per week doing work for each three-credit course.

Example 10.1: Determining the Null and Alternative Hypotheses (cont.)

Solution

In order to determine the hypotheses, we must first ask ourselves “What does the director want to gather evidence for?” The eLearning director’s concern is that students are being required to spend more than the 10 hours per week on coursework that the syllabi state will be required. Hence, she is investigating whether the mean is greater than 10 hours. Therefore, the research hypothesis, is The logical opposite of this hypothesis is m ≤ 10, which is the null hypothesis, H0.

Example 10.1: Determining the Null and Alternative Hypotheses (cont.)

Note that the null hypothesis contains a statement about what is currently believed to be true regarding the population mean, that is, m = 10.

Thus, the two hypotheses are written as follows.

Example 10.2: Determining the Null and Alternative Hypotheses

Determine the null and alternative hypotheses for the following scenario.

After learning that Peter Colat of Switzerland could hold his breath for 19 minutes and 21 seconds under water, a group of students decided to conduct their science fair project around the amount of time an average adult can hold his or her breath. Part of the group found research to suggest that adults can hold their breath for 30.0 seconds when submerged in lukewarm water.

Example 10.2: Determining the Null and Alternative Hypotheses (cont.)

The other half of the group found evidence to suggest that the mean time is different if the water temperature is colder. The group would like to test this claim about adults holding their breath in cold water.

Solution

We begin by determining what the group hopes to gather evidence in favor of. They are investigating whether the mean amount of time that an adult can hold his or her breath underwater when the temperature is cold is different from the reported time of 30.0 seconds when the water is lukewarm.

Example 10.2: Determining the Null and Alternative Hypotheses (cont.)

Note that they are simply interested in whether the time is different than 30.0 seconds, not necessarily whether it is longer or shorter than 30.0 seconds. Therefore, we write the research hypothesis mathematically as Ha : m ≠ 30.0. The logical opposite is m = 30.0 and is the null hypothesis, H0 . Therefore, the two hypotheses are written as follows.

Example 10.3: Determining the Null and Alternative Hypotheses

Determine the null and alternative hypotheses for the following scenario.

A leading news authority claims that the President’s job approval rating has dropped over the past three months. Previous polls put the President’s approval rating at a minimum rate of 56%. The president’s chief of staff is concerned about this news report since it is an election year, and he wants to run a test on the claim.

Example 10.3: Determining the Null and Alternative Hypotheses (cont.)

Solution

The news authority’s claim refers to a population proportion, which we write symbolically as p. The claim being tested is that the President’s approval rating has dropped since a previous poll estimated it to be at least 56%. Thus, the research hypothesis is that the President’s approval rating is less than 56%, written mathematically as Ha : p < 0.56.

Example 10.3: Determining the Null and Alternative Hypotheses (cont.)

The logical opposite of the research hypothesis is the null hypothesis, H0: p ≥ 0.56. Therefore, we have the following hypotheses.

Example 10.4: Determining the Null and Alternative Hypotheses

Determine the null and alternative hypotheses for the following scenario.

A survey of Super Bowl viewers found that 8.4% said that they were influenced to buy products from companies that advertise during the game. Because of rising costs, one company president has threatened the advertising team with not renewing the Super Bowl ad spot unless they get the results they want. He says that unless their company’s ad stimulates at least 8.4% of viewers to buy products, then they will not be airing an ad during the Super Bowl next year.

Example 10.4: Determining the Null and Alternative Hypotheses (cont.)

Solution

In order to determine the hypotheses, we need to ask, “What is the company president researching?” He does not want to cancel next year’s Super Bowl ad unless there is overwhelming evidence that the percentage of viewers persuaded to buy the company’s products is too low. So the research hypothesis is that the percentage is less than 8.4%, written mathematically as Ha: p < 0.084.

Example 10.4: Determining the Null and Alternative Hypotheses (cont.)

The mathematical opposite of p < 0.084 is p ≥ 0.084. Hence, the null and alternative hypotheses are written as follows.

Do I write Hypotheses (cont.)hypotheses using μ (mu) or p ?

If the hypotheses are about a population mean, use μ.

If the hypotheses are about a population proportionuse p.

If the hypotheses are about a population percentage, it is really a proportionproblem, so p is the right letter.

Where does the “=” live? Hypotheses (cont.)

The Null Hypothesis, H0, always has the “=” sense: = or ≤ or ≥.

The Alternative Hypothesis, Ha, always has a strict inequality: ≠ or < or >

The possibilities are:

- H0:μ = some value and Ha:μ≠ that value
- H0:μ≤ some value and Ha:μ> that value
- H0:μ≥ some value and Ha:μ< that value
And similarly for proportions with p.

Test Statistic Hypotheses (cont.)

Test Statistic

A test statistic is the value used to make a decision about the null hypothesis and is derived from the sample statistic.

A sample statistic is said to be statistically significant if it is far enough away from the presumed value of the population parameter to conclude that it would be unlikely for the sample statistic to occur by chance if the null hypothesis is true.

Test Statistic Hypotheses (cont.)

Test Statistic (cont.)

The level of significance, denoted by a, is the probability of making the error of rejecting a true null hypothesis in a hypothesis test; a = 1 -c.

The confidence level, c, represents the big area in which Ha “loses”,

because we fail to reject H0,

for lack of strong enough evidence.

The Level of Significance, α (alpha), represents the area in the tail(s), which is the region in which Ha “wins”,

because we reject H0.

Test Statistic Hypotheses (cont.)

Conclusions for a Hypothesis Test

• Reject the null hypothesis.

• Fail to reject the null hypothesis.

- Your hypothesis test will always conclude with one of these two endings.
- Either you have strong enough evidence to reject H0.
- Or you don’t have strong enough evidence.
- We do NOT say we “accept Ha” or “accept H0”.
- Rather, we say we “fail to reject H0”.

Example 10.5: Interpreting the Conclusion to a Hypothesis Test

At Northwest Mississippi Community College, syllabi for online classes state that students should expect to spend 10 hours per week doing coursework for each three-credit-hour class. The director of eLearning for the community college is concerned that students are being required to spend more than 10 hours per week doing work for each three-credit course. A hypothesis test with a 5% level of significance is performed on the director’s claim. The result is to reject the null hypothesis. Does the conclusion support the director’s claim?

Example 10.5: Interpreting the Conclusion to a Hypothesis Test (cont.)

Solution

We must first review the null and alternative hypotheses in order to evaluate the conclusion. Recall from Example 10.1 that the two hypotheses for this hypothesis test are as follows.

Example 10.5: Interpreting the Conclusion to a Hypothesis Test (cont.)

Since the null hypothesis was rejected, the evidence supports the alternative hypothesis, that is, the research hypothesis. Thus, the eLearning director can be confident in her assessment that students study more per week, on average, than the time stated on the syllabi, with a 5% level of significance.

Example 10.6: Interpreting the Conclusion to a Hypothesis Test

Based on historical data, the Board of Education for one large school district believes that the percentage of high school sophomores considering dropping out of school is at least 10%. A high school counselor in the district claims that this percentage is too high. A hypothesis test with a = 0.02 is performed on the counselor’s claim. The result is to fail to reject the null hypothesis. Do the findings support the counselor’s claim?

Example 10.6: Interpreting the Conclusion to a Hypothesis Test (cont.)

Solution

We need to ask, “What does the counselor hope that the data will show evidence for?” The school counselor claims that the percentage of high school sophomores in this district that are considering dropping out of school is less than the assumed rate of at least 10%. Therefore, the research hypothesis is written mathematically as Ha: p < 0.10, and its logical opposite is H0: p ≥ 0.10, so we write the hypotheses as follows.

Example 10.6: Interpreting the Conclusion to a Hypothesis Test (cont.)

Since the test resulted in a failure to reject the null hypothesis, the evidence is not strong enough at this level of significance to support the counselor’s claim that the presumed rate is too high. Remember that performing a hypothesis test does not prove the null hypothesis to be true. Therefore, we can only say that there is not sufficient evidence to reject it.

Test Statistic Test (cont.)

Performing a Hypothesis Test

1. State the null and alternative hypotheses.

2. Determine which distribution to use for the test statistic, and state the level of significance.

3. Gather data and calculate the necessary sample statistics.

4. Draw a conclusion and interpret the decision.

HAWKES LEARNING SYSTEMS Test (cont.)

math courseware specialists

Hypothesis Testing

10.5 Types of Errors

Definitions:

- Type I Error – the probability of rejecting a true null hypothesis, a.
- Type II Error – the probability of failing to reject a false null hypothesis , b.

Does your Decision agree

with Reality ?

Consider: what might lead to an error? Who or What is at fault?

Example 10.7: Determining the Type of Error Test (cont.)

A television executive believes that at least 99% of households in the United States have at least one television. An intern at the executive’s company is given the task of using a hypothesis test to determine whether the percentage is actually less than 99%. The hypothesis test is completed, and based on the sample collected, the intern decides to fail to reject the null hypothesis. If, in reality, 96.7% of households own a television set, was an error made? If so, what type?

Example 10.7: Determining the Type of Error (cont.) Test (cont.)

Solution

Let’s begin by writing the null and alternative hypotheses by asking “What is the intern gathering data for?” Because a hypothesis test is being used to determine if the percentage is less than 99%, the research hypothesis is Ha : p < 099.. The logical opposite is p ≥ 0.99, and this is the null hypothesis. So the two hypotheses are written as follows.

Example 10.7: Determining the Type of Error (cont.) Test (cont.)

The hypothesis test fails to reject the null hypothesis. However, since the reality is that p = 0.967 and thus, the null hypothesis is false, the intern failed to reject a false null hypothesis. This is a Type II error.

Example 10.8: Determining the Type of Error Test (cont.)

Insurance companies commonly use 1000 miles as the mean number of miles a car is driven per month. One insurance company claims that, due to our more mobile society, the mean is more than 1000 miles per month. The insurance company tests its claim with a hypothesis test and decides to reject the null hypothesis. Assume that in reality, the mean number of miles a car is driven per month is 1250 miles. Was an error made? If so, what type?

Example 10.8: Determining the Type of Error (cont.) Test (cont.)

Solution

Begin by writing the null and alternative hypotheses. The insurance company wishes to gather data in support of the statement that the mean is more than 1000 miles. Therefore, the research hypothesis is

Ha : m > 1000. The opposite is m ≤ 1000. Thus, the null and alternative hypotheses are written as follows.

Example 10.8: Determining the Type of Error (cont.) Test (cont.)

The decision was to reject the null hypothesis. The null hypothesis is false since m = 1250, so the decision was to reject a false null hypothesis, which is a correct decision.

Example 10.9: Determining the Type of Error Test (cont.)

A study on the effects of television-viewing on children reports that children watch a mean of 4.0 hours of television per night. Kiko believes that the mean number of hours children in her neighborhood watch television per night is not 4.0. She performs a hypothesis test and rejects the null hypothesis. Assume that in reality, children in her neighborhood do watch a mean of 4.0 hours of television per night. Did she make an error? If so, what type?

Example 10.9: Determining the Type of Error (cont.) Test (cont.)

Solution

Begin by writing the null and alternative hypotheses. Kiko wishes to gather data in support of her belief that the mean is not 4.0 hours per night. Therefore, her research hypothesis is written mathematically as

Ha: m ≠ 4.0. The logical opposite is H0: m = 4.0. Thus, the null and alternative hypotheses are written as follows.

Example 10.9: Determining the Type of Error (cont.) Test (cont.)

The decision was to reject the null hypothesis, when in reality, m = 4.0, so Kiko rejected a true null hypothesis. This is a Type I error.

Download Presentation

Connecting to Server..