Section 10.2

Chi-Square Tests and the F-Distribution

143

CHAPTER

10

Section 10.2 4Example 3 (pg. 526) Chi-Square Independence Test Test the hypotheses: H o :The number of days spent exercising per week is independent of gender vs. H a : The number of days spent exercising per week depends on gender. The first step is to enter the data in the table into Matrix A. On the TI-83, press MATRX, highlight EDIT and press ENTER. On the TI-83 Plus, press 2nd [MATRX] . (MATRX is found above the x-1 key.) Highlight EDIT and press ENTER.

On the top row of the display, enter the size of the matrix. The matrix has 2 rows and 4 columns, so press 2 , press the right arrow key, and press 4. Press ENTER. Enter the first value, 40, and press ENTER. Enter the second value, 53, and press ENTER. Continue this process and fill the matrix.

Press 2nd [Quit] . To perform the test of independence, press STAT, highlight TESTS, and select C: χ2-Test and press ENTER.

144 Chapter 10 Chi-Square Tests and the F-Distribution

For Observed, [A] should be selected. If [A] is not already selected, press MATRX, highlight NAMES, select 1:[A] and press ENTER. For, Expected, [B] should be selected. Move the cursor to the next line and select Calculate and press ENTER.

The output displays the test statistic and the P-value. Since the P-value is greater than α , the correct decision is to Fail to Reject the null hypothesis. This means that the number of days per week spent exercising is independent of gender. Or, you could highlight Draw and press ENTER.

This output displays the χ2 –curve with the area associated with the P-value shaded in. The test statistic and the P-value are also displayed.

3

Section 10.2

145

4Exercise 13 (pg. 529) Chi-Square Test for Independence Test the hypotheses: H o :The Result (improvement or no change) is independent of Treatment (drug or placebo) vs. H a : The Result depends on the Treatment. The first step is to enter the data in the table into Matrix A. Press MATRX, highlight EDIT and press ENTER. On the top row of the display, enter the size of the matrix. The matrix has 2 rows and 2 columns, so press 2 , press the right arrow key, and press 2. Press ENTER. Enter the first value, 39, and press ENTER. Enter the second value, 25, and press ENTER. Continue this process and fill the matrix.

Press 2nd [Quit] . Press STAT, highlight TESTS, and select C: χ2-Test and press ENTER. For Observed, [A] should be selected. If [A] is not already selected, press MATRX, highlight NAMES, select 1:[A] and press ENTER. For, Expected, [B] should be selected. Move the cursor to the next line and select Calculate and press ENTER.

The output displays the test statistic and the P-value. Since the P-value is less than α , the correct decision is to Reject the null hypothesis. The correct recommendation is to use the drug as part of the treatment.

3

146 Chapter 10 Chi-Square Tests and the F-Distribution

Section 10.3 4Example 3 (pg. 537) Performing a Two-Sample F-Test Test the hypotheses: H o : σ 1 ≤ σ 2 vs. H a : σ 1 > σ 2 . The sample statistics 2

2

2

2

are: s1 = 144, n1 = 10, s 2 = 100 and n2 = 21. Press STAT, highlight TESTS and select D:2-SampFTest. For Inpt, select Stats and press ENTER. On the next line, enter s1 , the standard deviation under the old system. The standard 2

2

deviation is the square root of the variance, so press 2nd [

] and enter 144.

Enter n1 on the next line. Next, enter the standard deviation under the new system by pressing 2nd [ ] 100. Enter n 2 on the next line. Highlight > σ 2 for the alternative hypothesis and press ENTER. Select Calculate and press ENTER.

The output displays the alternative hypothesis, the test statistic, the P-value and the sample statistics. Since the P-value is greater than α , the correct decision is to Fail to Reject the null hypothesis. There is not enough evidence to conclude that the new system decreased the variance in waiting times. You could also select Draw and press ENTER.

Section 10.3

147

The output displays the F-distribution with the area associated with the P-value shaded in.

3

148 Chapter 10 Chi-Square Tests and the F-Distribution 4 Exercise 19 (pg. 540)

Comparing Two Variances

Test the hypotheses: H o : σ 1 ≤ σ 2 vs. H a : σ 1 > σ 2 . The sample statistics 2

2

2

2

are: s1 = 0.7, n1 = 25, s 2 = 0.5 and n2 = 21. Press STAT, highlight TESTS and select D:2-SampFTest. For Inpt, select Stats and press ENTER. On the next line, enter s1 , the standard deviation under the old procedure. Enter n1 on the next line. Next, enter the standard deviation and sample size for the new procedure and press ENTER. Highlight > σ 2 for the alternative hypothesis and press ENTER. Select Calculate and press ENTER.

The output displays the alternative hypothesis, the test statistic, the P-value and the sample statistics. Since the P-value is less than α , the correct decision is to Reject the null hypothesis. The data supports the hospital's claim that the standard deviation of the waiting times has decreased.

3

Section 10.4

149

Section 10.4 4 Example 2 (pg. 547) Performing an ANOVA Test Test the hypotheses: H o : µ1 = µ 2 = µ 3 vs. H a : at least one mean is different from the others. Enter the data into L1, L2, and L3. Press STAT, highlight TESTS and select F:ANOVA( and press ENTER, 2nd [L1] , 2nd [L2] , 2nd [L3].

Press ENTER and the results will be displayed on the screen.

The output displays the test statistic, F = 6.13, and the P-value, p = .00638. Since the P-value is less than α , the correct decision is to Reject the null hypothesis.

150 Chapter 10 Chi-Square Tests and the F-Distribution This indicates that there is a difference in the mean flight times of the three airlines. The output also displays all the information that is needed to set up an ANOVA table like the table on pg. 544 of your textbook. Variation Between Within

Sum of Squares 1806.5 3978.2

Degrees of Freedom 2 27

Mean Squares 903.2 147.3

F 903.2 / 147.3

Note: The TI-83 labels variation "Between" samples as variation due to the "Factor". Also, variation "Within" samples is labeled as "Error". The final item in the output is the pooled standard deviation, Sxp = 12.14.

3

Section 10.4

151

4 Exercise 5 (pg. 549) Performing an ANOVA Test Test the hypotheses: H o : µ1 = µ 2 = µ 3 vs. H a : at least one mean is different from the others. Enter the data into L1, L2, and L3. Press STAT, highlight TESTS and select F:ANOVA( and press ENTER, 2nd [L1] , 2nd [L2] , 2nd [L3]. Press ENTER and the results will be displayed on the screen.

The output displays the test statistic, F = 0.77, and the P-value, p = 0.4752. Since the P-value is greater than α , the correct decision is to Fail to Reject the null hypothesis. The data does not support the claim that at least one mean cost per month is different.

3

152 Chapter 10 Chi-Square Tests and the F-Distribution

4 Exercise 14 (pg. 552) Performing an ANOVA Test Test the hypotheses: H o : µ1 = µ 2 = µ 3 = µ 4 vs. H a : at least one mean is different from the others. Enter the data into L1, L2, L3 and L4 . Press STAT, highlight TESTS and select F:ANOVA( and press ENTER, 2nd [L1] , 2nd [L2] , 2nd [L3] , 2nd [L4]. Press ENTER and the results will be displayed on the screen.

The output displays the test statistic, F = 8.46, and the P-value, p = .0002196. Since the P-value is less than α , the correct decision is to Reject the null hypothesis. The data supports the claim that the mean energy consumption for at least one region is different.

3

Technology

4 Technology (pg. 563)

153

Teacher Salaries

Exercise 2: One method of testing to see if a set of data is approximately normal is to use a Normal Probability Plot. To test the data in L1, press 2nd [STAT PLOT] , select Plot 1 and turn ON Plot1. For Type, select the last icon on the second line. For Data List, press 2nd [L1]. For Data Axis, select X. For Mark, select the first icon. Press ZOOM and 9 to display the normal plot.

Data Data that is approximately normally distributed will have a plot that looks fairly linear. Although this graph is not perfectly straight, it is fairly linear and it is therefore reasonable to conclude that the data is approximately normal. Repeat this process to check the other datasets for normality. 2 Exercise 3: Test the hypotheses: H o σ 1 = σ 2 vs. H a : σ 1 ≠ σ 2 for each pair 2

2

2

of samples. This test requires values for s1 and s 2 . To obtain these values, enter the data into L1, L2 and L3. To find the standard deviation for L1, press STAT, highlight CALC and press L1 ENTER. To find the standard deviation for L2, press STAT, highlight CALC, press [L2] ENTER. Repeat this process to get the standard deviation of L3. For each comparison, press STAT, highlight TESTS and select D:2-SampFTest. 2 2 For Inpt, select Stats and press ENTER. To compare σ 1 andσ 2 , enter s1 , the sample standard deviation. Enter n1 , the corresponding sample size, on the next line. Next, enter s 2 and n2. Highlight ≠ σ 2 for the alternative hypothesis and press ENTER. Select Calculate and press ENTER. Repeat this process to 2

compare σ 1 andσ 3 . Finally compare σ 2 andσ 3 . 2

2

2

2

Exercise 4: Test the hypotheses: H o : µ1 = µ 2 = µ 3 vs. H a : at least one mean is different from the others. Press STAT, highlight TESTS and select F:ANOVA( and press ENTER, 2nd [L1] , 2nd [L2] , 2nd [L3] . Press ENTER and the results will be displayed on the screen. Exercise 5: Repeat Exercises 1 - 4 using this second set of data.