* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download File
Survey
Document related concepts
Transcript
Lesson 6: Section 9.3 (part 2) Recognize paired data and use one-sample t procedures to perform significance tests for such data. Comparative inference is used more than onesample inference because those studies are more convincing. Paired data involves making two observations on the same individual or one observation on each of two similar individuals. If the conditions for inference are met, we can use one-sample t procedures to perform inference about the mean difference µ0. These methods are sometimes called paired t procedures. For their second semester project in AP Statistics, Leighton and Brooke decided to investigate which line was faster in the supermarket: the express lane or the regular lane. To collect their data, they randomly selected 15 times during a week, went to the same store, and bought the same item. However, one of them used the express lane and the other used a regular lane. To decide which lane each of them would use, they flipped a coin. If it was heads, Leighton used the express lane and Brooke used the regular lane. If it was tails, Leighton used the regular lane and Brooke used the express lane. They entered their randomly assigned lanes at the same time, and each recorded the time in seconds it took them to complete the transaction. Time in express lane Time in regular lane (seconds) (seconds) 337 342 226 472 502 456 408 529 151 181 284 339 150 229 357 263 349 332 257 352 321 341 383 397 565 694 363 324 85 127 Carry out a test to see if there is convincing evidence that the express lane is faster. Since these data are paired, we will consider the difference in time (regular – express). Here are the 15 differences. In this case, a positive difference means the express lane was faster. Regular - Express (seconds) 5 246 -46 121 30 55 79 -94 -17 95 20 14 129 -39 42 STATE want to test the following hypotheses at the 0.05 significance level H0: µd =0 Ha: µd > 0 where µd= the true mean difference (regular – express) in time required to purchase an item at the supermarket. We PLAN If conditions are met, we will perform a paired t test for µd. Random: A random sample of times to make the purchases was selected, and the students were assigned to lanes at random. Normal: We don’t know if the population distribution of difference is approximately Normal and we don’t have a large sample size, so we will graph the differences and look for any departures from Normality (draw a dot plot or box plot or Normal Probability plot). Slightly skewed right with no outliers, so it is reasonable to use t procedures for these data. Independent: One mean time should not affect another since we randomly selected the times to conduct the study. DO Using One-Variable STATS in the calculator, the sample mean difference is x 42.7 seconds with a standard deviation of sx = 84.0 seconds. 42.7 0 Test statistic: t 84.0 15 P-value P(t>1.97) using the t distribution with 15 1 = 14 degrees of freedom. Using Table B, I find that the P-value is between .05 and 0.025. Using technology, P-value = tcdf(1.97,100,14) = 0.034. CONCLUDE the P-value is less than (.034<.05), we reject the null hypothesis. There is convincing evidence that the express lane is faster than the regular lane, on average. Since When you’re working with paired data, the Independent condition refers to the differences. When calculating the P-value, be sure to report the degrees of freedom and please be specific if with your method (how you used your calculator or Table B). Significance does not mean important!! Example: Suppose that a large school district implemented a new math curriculum for the current school year. To see if the new curriculum is effective, the district randomly selects 500 students and compares their scores on a standardized test after that curriculum change with their scores on the same test before the curriculum change. The mean improvement xd 0.9 with a standard deviation of sd = 12.0. (a)Calculate the test statistic and P-value for a test of H0: µd =0 and Ha: µd > 0. 0.9 0 t 1.68 12 500 P-value = tcdf(1.68, 100, 499) = 0.047 (b) Are the results significant at the 5% level? Since the P-value is less than 0.05 (barely), the results ARE significant at the 5% level. Then we reject the null of no difference and conclude that there was an improvement. (c) Can we conclude that the new curriculum is the cause of the apparent increase in scores? No! Even though the increase was statistically significant, we cannot conclude that the new curriculum was the cause of the increase (we would need a control group for comparison and random treatment of assignments!) If there is a lack of significance, don’t ignore it! Address that you have absence of evidence! Example: When an AP Statistics class did caffeine and pulse rates experiment, the confidence interval for the difference in average pulse rate changes was (-2.52, 4.59). This means that drinking soda with caffeine can increase pulse rates by as much as 4.59 beats per minute, on average OR decrease pulse rates by as much as 2.52 beats per minute, on average. (compared with drinking soda with no caffeine). Because of this, it seems that caffeine does not affect pulse rates (our results are not statistically significant), but we need more data to estimate this effect. Do not use inference when… Your experimental design is poor (you can’t make inference when your data collection was biased!) Don’t use it when you already have the entire population’s parameters! Example: Suppose that you wanted to know the average GPA for students at your school who are enrolled in AP Statistics. Since this isn’t a large population, you conduct a census and record the GPA for each student. Is it appropriate to construct a one-sample t interval for the population mean GPA? NO!! You already have all the GPAs for every student – we only use inference to make a conclusion about something unknown about the population from SAMPLES Search data for patterns; however, do NOT search for significance! Running one test and reaching the significance level is reasonably good evidence that you found something. Running 20 tests and finding ONLY ONE out of 20 reaching the significance level is NOT reasonably good evidence. Assigned reading: p. 565-587 HW problems: Lesson 5 Lesson 6 Check p. 564 #57-60; p. 586 #63, 67, 71, 73 p. 588 #75, 77, 80, 82, 89, 94-97, 99-104 answers to odd problems.