University of Florida



STA 3024 Fall 2015Exam 1 RIPOLNAME:______________________________________Section: 4432 - 2nd pd TEST FORM CODE: B (top right corner of scantron) UF ID # ____________________________________ SPECIAL CODE: 21 (on bottom of scantron)"On my honor, I have neither given nor received unauthorized aid on this examination."Signature: _________________________Questions 1 – 5 Coffee is one of the most consumed beverages in the world. It can have a stimulating effect in humans due to its caffeine content. The effect of coffee on human health has been a subject of many studies; however, results have varied in terms of coffee's relative benefit.?The majority of recent research suggests that moderate coffee consumption (4 eight-ounce cups per day or less) is benign or mildly beneficial in healthy adults. For each of the following statements, determine the type of parameter it is estimating. Note that you can use each alternative more than once and you don’t need to use them all.one meanone proportiondifference of two independent meansdifference of two independent proportionsmean of matched pair differencesJust over 83 percent of U.S. adults drink coffee at least occasionally.Those who drink coffee at least once per week represent about 75 percent of the population.41 percent of 18-24 year olds are drinking coffee each day —up from the 31 percent of this age group who said they had a daily cup of coffee during 2010. Men drink as much coffee as women; each consuming an average of 1.6 cups per day.?Most adult coffee drinkers said their lifelong habit began during their teenage years, with an average age of 16.5. 6. When interpreting confidence intervals, we look for zero to determine significant differences any time:a) the confidence interval is for comparing two groupsb) the confidence interval is for comparing two independent meansc) the confidence interval is done with Bonferroni’s method for multiple comparisonsd) the confidence interval is done with Tukey’s or Fisher’s method for multiple comparisonse) all of the aboveQuestions 7-10 An experiment was conducted to compare 5 groups, with 8 observations each. An ANOVA F test was run on the data. Here are a portion of the results. Source df SS MS F Treatments Error 289.333 Total 12312 7. The value of the test statistic is: a) 1.89b) 0.78 c) 1.51 d) 10.29 e) 5.368. The best estimate for the common standard deviation is: a) 87.290b) 16.703c) 46.667d) 17.009 e) 83.7149. Under the null hypothesis, the test statistic has an F distribution with degrees of freedom in the numerator and denominator: a) 5 and 8b) 8 and 40c) 4 and 35d) 7 and 32e) 4 and 7 10. When comparing the group means to each other, how many comparisons do we make?5b) 8c) 10d) 28e) 4011. When using Bonferroni’s method for multiple comparisons:a) we find significant differences whenever two means are less than the margin of error apartb) we can order the means and draw lines connecting those that are significantly differentc) we adjust the individual confidence level of each interval so we can achieve the desired family confidenced) we don’t need to assume equal variances for all the groups since the sample sizes have to be the samee) all of the above12. A recent study reported by identified the majors with the highest average starting salaries for recent graduates. The top 5 on their list were Computer Science ($66,800), Engineering ($65,000), Mathematics and Statistics ($60,300), Economics ($58,600) and Finance ($58,000). With this information we can conclude that the average starting salary for Mathematics and Statistics graduates is:a) significantly less than that for Computer Scienceb) significantly less than that for Engineeringc) significantly more than that for Economicsd) significantly more than that for Financee) none of the above356616023876000Questions 13 – 16 These data concern advertising by a company in local newspapers. The same ad was placed in different sections of the newspapers (news, business and sports sections) on different days of the week, and the number of inquiries resulting from each advertisement recorded.13. The response variable studied here is:a) sectionb) dayc) sportsd) inquiriese) Friday14. This study has ___ factors and ___ treatments:a) 2, 15b) 5, 3c) 3, 5d) 1, 5e) none of those15. The plot of means suggests that there will be:a) no main effectsb) no interactionsc) interactionsd) one significant factore) more than one significant factor16. Which of the following statements is correct?a) The best day to advertise in the news section is Thursday.b) Overall, advertising in the business section seems to do worse than the rest.c) Advertising in the sports section is best on Mondays.d) The best section to advertise in depends on the day of the week.e) all of the above17. On an F distribution with df1=5, df2=15 we reject the null hypothesis at α = 0.05 if:a) the test p-value is larger than 0.05b) the test statistic is smaller than 2.90c) the test statistic is larger than 2.90d) both a and b are correcte) both a and c are correct42291005969000Questions 18 – 26 A researcher is studying treatments for agoraphobia with panic disorder. The treatments are the drug Imipramine at the doses 1.5 mg per kg of body weight and 2.5 mg per kg of body weight. There will also be a control group given a placebo. Thirty patients were randomly divided into three groups of 10 each. One group was assigned to the control, and the other two groups were assigned to the two treatments. After 24 weeks on treatment, each of the subject's symptoms were evaluated through a battery of psychological tests, where high scores indicate a lessening of symptoms. Results are given below. Source DF SS MS F PFactor 2 2257.3 1128.6 33.33 0.000Error 27 914.2 33.9Total 29 3171.5 Individual 95% CIs For Mean Based on Pooled StDevLevel N Mean StDev -------+---------+---------+---------+--control 10 70.400 5.296 (----*----)dose=1.5 10 84.900 5.626 (----*----)dose=2.5 10 91.100 6.471 (----*----) -------+---------+---------+---------+-- 72.0 80.0 88.0 96.0Pooled StDev = 5.819Tukey 95% Simultaneous Confidence IntervalsAll Pairwise ComparisonsIndividual confidence level = 98.04%control subtracted from: Lower Center Upper ---+---------+---------+---------+------dose=1.5 8.041 14.500 20.959 (------*-----)dose=2.5 14.241 20.700 27.159 (------*-----) ---+---------+---------+---------+------ -10 0 10 20dose=1.5 subtracted from: Lower Center Upper ---+---------+---------+---------+------dose=2.5 -0.259 6.200 12.659 (-----*------) ---+---------+---------+---------+------ -10 0 10 20 18. How many replications are used in this study?a) 30b) 29c) 27d) 10e) 319. At a significance level of 0.05 we can conclude from the ANOVA test that the population means of the three groups are:a) all equalb) all differentc) not all equal d) equal to their variances e) not equal to their variances20. When interpreting the individual confidence intervals for means provided by the output, what would be the family confidence level?a) 95% b) 98%c) 85%d) 96% e) 94%21. When making all pairwise comparisons for this problem using the Bonferroni procedure with 97% family confidence, what should be the value of t to use?a) 2.756b) 2.771c) 2.462d) 2.473 e) 2.05222. Do the results of the All Pairwise Comparisons procedure on the output agree with the results of the Individual Confidence Intervals also provided on the output?Yes, they both found significant differences in one of the three comparisons.Yes, they both found significant differences in two of the three comparisons.No, the Individual CIs were able to find more significant differences.No, the All Pairwise Comparisons procedure was able to find more significant differences.It’s impossible to compare the results of these procedures, since their purposes are completely different.23. If the margin of error for the Bonferroni procedure is 7.23, the results can be summarized as:a)Control b)Controlc)Controld)Controle)ControlDose 1.5Dose 1.5Dose 2.5Dose 2.5Dose 1.5Dose 2.5Dose 2.5Dose 1.5Dose 1.5Dose 2.524. Are there any problems with the assumptions necessary for the validity of these statistical analyses?No – there seems to be no obvious problem with any of the assumptions.Yes – there is a problem with the assumption of normality. Yes – there is a problem with the assumption of equal variances.Yes – there is a problem with the assumption of random samples. Yes – there is a problem with the assumption of large enough samples.25. If we wanted to determine if the average score of patients taking Imipramine at 2.5 mg per kg of body weight is higher than those taking the drug at 1.5, the correct test to perform would be:an F test where Ho:?1=?2= ?3an F test where Ha:?1<?2< ?3a t test where Ha: ?2- ?3>0a t test where Ha: ?2- ?3<0a t test where Ha: ?2- ?3≠026. Based on the confidence interval provided on the output comparing the dose2.5 and dose1.5 treatments we can predict that the test in the previous question will result in:a) a p-value = 0.000b) a p-value < 0.000c) very strong evidence to say there is NO significant differences between those two treatmentsd) very strong evidence to find significant differences between those two treatmentse) not enough evidence to find significant differences between those two treatments ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download