STAT 110 – Fall 2009 – Section 4 - Exam 1 FORM B Name________________________ Answer each of the twenty-five questions below on the scantron sheet using a number two-pencil. Be sure to bubble in your STAT110 ID (as passed around) and mark which form you are using. 1) The mathematics department wants to determine if more computer based exercises help students learn calculus better. There are two possible designs proposed for the study: (A) The researchers randomly assign half of the 160 students registered for the 10:10 MWF time slot to sections using the computers, and the other half to sections not using the computers. At the end of the semester they compare the students’ performance on a common final exam. (B) Offer two 10:10 MWF sections of 40 students each that use the computers, and offer two 10:10 MWF sections of 40 students that don’t use the computers and allow the students to choose their section. At the end of the semester they compare the students’ performance on a common final exam. Which of these two set-ups could provide evidence that computer usage caused a change in performance? A) (A) B) (B) C) Both (A) and (B) D) Neither (A) nor (B) E) Can’t tell from the statement of the problem 2) If a neighborhood association collects sample data about the number of speeding cars through a neighborhood by sitting on a street corner during rush hour, what type of sample has been collected? A) Cluster sample B) Convenience sample C) Simple random sample D) Stratified random sample E) Voluntary response sample 3) If a neighborhood association had a notice in their newsletter inviting residents to email the association president with any comments about the speeding problem, what type of sample would be collected? A) Cluster sample B) Convenience sample C) Simple random sample D) Stratified random sample E) Voluntary response sample 4) An experiment that always over-estimates the value it is trying to determine is said to be A) A census B) An experiment C) An observational study D) Biased E) Highly variableQuestions 5-13 are based on the following description The University of South Carolina decides to conduct a survey of its 27,488 students by telephoning a simple random sample of 400 students from the university’s phone directory. Of the 400 contacted, 238 felt that it was worth raising tuition in order to ensure that courses required for students to graduate on time were always offered as scheduled, and 162 were opposed to this tuition increase. 5) The sampling frame in this example is: A) The 400 people chosen to receive the survey B) The 27,488 students on the USC campus C) The percentage of people in the sample who favor increasing tuition D) The percentage of people in the population who favor increasing tuition E) The students listed in the university telephone directory 6) The target population in this example is: A) The 400 people chosen to receive the survey B) The 27,488 students on the USC campus C) The percentage of people in the sample who favor increasing tuition D) The percentage of people in the population who favor increasing tuition E) The students listed in the university telephone directory 7) If in actuality, 562 numbers were used in order to get the 400 responses desired. What term is used to describe the 162 subjects that chose not to answer this sample survey? A) Dropout B) Nonadherer C) Nonresponse D) Refusal E) Undercoverage 8) If instead of using a simple random sample of 400 numbers they had used every 68th person on the list (1st, 69th, 137th, etc…), it would have been a : A) Cluster Sample B) Matched Pairs Design C) Multi-stage Sample D) Stratified Random Sample E) Systematic Sample 9) Wording the question in such a way that it made the problem sound more severe (“many courses would not be offered”) or less severe (“only a few courses would not be offered”) would likely change the outcome of the survey and misrepresent the students’ actual views. This is an example of a: A) Nonsampling Error B) Processing Error C) Random Sampling Error D) Sampling Error E) UndercoverageREPEATED FROM THE PREVIOUS PAGE Questions 5-14 are based on the following description The University of South Carolina decides to conduct a survey of its 27,488 students by telephoning a simple random sample of 400 students from the university’s phone directory. Of the 400 contacted, 238 felt that it was worth raising tuition in order to ensure that courses required for students to graduate on time were always offered as scheduled, and 162 were opposed to this tuition increase. 10) The percentage ( pˆ) of those surveyed who favored the tuition increase was: A) %4.0004.02381== B) %0.5050.04001== C) %5.6065.02381== D) %5.59595.0400238== E) %1.68681.0239162== 11) What would the margin of error be at the 95% confidence level? A) %4.0004.02381== B) %0.5050.04001== C) %5.6065.02381== D) %5.59595.0400238== E) %1.68681.0239162==12) The parameter is: A) The 400 people chosen to receive the survey B) The 27,488 students on the USC campus C) The percentage of the 400 students who favor increasing tuition D) The percentage of all USC students who favor increasing tuition E) The students listed in the university telephone directory 13) The statistic is: A) The 400 people chosen to receive the survey B) The 27,488 students on the USC campus C) The percentage of the 400 students who favor increasing tuition D) The percentage of all USC students who favor increasing tuition E) The students listed in the university telephone directory 14) It is believed that seniors and graduate students would be more concerned about having the courses they need to graduate being offered than would freshmen, sophomores, and juniors. Much like using blocks in an experiment, using a stratified random sample to help account for this would mean we could no longer use the usual formula for margin of error. It would however: A) Reduce the bias of the parameter B) Reduce the bias of the statistic C) Reduce the non-sampling errors. D) Reduce the variability of the parameter E) Reduce the variability of the statisticQuestions 15-16 are based on the following paragraph A recent poll found that 57% of South Carolina voters opposed the proposed federal bailout plan. The margin of error for this result is +/- 3 percentage