You are on page 1of 8

Midterm 1 – STAT 231 Date: February 3rd , 2011: 4-30-6-00 p.m.

_________________________________________________________________
________

NAME: (Print) ____________________, _________________________ Family Name Given Name UW Student ID Number: ___________________________

Question 1 Question 2 Question 3

Total Points 10 points 20 points 10 points

Points Scored

They select 3000 names from the list and call assess what proportion of eligible voters plan to vote and. what proportion Six weeks before a national election in Canada. Midterm 1 Spring 2010 Solutions 1. A polling company has a list of eligible voters and their telephone the in table below. assess what proportion of eligible voters plan to vote and. of those. (6 marks) thebefore following and thenelection fill in the spaces based on what you have SixRead weeks a national in Canada. a political party commissions a poll to numbers in the 40 largest cities in Canada. of those. (6 marks) Read the following and then fill in the spaces based on what you have read.1) Read the following and then fill in the spaces based on what you have read: Midterm 1 Spring 2010 Solutions 1. Of these. what proportion them. They select 3000 names from the list and call them. Of these. support the party. This could lead to study errorThe polling firmed only called eligible voters in urban areas. There is ! mark for each correct answer. There is ! mark for each correct answer. This could lead to study error [ urban eligible voters may have different views that rural eligible voters – this is a [ urban eligible may have different views that rural eligible voters – this is a difference between the targetvoters and study populations] Many of the people called refused to participate. This could lead to [ this is a possible sampledifference error between the study population and the sample] difference between the target and study populations] [ this is a possible difference between the study population and the sample] . 1104 eligible voters agree to participate in the survey with the results support the party. a political party commissions a poll to read. 1104 eligible voters agree to participate in the survey with the results Support party summarized in the table below. A polling company has a list of eligible voters and their telephone Plan to Vote YES Support party YES 351 Plan to Vote YES NO 107 NO YES NO 351 107 381 265 NO 381 265 A unit is an eligible voter [ eligible a votervoter is not an acceptable answer] A unit is an [ a voter is not an acceptable answer] The target population is all eligible voters A variate is whether or not an eligible voter plans to vote or whether or not an eligible The target population is all eligible voters [note the target population must be a collection of units as defined above] [note the target population must be a collection of units as defined above] A variate isvoter whether or not an eligible supports the party voter plans to vote or whether or not an eligible voter supports the party [a variate is a quantity or characteristic associated with a unit] [a variate is a quantity or characteristic associated with a unit] The studyis population all eligible voters on the list The study population all eligible is voters on the list [ these the units might [ these are the unitsare that might be that included in be theincluded sample] in the sample] The sampleThe is the 1104 eligible voterseligible who respond towho the questions eligible sample is the 1104 voters respond or to the the3000 questions or the 3000 eligible voters who voters were called who were called [ the first answer is better – since the sample is the set of units the of study population [ the first answer is better – since the sample isfrom the set units from the study population on which we get values for the variated] on which we get values for the variated] The polling firmed only called eligible voters in urban areas. numbers insummarized the 40 largest in cities Canada. This could lead to sample error Many of the people called refused to participate.

ABBA.1. [if you could not solve part a). Attribute 1 proportion of units who plan to vote estimate: 732/1104 Attribute 2 of those who plan to vote. 2 for the probability function]. use p ( y ) 1/ 3. the subject is asked to decide which two glasses are brand A. there are 6 possible orders { AABB. ABAB. A numerical measure of the association between planning to vote and support of a Political party 2. Enter your answer in the table below. In a tasting experiment. the proportion estimate: 351/732 who support the party [ note the condition in the second attribute ] In a formal statistical analysis. two of his brand and two of a competing brand (B). E[Y ] sd [Y ] 0 u 1/ 6  1u 4 / 6  2 u 1/ 6 1 Var[Y ] (0  1) 2 u1/ 6  (1  1) 2 u 4 / 6  (2  1) 2 u 1/ 6 1/ 3 2 . BBAA. BAAB} we have y 1 with probability 4/6 If the outcome is {BBAA} we have y 0 with probability 1/6 We get y p( y ) 0 1/6 1 4/6 2 1/6 b) (3 marks) Find the mean and standard deviation of Y. BABA. In each case. BAAB} each with probability 1/6 If the outcome is { AABB} we have y 2 with probability 1/6 If the outcome is { ABAB. Many beer drinkers claim that they can identify their favourite brand (A) based on is taste alone. Based on the smell and taste. a reasonable probability model to describe the data is a multinomial distribution. a subject is given 4 glasses of beer. If the subject guesses which beer is which. describe the attribute and provide an estimate based on the given data (leave your estimates as fractions). The response variate is the number of correctly identified glasses a) (2 marks) If the subject cannot identify his brand and is just guessing. ABBA. BABA.There are two attributes of interest in the target population. y 0. Suppose we order the glasses A A B B. specify the probability function of a discrete random variable Y that would describe the behaviour of the response variate. The subject is told that there are two glasses RELATIVE RISK of each brand (A or B).

") distributed random variable. The data is denoted by c1.809 kmph________________________ c) The Arithmetic mean of two numbers is 10 and the sample standard deviation is 2!2. b) A man travels 60 kilometers at 25 kmph.2) Fill in the blanks for the following questions: (2+2+2+2+4+8=20 points) a) The “Plan” part of the PPDAC analysis may involve one of the following aspects: a ___Descriptive___________ aspect.. Mean cost = $2500 Range of Costs= $7500 Sample Standard Deviation of Costs= $5500 The relationship between Cost and Revenue is given by ! .... then the likelihood function is ! 10 L(" ) = $ " x i (1 # " )1# x i i =1 10 =" i =1 % xi (1 # " ) 10 # % x i i =1 10 ! e) Suppose we have data for the costs of production for a firm every month from January 2001 to December 2002. The following information s calculated from the data set.x10 are a sequence of independent realizations of a Bernoulli(") (or Binomial(1. and 100 km at 30 kmph. His average speed for the entire journey is __23. x 3 .. x 2 . d) If x1. 40 kilometers at 15 kmph. The numbers are _12________ and ______8___. c 2 .c 24 .. a ____Causative_______ aspect or a ___Predictive_____________aspect.

.5328__________________ Let W = max{ X1.500_____________ The Range of the Revenue is ________$52.X2. The minimum score you need to get in order to achieve an A is __84.500________ The variance in revenues of the data set is __38500 $^2____________ The scores in the Final of STAT 231 has a Normal distribution with mean 75 and standard deviation 6. The corresponding random variables are denoted as X1. you need to be in the top 5 percentile of the class..7967)^9=0.129___________________ ! In order to get an A.... The average Revenue for our data set is ___--16. i=1. X 2 .. _ (i) (ii) The probability that the sample mean X lies between 74 and 77 is _________0.…24 Answer the following questions: (i) (ii) (iii) (iv) f) ! Converting the data from Costs to Revenues is an example of a ________Affine Transformation_________of the Data.87___________________ (iii) ! . The probability that W is less than 80 is ___(0.X 9 } .Ri = "7c i + 1000 .X9.2. An independent sample of size 9 is drawn from this distribution.

2 i =1 " xi 6 ! ! . (8 points) a) Find the Maximum Likelihood estimate for ".3) The number of failures before the first success can be described as a random variable with probability function f ( y ) = (1 " # ) y # . based on the y’s.2.8 0.3(up to three decimal places) The MLE = 1/(1+4)=1/5=0.36 0. y3.36 = R(" ) = = 6 " xi 0. the first order condition gives us n n ^ #x = i =1 i ! .. y=0.2 L(0.1.….2 6 L(0.2) 6 i =1 0. Suppose we have n independent observations y1.7 24 0. which gives us the solution " ^ 1$" ^ ! "= n n = 1 _ n + # xi i =1 1+ x ! b) Suppose that n=6.…yn. Evaluate the Relative Likelihood function at "=0. Show all your calculations.7 0. ! n yi L(" ) = $ (1 # " ) " i =1 = (1 # " ) i =1 % xi n "n n The log-likelihood function l(" ) = # x i ln(1 $ " ) + n ln(" ) i =1 ! Taking derivatives. and the sample average is 4 days..3) 0. y2.8 24 0.

2 ! ! Scratch Copy . evaluate the log likelihood function at the maximum likelihood point. n The log-likelihood function = l(" ) = # x i ln(1 $ " ) + n ln(" ) i =1 24 ln0.c) Using the values in (b).8 + 6ln0.

8340 0.9989 0.0 1.8078 0.9319 0.9573 0.07 0.6443 0.9772 0.9357 0.8485 0.9952 0.9115 0.9345 0.5279 0.5239 0.9966 0.9857 0.9977 0.8289 0.9981 0.7190 0.8686 0.9934 0.7324 0.9962 0.9 3.9989 0.9875 0.5557 0.9713 0.2 1.6 1.9788 0.5596 0.9633 0.9599 0.7580 0.9279 0.8554 0.9977 0.5 2.9664 0.6331 0.4 0.5 0.9956 0.9986 0.1 0.9505 0.1: The Cumulative Distribution Function F (x) = P (X ≤ x) for a G(0.9292 0.6141 0.1 2.9972 0.6026 0.9854 0.9686 0.9967 0.3 1.9906 0.02 0.9332 0.9878 0.9987 0.5080 0.03 0.9591 0.9951 0.08 0.9625 0.6772 0.8186 0.9049 0.8264 0.9706 0.7734 0.9979 0.7995 0.8830 0.9938 0.7019 0.9953 0.9394 0.9207 0.9850 0.9893 0.8621 0.9032 0.9974 0.7673 0.9864 0.9756 0.9750 0.9965 0.9382 0.2 0.5199 0.9767 0.5517 0.9969 0.7454 0.5160 0.6179 0.9980 0.8770 0.6217 0.9940 0.9949 0.9955 0.9916 0.6 0.6950 0.9968 0.8925 0.9911 0.9608 0.9535 0.8023 0.9971 0.6736 0.6517 0.6406 0.9976 0.9616 0.9918 0.9265 0.9913 0.9959 0.5000 0.5832 0.7123 0.6664 0.9982 0.5636 0.9429 0.7486 0.6255 0.8212 0.9904 0.7852 0.9817 0.9932 0.9984 0.9964 0.9452 0.9236 0.9564 0.8133 0.9798 0.9989 0.9884 0.9981 0.5793 0.9945 0.9545 0.5675 0.9985 0.9922 0.5120 0.9925 0.9846 0.9656 0.9898 0.5910 0.5478 0.8365 0.8238 0.9162 0.9830 0.6985 0.9418 0.9826 0.7157 0.9066 0.06 0.9719 0.0 0 0.5398 0.7 1.8438 0. APPENDIX: STATISTICAL TABLES Cumulative distribution for a G(0.9973 0.9987 0.9990 0.8962 0.9732 0.9370 0.9868 0.9838 0.9984 0.8708 0. 1) random variable .7224 0.8531 0.9131 0.9693 0.9726 0.5987 0.9979 0.9554 0.3 2.8729 0.8599 0.7357 0.9 2.9222 0.6480 0.5319 0.9082 0.8413 0.8869 0.9515 0.176 CHAPTER 8.6915 0.7764 0.9484 0.5871 0.9306 0.8997 0.8944 0.8577 0.9871 0.5438 0.6879 0.8315 0.8749 0. 1) random variable x 0.01 0.0 0.9582 0.9985 0.7967 0.9881 0.9495 0.7823 0.8 1.7549 0.9463 0.9699 0.9987 0.9474 0.7910 0.8 2.9761 0.9983 0.9957 0.9793 0.7291 0.8888 0.05 0.2 2.7088 0.6808 0.7517 0.9678 0.8643 0.8 0.9834 0.8106 0.7 0.8389 0.0 2.9974 0.9406 0.9961 0.04 0.8051 0.6844 0.9441 0.4 1.9929 0.9990 Table 8.6368 0.9671 0.6554 0.9251 0.9803 0.8159 0.9842 0.6 2.9783 0.8790 0.6628 0.7422 0.5714 0.7939 0.8461 0.9931 0.8849 0.9982 0.9936 0.9099 0.9778 0.9960 0.9970 0.9649 0.7389 0.7054 0.7794 0.6103 0.8508 0.9920 0.6064 0.4 2.9177 0.9948 0.5359 0.8980 0.9192 0.9986 0.9975 0.8810 0.6293 0.9641 0.9861 0.9927 0.9963 0.9887 0.5 1.5040 0.9015 0.9988 0.1 1.9147 0.9943 0.9808 0.8907 0.9 1.7611 0.09 0.9909 0.9821 0.7881 0.9896 0.9988 0.9978 0.9941 0.5948 0.7257 0.8665 0.7 2.9744 0.6591 0.5753 0.7704 0.9812 0.9946 0.3 0.6700 0.9525 0.9890 0.9901 0.9738 0.7642 0.