Scribd
Upload
16 views24 pages

Normal Distribution Notes

The document discusses the properties and applications of the normal distribution, a key concept in statistics characterized by its bell-shaped curve and symmetry. It explains how the normal distribution can be used to model various natural phenomena and describes methods for calculating probabilities and z-scores using tables and calculators. Additionally, it provides examples of how to apply the normal distribution in real-world scenarios, such as measuring life expectancy and weights in packaging.

Uploaded by

jsminee803
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
16 views24 pages

Normal Distribution Notes

The document discusses the properties and applications of the normal distribution, a key concept in statistics characterized by its bell-shaped curve and symmetry. It explains how the normal distribution can be used to model various natural phenomena and describes methods for calculating probabilities and z-scores using tables and calculators. Additionally, it provides examples of how to apply the normal distribution in real-world scenarios, such as measuring life expectancy and weights in packaging.

Uploaded by

jsminee803
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
The Normal Distribution am NM Ne RMR NI IMAG MRL RTI A + PROPERTIES, + USING TABLES. -> AMODEL FOR MEASUREMENT > COMMONLY USED PROBABILITIES AND Z-SCORES. “9 FINDING MEANS AND STANDARD DEVIATIONS. + NORMAL APPROXIMATION TO THE BINOMIAL. «> CHECKING NORMALITY s™ of the most useful probability distributions are symmetric and bell-shaped: Sr In this topie we study one such distribution called the normal distribution (not all sym- ‘metric, unimodal distributions are normal) The normal distribution is valuable in describing various natural phenomena, especially those involving growth or decay. However, the real importance of the normal distribution in statistics is that it can be used to describe the results of sampling procedures, | ‘The normal distribution can be viewed asa limiting case of the binomial, More specifically, we start with any fixed probability of success p and consider what happens as n becomes arbitrarily large. To obtain a visual representation of the limit, we draw histograms increasingly large n for each fied p. Histograms for p = 0.5 ond fr n= 4 10, and 25, respectively ‘THE NORMAL DISTRIBUTION 255 ogra aera or p= 07 ond for n= 4,10, ond 25, respec For greater claity, different scales have been used for the histograms associated with dif- ferent values of n. From the diagrams above itis reasonable to accept that as n becomes larger without bound, the resulting histograms approach a smooth bell-shaped curve. This is the curve associated with the normal distribution. PROPERTIES OF THE NORMAL DISTRIBUTION ‘The normal distribution curve is bell-shaped and symmeitic and has an infinite base, Long, flat-looking tails cover many values but only a small proportion of the area. The flat appeat ance ofthe tails is deceptive. Actually, far out in the tails, the curve drops proportionately at an ever-increasing rate. In other words, when two intervals of equal length are compared, the one closer to the center may experience a greater numerical drop, but the one further out in the tal ‘experiences a greater drop when measured as a proportion of the height at the beginning of the interval ‘The mean here is the same as the median and is located at the center. We want a unit of ‘measurement that applies equally well to any normal distribution, and we choose a unit that arises naturally out of the curve's shape, There is a point on each side where the slope is steep- est, These two points are called points of inflection, and the distance froma the mean to ei point is precisely equal to one standard deviation. Thus, itis com under the normal curve in terms of z-scores (recall from Topic 2 that ‘multiples of standard cleviations from the mean}. Point of inflction —>, ceo Point of inflection 284 Ap sTATISYICS a AL eee ee err eH ik rd doatos von cintdtocnmce ‘Mathematically, the formula fol the normal curve turns out to be y= e-®”2, where yis the relative height above z-score. (By relative heigh we mean proportion ofthe height ‘mean. However, our interest isnot so much in relative heights under the notmal curve as in proportionate areas. The probability associated with any interval under the normal curve is equal tothe proportionate area found under the curve andl above the interval [Link] property ofthe normal curve is that approximately 68% of the area (and thus 68% ofthe observations) falls within one standard deviation of the mean, approximately 5% ofthe area falls within two standard deviations ofthe mean, and approximately 9.7% of the area falls within three standard deviations of the mean, LE LON BOX wenn owe nw r2o pa eae yave the USING TABLES OF THE NORMAL DISTRIBUTION ‘Table A in the Appendix shows areas under the normal curve, Specifically, this table shows the area to the left ofa given point. It shows, for example, that to the left of a z-score of 1.2 there 180.8849 of the area: 12 score Because the total area is 1, the area to the tight ofa point ean be calculated by calculating 1 ‘minus the area to the left of the point. For example, the atea to the right of a z-score af ~ 2.13, is 10,0166 = 0.9834. (Alternatively, by symmetry, we could have said that the area to the right 0f-2.13 is the samo as that to the left of +2.13 and then read the answer directly from Table A.) ‘The notation NGuo) stands for anormal ‘model with ‘mean and standard deviation «. ‘The area between two points can be found by subtraction. For example, the area between the z-scores of 1.23 and 2.71 is equal to 0.9966 - 0,8907 = 0.1059, 123 271 xscor0 Given a probability, we can use Table A in reverse to find the appropriate z-score. For example, to find the z-score that has an area of 0.8982 t the left oft, we search for the prob- ability closest 100.8962 in the bulk of the table and read off the corresponding z-score. n this cease we find 0,8980 with the z-score 1.27. 127 z-score ‘To find the z-score with an area of 0.72 to the right ofit, we look for 1-0.72 = 0.28in the bulk ofthe table, The probability 0.2810 corresponds tothe z-score -0.58, USING A CALCULATOR WITH AREAS UND! ANORMAL CURVE On the TI-84, normaledt{lowerbound, upperbound) gives the area (probability) between two z-scores, while inviora(aree) gives the £-seore with the given area to its let. So we could have calculated above: normaledé(-100, 1.2) = 0.6859 normaledf(-2.13, 100) = 0.9634 normaledé(I,23, 2.71) = 0.1080 AnvNorm(.99 an invilorm(.28) = -0.5828 ‘The TI-84 al has the capability of working directly with raw scotes instea® of z-scores. In this case, the mean and standard deviation must be given: normalcdf{lowerbound, upperbound, mean, SD) invWorm(area, mean, SD) 256 Ap sTaTistics a e ee. THE NORMAL DISTRIBUTION AS A MODEL, FOR MEASUREMENT Insolvinga problem using the normal distribution asa model, drawing a picture of a normal ‘curve with horizontal lines showing raw scores and z-scores is usually helpful. ‘> EXAMPLE 111 The life expectancy of a particular brand of lightbulb Is normally distributed with a mean ‘f 1500 hours and a standard deviation of 75 hours, a ee a ais 135014280 —aTS SDS Hours 9 24 ° 7 2 3 rscord What isthe probability that a lightbulb will last loss than 1410 hours? Anower: the sseot of Is IOEHEOD 1.2, and fom ‘Table A tha probability to the loft of =12 Is 0.115I. [On the T184, ‘normaled(0, 1410, 1500, 75) = 0.1151 and normatcal(-10, -12) * 0.11511 Segarra Oeste (continued) ‘THE NORMAL DISTRIBUTION oe: Dont use the normal model it ‘he distribution Isn't symmetric ‘and unimodal. ot Draw a pieturel 287 'b, What is the probability that a lightbulb wil last between 1563 and 1648 hours? Answer: The z-score of 1563 is $8853809_ = 0,84, and the z-score of 1648, is 16481500 5 197, In Table A, 0.84 and 1.97 give probabilties of 0.7995 and a probability of 1762 and 0.9756, respectively. Thus between 1563 and 1648 there 0.9756 ~ 0.7995 = 0.1761, [normaledf(ts63, 1648, 1500, 75) normaledt(.84, 1.97) = 0.1760.) \ \ W800 1863 (eas Hous tt oom 197 score What is the probability that a lightbulb wil last between 1416 and 1677 hours? Answer: the z-score of 146 is 183800 = 12 andthe z-score of 1677 is 197753800 256, n Table A, 12 ges a proba aves a probobility of 0.9909. Thus betweon 116 and 167 there ea probablity of 0.9909 - 01816 = 0.8595 norraes146, 177, 100, 75) = 0.8506 and ormaledif-112, 2.36) = 0.8595.) lity of 0.1314, and 236 258 ap staristics ‘> EXAMPLE 11.2 ‘Appacking machine Is set to fil a cardboard box with a mean of 16.1 ounces of cereal. Suppose the amounts per box form a normal distribution with a standard deviation ‘equal to 0.04 ounce. V \ 1598 1602 18061610 tata 1eT6 —_To2r Ounces nn oa rt ad ec Terrneer weieeee: eeinaa v 2 3 rscore 2, What percentage of the boxes will end up with at least 1 pound of coreal? Answers The zacore of 6 188.1 = -2.5, nc -25 in Table Acives 0 probability of 0.0062. Thus the probability of having more than 1 pound in a box Is 1 - 0.0062 = 0.9938 oF 99.38%, [normaledf(16, 1000, 16.10, .04) = 0.9928 and normaledf(-2.8, 10) = 0.9938.) b. Ten percent of the boxes will contain less than what number of cunces? Answer: tn Table A, we note that . area (actually 0.1003) is found to the left of 8 -1.28 z-score, (On the TI-84, invNorm(1) = -1.2816.] Converting the z-score of (oF 16.1 = 1.28(0.04) = 16.049 16.049.) x= 161 ‘0.04 ‘ounces. [Or directly on the 11-84, InvNormC!, 16.10, .04) 128 Into a raw score yields -1.28 Ounces (continued) THE HORMAL DISTRIBUTION 259 Eighty percent of the boxes will contain more than what number of ounces? Answer: In Table A, we note that 1 0.8 = 0.2 area (actually 0.2005) is found to the left of a z-score of -0.84, and so to the right of a -0.84 z-score must bo 180% of the area, [Or invNorm2) = 0.8416.} Converting the z-score of -0.84 into ‘a raw score yiolds 16:1 ~ 0.84(0.04) = 16.066 ounces. [Or directly on the TI-84, InvNorm¢2, 16.10, .04) = 16.066.) Tes 1610 Ouices oy vat} z-score The middle 90% of the boxes will be between what two weights? Answer: Ninety percent in the middle leaves five percont in each t ‘A, we note that .0505 area is found to the left ofa z-score of ~1.64, while 0.495 ‘area is found to the left of a z-score of ~L65, and so we use the z-score of 1.645, Then 90% of the area Is betwoen z-scores of ~1645 and 1.645. {Or InvNorm(05) = 1.6449 and invNorm(.95) = 1.6449,] Converting z-scores to raw scores yiolds 16.1 - 1.645(0.04) = 16.034 ounces and 16.1 + 1.645(0.04) = 16.166 ‘ounces, respectively, for the two weights between which we will find the middle 90% of the boxes. [lnvNorm(.05, 16.10, .04) = 16.034 and invNorm(.95, 16.10, 04) = 16.166.1 COMMONLY USED PROBABILITIES AND Z-SCORES ‘As can be seen fiom Example 11.24, there is often an interest in the limits enclosing some specified middle percentage of the data. For future reference, the limits most frequently asked for are noted below in terms of z-score 0 “W645 1.645 290° AP STATISTICS Ninety percent of the values are between z-scores of ~1,645 and 41,645, 95% of the values are between z-scores of 1.96 and +1.96, and 99% of the valu and 42.576, Sometimes the interest isn values with particular percentile rankings. For example, 1282 1.645 2.336 are between z-scores of -2.576 Ninety percent of the values ateWbelow a z-score of 1.282, 95% of the values are below z-score of L645, and 99% of the values are below a z-score of 2.326. ‘There ate cotresponding conclusions for negative z-scores: fi. “1282 2306 Ninety percent of the values are above a z-score of ~1,282, 95% of the values are above a z-score of -1.645, and 99% of the values are above a z-score of -2.826. It is also useful to note the percentages corresponding to values falling between integer z-scores. For example, Note that 6i.26% of the values are between z-scores of -1 and +1, 95.44% of the values are between z-scores of -2 and +2, and 99.74% of the values are between z-scores of -3 and +3. ‘> EXAMPLE 11.3 Suppose that the average height of adult mates in a particular locality is 70 inches with a standard deviation of 25 inches. & IF the distribution is normal, the middle 95% of mates are between what two heights? Answer: As noted above, the critical 2-scores in this case aro #.96, and so. {the two limiting heights aro 1.96 standard deviations from the mean. Therefore, 70 #1.96(2.5) = 70 44.9, or from 65.1 to 74.9 inches. . Ninety percent of the heights are below what value? Answer: The critical z-score is 1.282, and so the value in question is 70 + 1.282(2.5) = 73.208 inches. Ninety-nine porcent of the heights ate above what value? Answer: The critical z-score is -2.326, and so the value in question is 70 ~ 2.326(2.5) = 64.185 Inches. . What percontage of the heights are between z-scores of 41? Of 42? Of 437 Answer: 68.26%, 95.44%, and 99.74%, respectively. “TH NONMAL DISTRIBUTION 261 FINDING MEANS AND STANDARD DEVIATIONS Ifwe know that a distribution is normal, we can calculate the mean [Land the standard devia- tion 6 using percentage information from the population, ‘> EXAMPLE 11.4 Given a normal distribution with a mean of 25, what isthe standard deviation if 18% of the values are above 297 Answer: \Looking for a 0.82 probability in Table A, we note that the conlesponding z-score is 0.92. Thus 29 ~ 25 = 4 is cqual to a standard deviation of 0.92, that is, = eas 0.926 = 4, and o= GS5 = 4.35. Raw score ‘w> EXAMPLE 11.5 Given # normal distribution with @ standard deviation of 10, what is the moan It 21% of the values are below 50? Table A leads us to a z-score of -0.01. Thus 80 + 0.81(10) = $8.1 Ansiver: Looking for 80.21 probability 50 is -0.81 standard deviation from the mean, and so, Row score 62 AP STATISTICS ‘> EXAMPLE 11.6 = SEES EEE aes Given a nortnal distribution with 60% of th values above 125 and 90% of the values above 10, what are the mean and standard deviation? Answer: Table A givos critical z-scores of 0,94 and -120, Thus we have PSOE 2-084 and MO=IK = 128, Solving the system (125 - = -0.840, i TO - 1 # 1.280) simultaneously gives 1 = 153.64 and = 34.09, \ NORMAL APPROXIMATION TO THE BINOMIAL. Many practical applications of the binomial involve examples in which nis large. However, | 2D for large n, binomial probabilities ean be quite tedious to calelat, Since te norma can be | When nis sal viowed asa limiting case of the binomial itis natural o use the normal to approximate the | YOM Sanot binomial in appropriate situations, toda to v tinuous with probabili- | approximate the The binomial takes valuos only at integers, while the normal is continuous with probabil: | Treat modal, ties corresponding to areas over intervals." Therefore, we establish a technique for converting fiom one distribution to the other. For approximation purposes we do as follows. Each bino- nial probability corresponds to the normal probabillty over a unit interval centered atthe desired vatue, Thus, for example, to approximate the binomial probability of eight successes ‘we determine the normal probability of being between 7.5 and 8.5, ‘=> EXAMPLE 11. ee eee eet Suppose that 18% of the cars coming out of an assembly plant have some defect. Ina dolivery of 40 cars what Is tho probability that exactly 5 cars have defects? Ansiter: The actual answors (349%; c015)%0.85), but cloths involves ‘a nontrivial calculation. [if one has a calculator such as the TH84, then bInompd40, 0.5, 5) = 0.1692, To approximate the answer using the normal, ‘we frst calculate the mean j and the standard deviation as follows: = np 2 40035) = 6 = (aati) = /ROTEABI(ORE) = 2280 ‘ continued) is THE NORMAL DISTRIBUTION 263 35 We then calculate the appropriate zscores: Sa = -0.66,and Soep = 22, Looking up the corresponding probabilities in Table A, we obtain a final answer of 0.4129 ~ 0.2546 = 0.1583. (The actual answar is 0.1692) we Cors with defect penne tis at score ven more useful are approximations relating to probabilities over intervals, ae kt) EXAMPLE 11.8 _ peblesiztsessicuecteceeceereeeeeeeeereereree f 60% of the population support massive Federal budget cuts, what Is the probability that in a survey of 250 people at most 155 people support such cuts? Answer: The actual answer Is the sum of 156 binomial expressions: (0478 +--+ (2824 (05)"%04" Howaver, a good approximation can be obtained quickly and easily by using the normal, We calculate j and o: = (mop) = (RO(OHOA) « 7.746 ; _ ‘The binomial of at most 155 successes corresponds to the normal probability of s155.5. 3855-150 77a of 0.7611. [0r binomedt(250, The z-score of 155.5 Is 0.71. Using Table A leads us to a final answer 185; 7605.1 Is the normal a good approximation? The answer, of course, depends on the error toler ‘ances in particular situations. A general rule of thumb is that the normals a good approxima- tion to the binomial whenever both np and n(1 ~ p) are greater than 10, (Some authors use 5 instead of 10 here.) CHECKING NORMALITY In the examples in this Topic we have assumed the population has a normal distribution. When you collect your own data, before you can apply the procedures developed above, you ‘must decide whether itis reasonable to assume the date come from @ normal population. In Jatet Topics we will have to check. for normality before applying certain inference procedures. ‘The initial check should be to draw a picture. Dotplots, stemplots, boxplots, and histo- {grams are all useful graphical displays to show that data is unimodal and roughly symmetric. > EXAMPLE 1.9 ‘The ages at inauguration of U.S. presidents from Washington to G. W. Bush wero: (87, 61, 57, 57, 58, 57, 61, $4, 68, SI, 49, 64, 50, 48, 65, 52, 56, 46, 54, 49, 51, 47, 55, 55, 54, 42, 51, 56, 55, 5I, 54, 51, 60, G1, 43, 55, 56, 61, 52, 69, 64, 46, 54). Can we conclude that the distribution Is roughly normal? Answer: Entering the 43 data points in a graphing calculator glves the following histogram: (continued) THE NORMAL DISTRIBUTION 265 Alternatively, we could have used a dotplot, stemplot, or boxplot: 23 667899 orrii12944444 ——Cb——+ 355586677778) mmengermengsererg orniiad Pearce 589 All the graphical displays indicate a distribution that is roughly unimodal and yr \-——— aa fe Amore specialized graphical display to check normality is the normal probability plot When. the data distibution is oughly norm, the plot is roughly a diagonal straight line. While chis plot more clearly shows deviations from normality, it is not as easy to understand asa histo- gram, The normal probability plot is difficult to calculate by hand; however, technology such as the 11-04 readily plots the graph. (On the'TI-B4, in STATPLOT the sith choice in TYPE gives tie normal probability plot) The normal probability plot for the data in Example 11.91 70 _° 6 : B 60 rat B ss # so, 45 : aot. a ttt BAS -1-5 0 5 1 15 2 Normal scores graph this close to a diagonal straight line indicates thatthe data have a distribution very loge to normal. Note that alternatively, one could have plotted the data (in this cas, inaugural ages) on the -eais and the normal scores on the yeas SUMMARY (0 w The notmal distribution is one particular symmetic, unimodal distribution. 2 Drawing pictures of normal curves with horizontal lines showing raw scores and z-scores is usually helpful | _& Areas (probabilities) under a normal curve can be fou using the given table orcaleu- lator fanctions, such as noxmaledf on TI calculators «Critical values corresponding to given probabilities can be found using the given table or calculator functions, such as invNorm on TLealculator. 12 The normal probability plot is very useful in gauging whether a given distibution of data is approximately normal: I the plot isneatly straight, the datas nearly normal. 203 AP STATISTICS QUESTIONS ON TOPIC ELEVEN: THE NORMAL DISTRIBUTION’: Multiple-Choice Questions Directions: The questions or Incomplete statements that follow are cach followed by five suggested answers or completions. Choose the response that best answers the a ‘uostion or complates the statement. \ 1. Which of the following is a true statement? (A) The area under a normal curve is always equal to 1, no matter what the mean and standard deviation are. (B) All bell-shaped curves are normal distributions for some choice of [L and 6, (C) The sinaller the standard deviation of a normal curve, the lower and more spread out the graph. (D) Depending upon the value of the standard deviation, normal curves with | different means may be centered around the same number ©) Depending upon the value ofthe standard deviation, the mean and median ofa} particular normal distribution may be different. | 2. Which of the following sa true statement? (4). The area under the standard normal curve between @ and 2s twice the area between O and 1. (©) Theatea under the standard normal curve between 0 and2 is half the wrea between ~2 and 2. (© Forthe standard normal curve the interquartile range is approximately 3. () Forthe standard normal curve, the range is 6. (2) For the standard normal curve, the azea tothe left of0.1 isthe same asthe area co the right of a9, 3. Populations P1 and P2 are normally distributed and have identical means, However, the standard deviation of P1 is twice the standard deviation of P2. What can be said about the percentage of observations falling within two standard deviations of the mean for each population? (A) The percentage forP1 is twice the percentage for P2, (@) The percentage for PL is greater, but nat twice as great, as the percentage for P2. (C) The percentage for P2 is twiee the percentage for PL (D) The percentage for P2is greater, but not twice as great, asthe percentage for PI ©) The percentages are identical. . ’ ‘THE NORHAL DISTRIBUTION 267 - 4, Consider the following two normal curves: Which hi i the larger mean and which has the larger standard deviation? (A) Larger moan, a; larger standard deviation, a (B) Larger mean, a; larger standard deviation, b (C)_ Larger mean, b; larger standard deviation, a (D) Larger mean, b; larger standard deviation, b (©) Larger mean, b; same standard deviation In Questions 5-15, assume the given distributions are normal. 5. Atrucking fitm determines that its fleet of trucks averages a mean of 12.4 miles per gallon with a standard deviation of 1.2 miles per gallon on cross-country hauls. ‘What is the probability that one of the trucks averages fewer than 10 miles per gallon? (a) 0.0082 (B) 00228 (© oar (©) 05228 wy 09772 6. A factory dumps an average of 2.43 tons of pollutants into a river every week. Ir the standard deviation is 0.08 tons, what is the probability that in a week more than 3 tons are dumped? (A) 0.2578 @) 0.2043 (© 06500 ©) 07157 @) 0.7422 260 AP STATISTICS 7. Am electronic product takes an average of 3.4 hours to move through an assembly line. If the standard deviation i 0.5 hour, what is the probability that an item will take between 3 and 4 hours? (a) 02119 (8) 0.2295 (© 0.3270 i (D) 0.3811 (B) 0.6730 \ \ 8, The mean score on a college placement exam is 500 with a standard deviation of 100. Ninety-five percent of the test takers score above what? (a) 260 (B) 336 © 405 (D) 414 (8) 664 9. The average noise level in a restaurant is 30 decibels with a standard deviation of 4 decibels, Ninety-nine percent of the time itis below what value? (a) 207 (B) 320 | (©) 334 () 37.8 @) 393 10. ‘The mean income per houschold in a certain state is $9500 with a standard deviation 0f$1750, The middle'95% of incomes are between what two values? (A) $5422 and $13,578 (8) $6070 and $12,930 (©) $6621 and $12,379 (D} $7260 and $11,740 (®) $8049 and $10,951 11. One company produces movie trallers with mean 150 seconds and standard devia- tion 40 seconds, while a second company produces trailers with mean 120 seconds and standard deviation 30 seconds, What is the probability that two randomly selected trailers, one produced by each company, will combine to less than three minutes? (a) 0.000 (B) 0.036 yO) 0.099 ‘ Y {D) 0.180 — (0.405 ‘THE HORMAL DISTRIBUTION 269 ae: 12, Jay Olshansky from the University of Chicago was quoted in Chance News as aguing that for the average life expectancy to reach 100, 16% of people would have to live to 120. What standard deviation is he assuming for this statement to make sense? (a) 217 B) 44 (© 22 D) 350 © ma \ \ 13, Cucumbers grown on a certain farm have weights with a standard deviation of 2 ‘ounces. What is the mean weight if 5% ofthe cucumbers weigh less than 16 ouncest (A) 13.92 (By 1430 © 1440 (D) 1488 (@) 18.70 14, 175% of all families spend more than $75 weekly for food, while 15% spend more than $150, what is the mean weekly expenditure and what isthe standard deviation? 15, Accoffee machine can be adjusted to deliver any fixed numbor of ounces of coffee. IF | the machine has a standard deviation in delivery equal to 0.4 ounce, what should be ‘the mean setting so that an 8-ounce cup will overflow only 0.5% of the time? (A) 6.97 ounces (B) 7.22 ounces (©) 7.34 ounces {D) 7.80 ounces 9.03 ounces 16, Assume that a baseball team has an average pitcher, that is, one whose probability of winning any decision is 0.5. If this pitcher has 30 decisions in a season, what is the probability that he will win at least 20 games? (a) 0.0505 (@) 02514 © o2raa x y (D) 0.3333, — (B) 0.4300 270 Ap. STATISTICS Bee 17. Given that 10% of the nails made using a cert manufacturing process have a length less than 2.48 inches, while 5% have a length greater than 2.54 inches, what are the ‘mean and standard deviation ofthe lengths of the nails? Assume that the lengths have ‘a normal distribution, (©The nbban and standard deviation cannot be computed from the infSkmation given, Froe-Response Questions Directions: You must show all work and indicate the methods you use. You will be 0.3) is fess than 05. 3. (€) All normal distributions have about 95% of their observations within two standard deviations of the mean, 4, (C) Curve ahas mean 6 and standard deviation 2, while curve bhas mean 18 and stan- dard deviation 1. 1-124 5.,(B) The z-score of 10 is From Table A, to the left of -2 is an atea of 0.02 8, (normalcd(0, 10, 124, 1.2) 3 6. (A) The zscore of 3 is = 0.65, From Table A, to the left of 0.65 is an area of 0.08 0.7422, and so to the right must be 1 0.7422 = 0.2578, (aormaledl(3, 1000, 2.43, 88) = 0.2589, 7. (8) The z-scores of 8 and 4 are 534 2-08 and 4534 ~ 1.2, respectively. From ‘Table A, to the left of 0.8 is an area of 0.2119, and to the left of 1.2 is an area of 0.8849, Thus between 3 and 4 is an area of 0.8849 ~ 0.2119 = 0.6730. {normaledf(3, 4, 34, 5) = 067314 8. ¢B) 195% of the area is to the right of a score, 5% is to the left. Looking for 0.05 in the body of Table A, we note the z-score of -1,645. Converting this to a raw score gives 500 ~ 1.645(100) = 336, linvNorm(.065, 500, 100) = 336.) 9. (G) The critical z-score associated with 99% to the lefts 2.326, and 30 + 2.826(4) = 39.3. linvNorm(.89, 20, 4) = 39.3) 10. (B) The critical z-scores associated with the middle 95% are 1.96, and 9500 1.96(1750) = 6070 and 12,930. {invNorm(.025, 9500, 1750) = 6070 and invNorm(.975, 9500, 1750) = 12,930) 160-270 LL. (BY fggy= 150 120270 and ayy YAOI =50. then P(2< A222) 9.036, 274 Ap SrANISTICS 12, (A) The critical z-score associated with 18% to the right oF 82% to theleft is 0.92. Then. 100 + 0.926 = 120 gives = 21.7. 13, (A) The critical z-score associated with 85% to the left is 1.04, Then st + LO4(2) = 16 aives w= 13.92. 14, (D) The critical zscores associated with 75% to the right (25% to the left) and with 15% to the tight (85% to the left) are -0.67 and 1,04, respectively. Then {u ~ 0.670 = 75, #14 1.040 = 150) gives = 104.99 and 0 = 43.86 15. (A) The critical z-scdke associated with 0.5% to the right (99.5% to the left) is 2.678. Then ¢+2.576(0.4) = 8 gives c= 697. 16, (A) Using the normal as an approximation to the binomial, we have 1 = 30(0.5) = 15, 0° = 30(0.5)(0.5) = 7. = 164, and 1 ~ 0.9495 = 0.0505, (Or Dinomedf(30, .5, 10) = 0.0494.) 17, (A) The respectively. Then (4t~ 1.2820 = 2.48, j.+ 1.6450 = tical z-scores for 10% to the left and 5% to the right are “1.282 and 1.645, 54) gives x= 2.506 and o = 0.0206. FRUE-RUSPONSE 1. (@) Let the random variable D be the difference in walking times (Steve-Jan). Then p= 30-25 =5and oy= \S%+4* = 6.403, For Steve to arive before Jan, the 2z4n,a000) | 2. (a) P(2740)2500)=—— Given independence, the probability that both components last 240 more hours is (0.829825) (0.665506) = 0.552. (b). The probability that both fail is (1 ~0.829825)(1 ~ 0.665506) = 0.056928. The probability that at least one doesn't fail is. 1 - 0.056923 = 0.943. v INVESTIGATIVE TASKS: L. (@) The announcer meant that to maintain his tabulated free throw percentage, Jordan would have to soon miss one. This is not a correct use of probability. If Jordan ‘makes a certain percentage of free throws, that probability applies to each throw THE NORMAL DISTRIBUTION 275 imespective of the previous throws, By the law of large numbers, in the long run the relative frequency tends toward the corect probability, but no conclusion is pos- sible about any given outcome, (b) Pisix ina row) = (0.9) Pllive, then a miss) P(nextlprevious five) = P(next (0.9)*(0.1) = 0.059 9 © et the digits 1 through @ stand for making a free thro, while 0 stands for a miss. {Look at blocks of five random numbers. fal stand for "makes" (no 0), look at the \ very next ligt to see ifit isa 0 or not, Keop a tally of how maby times five makes in row are followed by a make and how many times five makes ina row ace followed byamiss (@) This is a binomial distribution with n = 6 and p np = 69) =5.4 while the standard dovition is Ynp{l-P) (©) Since n= 40(1 ~0.90) = 4, the normal approximation tothe binomial is not recom- ‘mended, However, a direct binomial calculation yields P(S38)=1- P(236) - [(Zjosoar ' (Boar. ay +(e (0. Fa( ye 9)(0.1)+(0 | = 037 il so the mean is, fa(0.(0.1) = 0.735, 1 (Or simply use binomedf(40, 9, 35) on the TI-84,) With such a high probability, there is no evidence to conclude that Jordan's average has dropped! 2 (a) With z= 185, Table A gives a probability of 0.1977 =0.2 {) For exemple, letting 0 and 1 represent caplets containing less than 8500 units and 2 through 9 represent caplets containing more than 8500 units, we can read off digits until we Sind two containing less than 8500 caplets, Five simulations would yield 841770 67571761 31558251 50681 435410 giving 6,8, 5, and 6 tablets sampled before finding two with less than 8500 units apiece, ©) ‘he probability of « caplet having more than 9900 units is 0.5, and thus one would ‘expect to find two caplets with more than 9000 units much quicker than finding two caplets with less than 8500 units. Thus the first histogram results from looking for ‘wo caplets with more than $000 units and the second histogram results from look- ing for two caplets with less than 8500 units. xi Anmestimate forthe expected value is obtained by YieP(x) = 20.03) + 8(0.05) + 4(0.09) +5(0.10) +--+ 24(0.01)=9.44, 276 AP STATIGNICS

You might also like