Scribd
Upload
6 views26 pages

Statistics Notes

The document provides an introduction to statistics, covering key topics such as data representation, probability theory, and the normal distribution. It explains how to organize data using frequency tables and histograms, as well as how to calculate measures of central tendency like mode, mean, and median. Additionally, it discusses the fundamentals of probability, including definitions, tree diagrams, and set theory, along with practical examples and exercises.

Uploaded by

orabetswe1
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
6 views26 pages

Statistics Notes

The document provides an introduction to statistics, covering key topics such as data representation, probability theory, and the normal distribution. It explains how to organize data using frequency tables and histograms, as well as how to calculate measures of central tendency like mode, mean, and median. Additionally, it discusses the fundamentals of probability, including definitions, tree diagrams, and set theory, along with practical examples and exercises.

Uploaded by

orabetswe1
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
INTRODUCTION TO STATISTICS Inthis pr ofthe course we attempt o give you basis for the satis you wil need further on in you studies, ‘We inroduce three topics 1. The Representation and Handling of Data, 2. Probablty Theory 3. The Normal Distebuion. 1. The Representation and Handling of Data For this entire section we will we the following data sillusraton: The ages of rent diferent people: 28,24, 13,24, 28, 16,20, 16, 3, 22, 23, 20,22, 24,24, 28,24, 41,26, 4, 1.4, Frequency Tables ‘We need to be abl to orgaise data before we can draw any oochisions from it. For his ‘we will need what i called feguency able. Say, for example, you are given the ages of twenty different people: 28,24, 13, 24,28, 16,20, 16,23, 22,23, 20, 2,24, 24,28, 24,41, 24,41. inthis format, data is prety wales. Let's create abt of order. We draw vp a table of Frequencies. From his table we can see exactly how many mes each cbservation coeur Sometimes iis eore wef to divide the obsecrations into classes. We can, fr example, consider the ages fom 10 to 1, fora 20 to 29, from 30 to 39 and fom 40 fo 49. In this ase, the fegency able wil lok ike this: Gass] — Goan] Freaueney To, o x 229 ws 3030. = o suo 7, Tae ae 112. Histograms Daa can also be represented graphically. Thi lps us to visualise the big picture” at fone glance, One of tie moat popular ways todo this isto use a bar chert of histogram. ‘We ue the x-axis fer the observations or clases andthe y-axis for frequencies, For the previous me equency table, respectively, te histograms wi ook s follows: HISTOGRAM HISTOGRAM 1019 2029-3039 40-49 ‘Age Groups 13, Relative Frequency Saying that e speci observation has a fequency of Sis't very usefil in itself. This could be considered seltvely large frequency ifenly 10 observations were made. I ‘however, 10000 000 observation Were mado, afequency of i negligible. To, ‘overcome this problem We somtimes prefer to use relative frequency. [Relative fequency = Frequency ofthe specific obsennton/dast l Tota numberof oboarvations ‘Tey to vey for youre that the relative hequencies of ha respective age in our example ae: ‘Age:13 Relative fequeny: 0.05 ‘Age: 16 Relative fequency: 0 ‘Age:20 Relative fequeny: 01 ‘Age:22 Relative eqoeney:0.1 Ags 23—Relave Roquney 01 Age 2A Relative oqueney 03 ‘Age28 Relative foquney 0.15 ‘Aged Relaive Soqueney 01 [Note that al the relative frequencies add up to 1. This is always the case, “Im sections 14 and 15. we shall se the following notation: 1 Jorthe mumberposition of n observation x, Jorthe ith observation 1 Forthe total number of observations 114, Mode, Mean and Median {tis ofen useful to represen an entre set of data witha single valu that gives us an idea ‘ofthe general vale of the observations. Fo this purpose we have the mode, mean and median ofa dataset. "The mode is simply the value that occurs the most times (Le the value withthe highest fequency.) Te shoul be clear to you that in ou case, the mode ofthe twenty ages is 24 “The mean is basically the average ofthe observations. To calculate the mean dp all he observations and divide by the numberof observations. ‘The mean ofthe twenty ages is therefore: (03% + 016%. + (20x29 + (222) +232) + 24x6)+ (2823) + (4142) 20 “Thc median isthe val hats found inthe mile of he dataset, when observations ace arranged fom smallest to legest. ‘Finding the median isa simple as ranging the differect observations from smallest to Target and taking the valu inthe middle If, however, we have an evea number of| observations we ake the average ofthe two values in the middle. So forthe twenty ages: Step 1: Arrange: 13,16 16;2020,22,22,23,23 2424,2424,2424.26.28,28,1,81 ‘Step 2: Because there are two valet nthe middle, we tke the average: median = 24+24 4 1S. Standard Deviation ‘Sometimes we need to have an idea of ovr the different values are spead sround the mean The most power Satistic we have for this purpose i the standard deviatlon. ‘We calulate the standard devistion by subtracting the mean from every value and squaring tis diference (one by one). We ten add up the answer, divide the sum by the number of observations end nally, take the square rot ofthis eau. ‘This process is mathematically denote by the formule: [SG -3 W ‘There are functions on scent calculators thet enable uo find the sendard deviation, ‘but we wil do so manually. The best way of doing this is by drawing up a table. We will ‘agin uso the twenty ages as an example: a |e -8 | -3) 23_|a75_| 140008 24 | -a25—| os, is [ize sas za [a8 | 0.025. 2a | 376 | 140005) 36 | 625 [0.005 eas | 195 | 1seas 30_| 22 | -295-| Sones a1 [ as [125 | tes x2 [20 | 425 ya0eos as_[ 22 [25 | 60025 14 [ 24 | a6 |. x6 [24 | 025 | os 6a [376 | ra080s 73936 Poms. 1a | 4 | 1675 [o0n 628 19 [ae [025 | oes (20 | “a1 | 1675 ann se itis tat $65, -3)°= 91575. Nowe EXERCISES 4 (1) 1. 111962, the outstanding debt per capita, forthe diferent US states were as follows: u ~ 1.1-Draw up a frequency table with classes (0,00), 00,1000) 1000,1500), ua $200, om2sn) 2500), 15; Wie ithe eae tequeny of he clas (50,1000? a 9 teen nye nent, re it Laierme ste BIE a8 Ble S/S aly PRE B/E S/S ORT RE OB 2.1, Draw up a frequency table with lasses: (15,20), (2025) ,[25,30). 122. Represent these clases on histogram, 2.3, What percentage of eachors earned $25 000 or more? 2. Complete the fllowing table: foooa] me] ieee Eee] ee], 4. Complete the following table and then: 4.1 draw a histogram frepresent the dat. ‘42, find the mode, mem and median, 2. Probability Theory 2A, Definition of Probability Probability i measure of how likly something isto happen. Asking “what faction of seadensbave brown eyes” isthe sae as asking “whats the probably of randomly selecting student wit brown ees?” Thus, probaly wil aways be a mumber between and I. A probability of 0 means tha something is impossible, while «probability of 1 means tht wll deity happen. Some examples of probable in everyday ie ar: There isa 60% (0.6 or 3) chance fr rain tomorow. fou fp a coin there isa pebebity of 50% (0.5 o£) that wil land on heads ‘Ifyou guess the answer toa multiple choice question wih S posible answer, the probability that you will gti rights } (02 or 20%), ‘Probsbility of an event happening = Smberof waysin which this specific eventcan happen ‘otal numberof possibiies ‘Let us apply this formula to practical problems Beample Ina class there are 10 men and 15 women If one person ie sclected randomly (ll ere ‘equally ely tbe chosen), what is the probability that wil be a woman? number of women in the class Tota number of stents inthe clas 1s 35 05 Probability of selecting wom Another example ‘When you roll ade (singular for ice), what ithe probability that it wil land on six? numberof ways torollasi Probability of rlingesix = ‘otal number of ways the die can land 22, Tree diagrams for Combined Probables We have now detesmied the probability of throwing six witha single die, What about ‘thoorng two ses with wo die? In this case we have o ask to questions: 1. What does the fst de land on? ‘2. What doos the second de land on? For question 1 there are 6 posible answers of which only 1 favourable (ands on six). “Thus, we can split question I into two Branches as fllows: ands on six (199) Does not land on sx (5 ways) ‘We now wit the probabilities on each branch ots “Note that these two branches fiom the same point) add upto 1. This is always the case, or question 2, we again have the same situation ‘We must now add our “ree” fr question 2 onto each of the branches of question 1 \ y i he ‘eeu bio ‘The final probebilties, shown at the endpoints, are found by multiplying probabilities long every “pat through the re. ‘The diferent possible outcomes and ther repectiveprobablliies are: ‘Twosixes: 4 One se: $458 No sixes: Another Example ‘The probability fora male being bore colour-lind is 10%. Fora female iis only 1% ‘Anne is pregnant, what isthe probability that her baby wil be born colou-bing? ‘Question 1: What isthe gender of the baby? (Question 2: eth baby colour Bling? ‘Note tat the probabilities for question 2 depend onthe szrwer to question 1. Thus the ‘tee must start With question I 0 Lev Coloured 0.05 vite P os 3 cotmeding 0005 a Femie a 099 Normal 0.495 “The probability tht the baby wil be bom colou-lind i 0.055 (5.5%). 23, Set Theory aud Probability Probability can also be defined in terms of sets, To do this we must frst define afew basic concept: ‘A set is collection of objects a ou ase these objets wil be the different posible ‘outcomes (wht can happen to an experiment (like olin ade) ‘The sample space isthe tof ell possible outcomes. An event i subset (part of) ofthe sample space. Consider the following: Job rolls adic and hopes that it wil land on an even number. Rolling the dies called the experiment ‘This experiment has sx possible outcomes. “Therefore the sample space is the set S = (1,23.456) —itis important to use the curly racks. ‘Landing on an even number i alld an event of thi speciic experiment. We assign a capt leter to this event, let's say E. So we define B= (the dielands onan even number} = (2.46). “The number of objects (clement) in $s written (8), 50 (8) = 6. ‘The numberof object (olements) in Eis written (Es n(E)= 3. Probability is now defined as n(E) divided by n(S), and is written PCE). ‘The probabil ofan event F ls writen PE) and defined by. a) 1-55) Is Soin this case, PE) = ED =05 2) Feampie “Two coin ae fipped, 1. Write down th ample space 2 Fad the probaly at both and on had 1.8 (GLE). CL, (CHD , (1D). Note ha (4,1) ina diferent outsome fom (1. 2E- (an) a= 1 nS)=4 ‘Two dic are rolled. 1. Write down the sample space. [Link] the probability tha the sum ofthe numbers tat the die land on is 7. 18=(G).0.2),0.3),.0.9,0.9.00, @1).02),03).09.25.05, GN.62).89).64).6.9.66, 41142)/49).40,6.9,60, GN6N.69.6069:65. 6D .62),63),64).6.9).68) }- 2B=(6),62),43).04),25).(1.0) B= 6 1(8)= 36 926, P= ag I 13 24, Ven diagrams ‘As we eve seca, events ate contained within the sample space and are subsets of te Sample soce. Ths canbe represented witha pctre called a Venn-dingram: s ‘We usually use a rectangle forthe sample space and ctl for evens. If we ook atthe example shout Jabu and his die, the Venn diagram wil lok as flows: s 4 25, Union and Intersection ‘The sion of evo events A andB, writen AWB, is the event that A wil take place, of B vl tke place, or both will tke place. la other words tha at est one wll ake place) In tems of ses ths means that “we throw the two sts together”: AUB ‘The erscton of two events A and B , writen AB, ithe vent that both A and B ‘wil tae place ‘Inte of sets this is where he two sets “overlap”, Bsampte ‘Thabo olla adie. Let A= {ie de lands onan even umber} 1B {the di lnds on amulple of 3) |. Write down AWB. 2. Wate down AAB. 3. Determine P(A), PB), P(A UB) and PCAMB). Is LA Bs Using the values above, you can verify fr yourself that: PAUB)= F(A) +PB)-PANB) ‘This elation is always rue and very usefl Ib Buample ‘Suppose 50% of students at a university can speak Adikams and 60% can speak Zul, ‘while 20% can speak both languages, what percentage of eudena cin speak at last one ofthese two languages. AUT)=P(A)+P@)~PANZ) =05+06-02 =09 ‘Thus, 90% of students can speak t least one of these languages. 26. Complement ofan Fvent ‘The complement ofan event A isthe event that A doesnot take place anda writen X. s For example the complement of {it wil rain tomorow) s(t will not rain tomorrow). Teshould make sense to you that: 2) PA) ‘This always true and, apni, very uefa. ‘Example Refer to our previous example of Anne's baby. Wht the probablity that her baby will ‘ot be bor colour-bling? {not colourbind) = 1-P(colour blind) 10085 0945 7 2.7. Mutually Exclusive Events _Matually exclusive events ae events that can never happen together. In terms of set, ‘ven thot re mutusly exclusive has bo inersecton, Mamematcally We say A end sre mutually exclusive FAMB = 2, tis very imporant to note tht events A and B ee mually exclusive the following le applies: P(AUB) = P(A) + PB) 728, Independent Events ‘other appens or nt. Independent event are events whose probabilities are not influenced by each other. Ih osher words, the Iklihoed ofthe one event happening doesnot depend on whether the fevents A and B ae independent, the following rule applies: PCARB) = PAD PB) 1% EXERCISES B (.1 and 2.2) 1. Determine the probability of each of the following events: 1.1. Drawing en ace fom a standard deck of cards 112 Selecting the winning horse in a rave of 8 horses. 13. Peking blue smart fom a packet of smartes wih 5 green ones, 7 brown ones, ‘rad ones, 8 be ones and7 Jello ones. 2. Consider conventional sx-fced dice, What isthe probability of throwing 2.f- m6 with single die? 22! a total of 7 with to dice? 23. double 6 with two dice? 2.4, two dice and geting exactly one 6? '3.A golfer has 70% chance of playing a god shot the uses the sorect cub and 50% chance of playing good shot fhe uses a wang cub. H has ony four els in his ‘ihag and chooses one at random. What is his chance of playing good shot? [Link] sunny day Sandy basa 60% chance of feeling good. Its cloudy this chance falls 025%. The weather forecast predicts that there is 840% chance that tomorow willbe a cloudy day. What is the probability that Sandy wil feel good tomorow? ‘5. Thoms plans to have two chien. What i the probability tht he will havea east onebay? [Link] contain 2rd balls, 3 yellow balls and 4 blu alls. 1f2 balls ae selected one ier the othe, without replacement, find the probability tht 6.1 both ball are the same colour? {62.00 red bllschosen? 7A litle ol lady keeps her money ina sugar bow. She has two R100 noes, four R50 ‘notes, five R20 notes and nine R10 notes. The old lay reaches ito the bow and grabs to notes. What isthe probobility that: 11. she draws (wo R20 notes? 72. total she grabs is ess than RSO? 5 EXERCISES C (23-28) 1. Write down the sample space foreach of the following statistical experiments 1.1. A coin is spun three times and the result ofeach suecessve throw is recorded as H (Gade) oT (ail). 12, Thereslt ofa cricket match ean bea win (W), alos (L) ora raw (D). Consider the ‘outcomes of two consecutive matches. 2. ight Ri-coins, numbered | to 8 replaced in container and shuffled propery. ‘One of the coins are draw (t random) from te container and its numbers recrded, 2.1 Write down the sample space for this expeviment [Let A= (an even number is recorded) B= (aprime number is recorded} en even nimber, greater than 3, is recorded) 12.2. Wee down, inst form, te following events listing alts elements: ABandc. (i) ALB, AUC and BUC. Gi AAB, ANCans BAC. 2.3, Caleulte he probabilities of ech ofthe events in 22. ‘2.4, Which ofthe events A,B and C ifany, are mutually exclusive? 4 3% At «school 21% of pupils were absent on Monday and 15% of pupils were absent on ‘Tucslay. 10% of pupils were absent on both days, what percentage of prpils were beat tall on these two days? What percentage of pupils were present on both days? 4.1FP(A)= 055, P)= 04 and POAMB) = 0.25, calcu 41.20). 42, P(ALB). 43. GAB). ‘5. The probability thet a rndomly chosen commerce student wil pass accounting is 0.7, the pobeblity that he will past Economics s.8 and the probability that be wil pass Doth 0.6. Find the probability thet: 5.1 hepasses at est one ofthe subjects 5.2. hepanses neither subject. 20 v 6. group of 10 000 people were tested fr colou-bindness nd he results ae shown in ‘the given table: ‘aie Fema Tot Colour-biiad 3 is 408. ‘Normal 4517] 095 9597 ([rotar 4910 | 5090. 10000) Doss the evidence seem to indicat that coloucblnluow is independent of gender? 7 Before going on an oversees holiday , you are advised to have inoclations aginst ‘holer and typhoid. Ifthe probability tat the inoculation agin cholera works is 0% and the probability tha the inoculation again typhoid works Is 95%, what isthe probability that: 1.1 nsther inoculation works? 172. both inoculations work? 173. exacly one works? Y 8A gio isto by her boytiend that she is “one in bilion®. She has a dimple in her chin, probability 0.01, eyes of diferent colours, probably 0.001 andi ebsolutely crazy shout mathematics, probability 0.0001. 1s her boySiznd's citim tue 3. Normal Distribution ‘When we look at reasonably large set of objects and measure the same aspect ofeach of ‘these objects, we get a disuibution With the word distribution we basically refe to how the fequences are disibuted over the different values For example, if we consider the lengibs of 100 diferent people, we my find that 50 ofthe people are taller than 176mm fd the eter 50 are shorter thn 176em. Typically the most people wil have length fround the average. The further we mave away ffom the average, the less people we will find. We would cenainly no expect many people tobe taller thn 210em. Likewise We ‘would not expect many people tobe shore than 142em. We can represent this szapicaly: Tas VWéem 210m (On this raph, the diffrent values that ean be measured are pu on the Xaxis. The Y-axis {s used fora measure of how likely itso td value close othe given X-value. The precise meaning ofthis graph il noc be discussed here. We will only use the raph to answer certain questions about the distribution. ‘You wil note tat ll the dsebutions we work ith wil have this basic shape, Its sometines refered to asthe "bell shaped cuve” or the “Gaussian distribution”, This kind of disritution ie very common. “There are vasitions in these graphs with respect to postion and “width. “The poston is determined by the mean (average), andthe “width” is determine by the standard deviation. A graph with a higher standard deviation willbe both “wider” and “fair” than ne witha smaller standard deviation. ‘We will find proportions and probebilties by looking t areas under the graph. ‘Thowe eras ae read from atabl. “The otal are under any normal dstbuton graph is 1. The graph is slays symmetrical with respec tothe vertical Hine through the mean and therefore the total rea lef ofthe teen equals the ttal are right ofthe mean and both are always equal to 0.5. Consider the following situation ‘Seppo the waist measurements () of 900 gis between 16 end 20 are normally Aiton with mean 6Sem and standard devition Sem, Fest ofa, we'll introduce some notation. We we the symbol 1 (nu) foe mean (average) an (sigma) for standard devistion, So 265 and ‘To summarise all his information we write X~ (65,28) The 25 is called the variance of the dsrbuton (square ofthe standard deviation), (Constethe question; “How many standard devistions away from the ean can walt ‘measurement of 75cm be found?” ‘Tosnsver this, we subtract the mean, 65, from this value,75, and dlvidethe answer by the standard deviation, 5: %. ‘= standard deviations. 3 ‘The number of standard deviations is also called the valu, =p] [very value (specific measurement) has a coresponding -value (aumer of standard 75). a x Tuan betas to NZ Qe ‘Therefore, our answer is 050.4772 = 0.0228. Another Example “The width of bolts of fabrics normally distributed with mean 950mm and standard deviation {Omm. What is the probability that a randomly chosen bot has a width of| beeen 947 and 958mm? POM < X <958) =e (03.

You might also like