INTRODUCTION TO STATISTICAL BIOLOGY Biostatistics Bratch of opel Hok des wilh the applcatinsoF shtichical mec in hlth-rdke! fel, meng publi halt, meine an bidegy Bio Life *Bioslatislics” Sctowe that dds wth eel, gern, smarsin,prakn a anf ch ~ apples shail mos ml an yl presi *Mathematcl Sitios date wh Neco ens hse eb fr ssh el “concerns the cleopmat of new neti of sll into ae ries debe kandege of cba mmthevatcs for its inplenerkion ‘PopledStoishes ee apo ck alle re tel pi nll a te dpe oe si ch ofl by ed probs I ines prying the nels of mlhematnl shox fo spite. sobject Aner, sich or exami, peyclloy , ara plc hel Beostahstcians andthe Confributions ‘bie Cutle~ Bj chron [nan cone he Ky ol pc mk fis ol appl them fo pits OF bidhgy,meaiie, and sang 2) Francis Gallon - Father of brestahshics ord epics 3) Kor! Pearson Founder of madern ctoishies ‘Pore File Ieee hin vc He pal ido of pn oe Branches of Statistics °Descriptie Sblistis “ins at seman ced pean dao in He fem which wl mle Hon ene fo unestd onl wpe ‘Tables | Graphs 2)Measives of Central Tadley 3) Measures of Rsxtion ‘)Mosieso Vavbly 2) Mensves of Condlobin° Inferential SHoisis ~ fies drawing ond making ceson oth pain bre on eile hie fn 0 snk 1) Esinahin Hypa tn Cosine Salish: °Promchle ~Assmes row dicibton nd inks tating of hy: bt he ppdain mon °Norpranehie—Da dst isnt asec dik dling of hypo epi mean Paometric Non - Foramelr' ¢ ‘ove semple pee Sign tet ee ee ee 9) fe llpdet | 2-tat Whang eet feats BE a Tang Frelron Parametric. statistics ~ Qppropriate generally fr interval avd ratio data. ~ requires large sonpk sie (nz20) Nonparanehic Statistics ~ a be used for nomina|| andl ordinal data: Heanako be used in infewal oF talto data ~ Cah be used! if the. sample size's not large enowyh Measurement He press of signing meaning nmr finials bas on te dayee to which they posses portivlor Chavaclerise Levels of measurement: ") mal Se Csi of finde setof posible aes orcalyoves Hat have worded sel ~ Lowest /Lenct presse Ordinal sae fn wt of vhs /okgpres 3) Ineal Scale - based on minke] ; there is no tre 2e0 pot 4) Ratio Scale has @ tive 2ee ‘most presse soakDescriptive ard summary stabistics,ord lata visualization: “Descphie Shits ~ spent: brselly loco; debe, ond sommari calc ech doin 0 I ca, meaning anol efficient ea S hie fo describe anol vndestand the Ralvres of o specific data sot boy gira, short Summonits about the sample and measures of the data *Summiy Statistics - Summaviae and provide infermortion abort your collected data a ime something alot the vabes in your data set ~ Ita tell where your meon lies or utelher your data is skewed Meacures of Cenhral Tendency ord position |) Mean ~Sum of 0 collection of nimbers. divided by the count of mmbos in the cllrkon fornvlas ungroyped dato - % ” X- Semple mean _ %~ Symbol for sonmabin rope dole Emit 5 Fagan othe ae X.~ class mack of the css infewal 1) total. number oF cbsenotions A pediatrician had 10 patients on a particular clinic day. The weight (in kgs) of her patients on that day were as follows: 7, 17, 12.6, 15.7, 17.7, 16, 16, 11.7, 17.5, and 12.6. Compute and interpret the mean weight of the patients. Solution : n=10 weight Cinkg) = F174 12.6, 15-9, 19-9, Ib ,IG, I.9, 19.5, and 12.6 Ke Sy _ That. Grrue sti tern etstins ie = h loConsider the distribution of weight (in pounds) of the 57 children at hospital X is given below. Compute and interpret the mean weights of these children. Weight 70-99 GO - 69 50-59 4o-49 30- 39 20-29 lo -19 n=37 Total : 2086.5 *Geomelric mean Pki-%e XX 2) Median wide cbsabins tat des ecbsralins nl ho ey pos of craig cbs: fom hghst lowest or vice verso ae aoaninr 4 Dae rob tleger my, (= Ungrouped data x-(3) (G+) Xc2it* obsevahion where i*-is the 2 closest infege grater than 3 ee aan © > meshion ieee sane teers Gene ase Lin ~ lower lass boundery Fy 7 SOMf oF the olass immediately Prescting the medion ccc . - the sizeof the intewal Foo frequency of medion chss 1 total. rumber of obsenatonsConsider the distribution of weight (in pounds) of the 57 children at hospital X is given below. Compute and interpret the mean weights of these children. 2) Mode - mos} common occuring Formulas “ingraupedl aktermine the. most occvring groped tyeln + Leela Wi Weight | fi < highest 19 702 19 2 t os fey teal fib i GO-09 | 4 55 W-59 4 5] = 19% + f14-8. 4 10 4o-44 | 1% f 49 30-39 |@"| 34 =lrs sen] 20-29 \[ig\| 24 lo-ig |@*| 54) Quortiles Manes tht dvd the cbsenatins ino fiw ei! pork Formlos: yroupedl data ac oe rot tokeger career ELE iF cbsntation ee is he 2 Chast inkyo- grate than ME grouped dato ~~ Lert 4 The distribution of systolic blood pressure (in mmHg) of 80 randomly selected patients at hospital Z are given below. Determine the first and third quartiles. J hk 60) _ 99 200-214 G4 (cs) ‘20)-14 195-199 Q, =49.5 +f — Fs Ieo-In4 ] = 105-93 mig 145-159 130-144 (Sey nk ox _ Mei i (G0)-5 = 15 100-114 Ca ae gs-aq | C4) 2. JHE 25 mm Jo-842)Deciles - cbvide tolal mmber of obsevliae ‘ake 10 aya pu Formdas: roupedl dota aa 7 free ere Des Dz i* observation where i" ste chat inkge- gral Han * grouped dato Deel oe ) The distribution of oe blood pressure (in mmHg) of 80 randomly selected patients at hospital Z are given below. Determine the fourth and ninth deciles. ink , COM. 39 io” 10 00-214 Dyeltt seg 2-H 4s 75-199 2 I@o-14) | | 145-|59 , ak (8X4) 5 lo lo 130-144 We-Ra | | Da 134 sede 45 00-1 Jo-846) Racentiles - divide tolal number of observalins inte (00 equal perks Formdos: rouped doa mate Tan rot tokger c(t fe Pract o/s P.2i* obsevation where iis the 2 Clon inkge ranks Han ig grouped data The distribution of systolic blood pressure (in mmHg) of 80 randomly selected patients at hospital Z are given below. Determine the 8" and 47" percentiles. 200 - 214 I95-199 I@o-174 145-159 130-144 1B-h9 1oo~ 4 85-99 o-34Meosures of variability ° Rvge ~ the iffoene beluem the lowest an hihst observations R-(H-L) *Inferquarhle range ~ the diffeeee beh the fic oe Hid gots TQR= Q;-Q, Fite beg een “Quartile Deviotion- divides the diffbere o Qs and O, ae hales off ang IR Qp= S22 bogo mag. mins. 2 ce Average Denation Ttheasne of absolvle variability that is affected by every indiichal observation — meon_of the absolve deviations of the individual obsenation from the mean. Apt 3, Iw-a /s- Consider the distribution of the number of hours spent in Internet Café per week by 50 medical students. Determine the AD. No. ofHours fi Xi FX XX S/Ki-K/ 54-56 5 55 275 7.98 39.90 51-53 7 $2 364 4.98 34.86 48-50 8 49 392 1.98 15.84 45-47 15 46 690 -1.02 15.30 42-44 10 43 430 = -4.02 40.20 39-41 5 40 _200 -7.02 35.10 50 2,351 181.20Solution: X= 225) /50-47.02 AD: Su fleil ~ Hz, 3.70 hours Standard lea “measure of dispersion that involves all observations in the aie rather than Thragh extreme obsewations a0 feet n (K Z oss. (a j ungrowped a No. ofHours fi Xi fii (Xi- XP? Si (Ki-XP 54-56 5 55 275 63.6804 318.4020 51-53 7 52 364 24.8004 173.6028 48-50 8 49 392 3.9204 31.3632 45-47 15 46 690 1.0404 15.6060 42-44 10 43 430 16.1604 161.6040 39-41 _5_ 40 _200 49.2804 246.4020 50 2,351 946.9800 X = 2351 /50-47. 02 sp- | eho = He Fhe sons h-I 4q Population Variance | Sample Variance o* = population variance | s* = sample variance x, =value of #” element x, =value of i element «= population mean ¥ = sample mean N = population size n= sample sizeNorma} distribution with Varying means ond) standard deviations * Refer +0 module ° Coefficient of Variation - ratio std 4o its mean ara ie welly expressal in poonge Cv=S2 xto0% © Standard Score ~ measures how mony etdi an observation is above or below the mean. Xik 27S Measures of relative variability —This measure is witless and is used when one wishes tocempare te scatter of one distribution with anolker dltribuhon. Measures of Shope - describe the dishietion of the dota vithin o. dolce! * Skewness ~The degie of asyrnehy of odictibutin ond it oko hates He direction of the dishibvtin (0) Negatively skewed (©) Normal (no skew) (c) Positively skewed Mean Median Mode Mode’ Mode . Median/“|\\ Median z z & Mean Mean | x “x { | -x The normal cue N represents a perfectly Positive direction symmetrical distribution —_—_—+ ¢ Negative direction sx~2i 6 (n-1) CSD)" SK is) — skeuedl 40 the right Sk is (-)>skewed to the left Sk is (0) normal distribution= Kurtosis ~ Meosure of degree of peakelness or flatness Platykurtic Distribution Normal Distribution —_Leptokurtie Distribution Mesokurtic Distribution ill lil. ku> oe (xi-x)! (n-") (sb)" olf KU > D- leptokwhic °lf ky =3 ~ mesokuehic IF KY <3 = platy kurtic:Measues of Correlation ~ reflect the shength of association “Correlation - a biveriae omips that nancies the sheigth oF exciton behaen to wile ou He direction of the relahionchp. Generally “Volues nent 2 inal a sheng positive assoiton “Values near ~I indicate a strong negative asseciation Valves oround 0 indicate o weak association ) Rearson's Conelahion h=number of observations X= Vvolve of x mm) ~ bath variables stould be hormally dishibuted. a Iineority ~ assumes 0 shajght line relahionship i homoscedashicty ~assumes that data 1s equally distibvled about the hegression ine 9) Spearman's Ronk Correlation nonparametric “does not carry ony assumptions about the distribution of data I = Spearman Rank Condon di = the difference beeen th ranks oF colesperding vriabes = number of observations Assumptions: “must be ot least ordinal and the scores on one voriable must be monotonically relate! to the other variable3) Kendall Rank Correlation i honporometri ic “measures the shength oF dependence between tuo variables Ne a of concordant Nd= Number of discordant [Ps + Concorhant: Ordered inthe same wy * Donon: Orde offal ~Two variables should be measured on on ordinal or continvous Scale. desirable if your dak would oppear to follow o monotonic relationship Measures of Change over time ~ describe the changes in the dal. over tine Linear growth World - Population IndexExponential Qrowth Human Population Growth Logistic growth Logistic Growth Carrying capacity Later growth slows to zero Point of maximum growth beforeit begins to slow Early growth accelerates quickly Population size TimeMethods of Pesentotion of dato © Textual preserbation ° Tabular and grophical presetation of dota ~Prepeney tbls ~ Bar chorts Pow ~ Line graphs — are “Histograms ard Feqvency ive C Cumulative Fiequercy ys m ard leaf bet Plots ~ Bex Plots ~ Scatter Plols Basie Probability Coepts andl Probability ° Methods of | ae ancl probabi lity ~ Statistical experiment “Sop it Sample Space edliscrete sample space °Continwys samp space ODisccte sample space cample #1 In recording the gender of two children of two families, enumerate the sample space. Solson Let b-bey | $= {(bb), (bg), Gb). Cgg) gra! @Continvovs sample space Example #2 In recording the number of particles emitted by certain radioactive source within a specified period of time, ist the sample space Solution: 2.2 {0,),.4 a loge rwmber Consists of alll nonnegortive integers~ Event °Simple ~ 1 ovtcome ind] - union of simple events ° Sire event all ovlcomes olmmpocsibe erent- no outcome ~Venn > [Gs ice muy tg ue E ofan , a both E, Co Sorat wet iF consist Patil Example #3 ‘Suppose that a survey was conducted to determine drug preference among 700 household heads in Caloocan City. Here are the results: 250 preferred drug A 75 preferred drug A and C 280 preferred drug B 80 preferred drug B and C 270 preferred drug C 30 preferred all drugs 100 preferred drug A and B Determine by means of a Venn-diagram the number of household heads (a) who preferred drug A only, (b) who preferred drug B onl (c) who preferred drug C only, (d) who are not interested in any of the drug, (€) who preferred exactly one of the three drugs and (f) who preferred exactly two of the three drugs. vs 4) 230- (30F70 +43) =105 ©) 260 - (304 99+ 30) = bo 9 €)290- (304 50 +48) = 145, (ss 4) 700~(105 #80148 20450 45 +30) = I25 2) 105 F130 Hs -3¢0 #)I+S0t45 =03 eo Tree Diagram ~ Refers to method of finding total cample points in o sample space by listing it systonatically, ¥ Wag na i-drawing hahnRules of Counting °Wttiplication rule Ion operation con be performed inn, ways, ord it exch of 1n, ways, then the tuo bperations con be performed together Example #5 +these 0 second operation can be performed ‘n in nxn, says fe If there are four doors to a ward in one of the hospital. In how many ways can a visitor enter one and leave by a different door? Solution: (3)(4)=|I2 woys *Generalizes! Multiplication Rule: Fam oprakon con be performed in, weysond if fr each of thse 0 second operation can be performed in Na woys, and if for each of the fist two, a third operation can be performed in Ns woyg, andl avon... then He sequence of operations can be performed together in 1x1... my, ways Example #6 How many ways in three different objects which may be represented by the letters b, i, and o without repetition and with repetition? Solution: without repetition= 3° 2" | with repetition > 3% 3*3 27 ways’ Permutations ~ tefers fo on arrangement or ordering of all or park of a set oF objects lenotal by pP, where r probability oF success Begweavets Ot least = > Nn or fewer = & Only one = [=] P(%=x]n 1P>o-8 Y= P(X=n-x[n, rp) Plxe x1 np>es)=P(xdn ~xJn sI-P) Pix > xn, p05 )=POXEn- x/nsl-p) hoisson Dshibuhion PCL)= ex Normal Deion SO Fa Sfarolaid Normal a X= X= vew Score a M= mean C-= st