Homework Solutions for Chapter 7: Hypothesis Testing - Prof. Spencer Muse | Assignments Data Analysis & Statistical Methods

ST301

Homework11Solutions

CH7

53.)(a)Thedatacan’tbenormallydistributedbecausethereareonly7possiblevalues(0‐6),all

integers.ThisdataactuallyhasaBinomialdistribution.

(b)Wehavetwolargesamples(n1=165andn2=212),sothesamplemeanswillhavenormal

distributions.Sothemethodsofthischapterareapplicable.

(c)H0:µINTERVENTION=µCONTROL

HA:µINTERVENTION≠µCONTROL(two‐sided)

HA:µINTERVENTION>µCONTROL(one‐sided)

Theone‐sidedalternativewouldreflecttheresearcher’sbeliefthattheinterventiongroupshouldhavea

highermeanscore.Thetwo‐sidedalternativeallowsforthepossibilitythattheinterventiongrouphasa

lowerscorethanthecontrol.

(d)















 ..

.

 .

  6.258

.



df=354(software)or164(approx).Eitherway,p<0.0001.

Withsuchasmallp‐value,RejectH0.Concludethatthereisasignificantdifferenceinmeanscore

betweentheinterventiongroupandthecontrolgroup.Sincethedatashowthattheinterventionscore

washigher,wecanconcludefurtherthattheinterventionincreasesmeanscoreovercontrol.

(e)95%CI:(xbar1–xbar2)±t*SEDIFF=(5.08–4.33)±t*(.

 )

t*dependsonthedfused,buttheintervalisabout(0.51,0.99).Thismeansweare95%confidentthat

themeanscorefortheinterventiongroupisbetween0.51and0.99pointshigherthanforthecontrol

group.

(f)Theresultsarenotgeneralizabletootherpopulations.



55.)n1=115,xbar1=18.0,s1=7.8

n2=220,xbar2=6.5,s2=3.4

(a)Ifthiscompany’sworkingconditionsaresimilarenoughtoothers’,thismaybeagoodenough

randomsample.

(b)Youmayusedf=137.1(software)or114(approximation)or100(closestvalueintableD).

Partial preview of the text

Download Homework Solutions for Chapter 7: Hypothesis Testing - Prof. Spencer Muse and more Assignments Data Analysis & Statistical Methods in PDF only on Docsity!

CH 53.) integers. (b) distributions. 7 We (a) haveThe This data two dataSo can’tlargethe actually methods besamples normally has of (na 1 thisBinomial =165distributed chapter^ Homework and distribution. n 2 are because = ST212), applicable.^30111 so^ Solutionsthere the sampleare only means 7 possible will havevalues normal (0‐6), all

(c) H H The higher lower (d)A (^) A (^) : : H (^) ܶ μμ (^) ൌone (^0) INTERVENTION (^) INTERVENTION : scoreμmean (^) INTERVENTION‐ ට௫ଵതsidedതത ೞതభమି (^) ௫ଶ (^) തthanത (^) തscore.ത alternative≠> μμthe = (^) CONTROL (^) CONTROL μThe CONTROLcontrol. two(two(one would‐sided‐‐sided)sided) reflect alternative the researcher’s allows for beliefthe possibility that the interventionthat the intervention group should group have has aa With between was (e) t* the depends95% meanhigher, such CI: ೙భthe scorea ା (^) on(xbarwesmall^ ೙మೞ interventionమమ the^ ൌ canfor^1 ට p – ହ.଴଼ିସ.ଷଷdf^ భ.భఱ ‐ (^) theభలఱconcludevalue,xbar used,మ^ ା intervention^ భ.భల 2 మభమ ) groupReject± మbut tSE^ ൌ 6.258 further the andDIFF H 0 interval. = (^) groupthethat Conclude(5.08 control the is df–is between^ 4.33)interventionabout= that^354 group. ± (software) (0.51, there t 0.51 (ටSince (^) ଵ.ଵହ is 0.99) increases ଵ଺ହ and a (^) మ significant^ theor൅ .0.99 This (^) ଵ.ଵ଺^164 dataଶଵଶ (^) meanpointsమ (approx).)means show difference score higher thatwe^ Either over arethe than in 95% interventioncontrol.meanway, for confident the p score^ <^ control 0.0001 score that. group. (f) 55.) n2 (a) random The=If n1 220,this results= sample. 115,company’sxbar2 arexbar1 = 6.5,not =working generalizable18.0,s2 = 3.4 s1 conditions = 7.8 to other are similarpopulations. enough to others’, this may be a good enough (b) You may use df = 137.1 (software) or 114 (approximation) or 100 (closest value in table D).

SE This higher (c) H drill T A=D (^) : H μ= (^) ௫ଵത means&ത (^0) DRILL&BLASTത (^) :തට (^) ି (^) ௫ଶ (^) μblastfor௡ଵത௦ത (^) DRILL&BLASTതభమത ൅drill we (^) workers௡ଶ௦ మమ> andare ൌ 0.7626μ = (^) CONCRETE μ95%blast (^) CONCRETE will confident workershaveI use a higher thethan that one forthemean CI‐ (^) sided=concrete true (xbar1 exposure alternativedifference – workers. xbar2) than inhere ± outdoor meant*SE becauseD =exposure (^) concrete (9.99, it’s 13.01) reasonable is workers. between for any to9.99 of assume the and df 13.01 theused. ට ೞ೙భభమା (^) ೙మೞమమ^ ൌ 15. |ݐ൐ 3.11ሻ

With significantly (d) 59.) there The (a) such is skewnessIt no makes a uncertainty smallhigher no doesp ‐thansensevalue,^ df notabout^ = the^137 to Rejectinvalidate meanstate its(software) value. H hypothesesexposure 0. (^) ourI conclude We^ oranalysis^114 want for in^ (approx).concrete thatterms tobecause hypothesize the of mean^ workers. xbar. Eitherwe have exposureXbar way,about large is^ p^ thethe <^ forsample 0.0001. sampletrue drill population andsizes mean blast (115 we workers andmean, observe; 220). is denoted (b) (^1010) (c) 61.) boysboys.WeThe (a) need methodsUsing is μ. not P software,independentto ofbe this small (^) thesection to ofp reject‐ (^) valuethe assume group H for 0. Athisthe oflarge alltesttwo kids, P is: groupslike P(because this| being gives the compared no group (^) (notereason of the are toall absolute (^) doubt (^) kids independent contains H (^0) value).. the Thewith group group 23 of of degrees level (b) negative). For or the evenof Thisfreedom.one the‐ sidedprobability 1% (^) level. Usingalternative issoftware, very that large; thisthe P differenceprobability= 0.9975. isSo is positive, lesswe would than the 0.005. Fail p‐ valueto SoReject weis: P(tcan H 0 >. reject ‐3.11) H (note 0 at the the 5%

Homework Solutions for Chapter 7: Hypothesis Testing - Prof. Spencer Muse, Assignments of Data Analysis & Statistical Methods

Related documents

Partial preview of the text

Download Homework Solutions for Chapter 7: Hypothesis Testing - Prof. Spencer Muse and more Assignments Data Analysis & Statistical Methods in PDF only on Docsity!