Download as pdf
Download as pdf
You are on page 1of 11
(Britney € Bailey | STAT 135 | Exam Due December 9, 10pm ET STAT 135, Take-home Final ‘This final assessment is a cumulative, handwritten exam and should reflect your individual understanding, It should take about 2 hours to complete, | would encourage you to have by your side: + your cell phone (or tablet) for scanning work (or annotating the pdf in an 2pp) = 2 pencil (or stylus) and paper (oF tablet) + 2 calculator and/or RStudio + a tworsided sheet of notes You may reach out to Professor Bailey via campuswite if you run into technical issues, but otherwise you should rely only on course material and R as needed to complete this exam, SHOWING WORK For fll cet, answer in complete sentences in context af the problem, and fully justify each response, showing all work necessary to demenstrate your knowledge of the process and/or concepts Write out steps. Indicate when you have used R as part of your work. Draw and label pictures (e.g., probability distributions). Define notation. Show formulas used in the process with appropriate notation. Show a sufficient amount of calculations along the way so | can track your reasoning. Unless explicitly told otherwise, answer in complete sentences (as if the question were not right there) in context and justify each response. If you do not justify your work and provide context in your answers, you will not receive full credit. EXAM FORMAT If possible, I would prefer that you fill in a printed copy of the exam or annotate the paf on a computer/tablet using a stylus if you have those capabilities. Do not type your answers into the pa. | expect most students will not have the option to print/annotate the pdf; in that case, please complete your work using your own paper with the problems clearly labeled and well-organized. Please do what you can to keep solutions from splitting across pages (you might want to match my spacing in the pdf), ‘This exam is due by 10pm ET on Wednesday, December 9. This is a hard deadline, so please plan accordingly. J understand the expectations and agree to the above rules. | affirm that | will neither give nor receive any unauthorized help on this exam and that I will comply with the Amherst College Honor Code. Signature: |< == —— ‘ernr A |[s0 Youre samme] so voure —[SooNns UME YOU wave NO wore uc || net beat A” | sacs | een ee Wr we eves B. Basa Over LEW OAT mene ngs YOU BIS CONVERSATON Gunso | | cere | | wmnco ver: uso Nor 150 (cus men eons BRE You'd think i'd be easy to just bet money against these people, but ‘you have to consider the probability of them paying up Source: XKCD 2370: Prediction Brittney & Bailey | STAT 136 | Final Problem 1 Anti-nausea medication A 1980 article in the New England Journal of Medicine described a study on the effectiveness of medications to combat nausea in patients undergoing chemotherapy treatments for cancer. In the experiment, 157 patients were divided at random into two groups. One group of 78 patients was given a standard ant nausea drug called prochlorperazine (PCPZ), while the other group of 79 patients received tetrahydrocannabinol (THC, the active ingredient in marijuana). Both medications were delivered orally and no patients were told which of the two drugs they were taking. The response measured was whether or not the patient experienced relief from nausea when undergoing chemotherapy. The data are shown below. Relief No relief THC 36 2B PCPZ 16 e 1.1 List two possible hypothesis testing procedures you could perform to examine whether there is a difference in effectiveness between the two medications Just state the names of the procedures + Two sample 3 - best + OW Square fev © how. ener 1.2 A. 95% confidence interval of the difference in effectiveness is (0.11, 0.39). Is this evidence of a difference in ‘effectiveness? Why or why not? Ba diMereve in effec hveness ec fnteyvak heen ON al Peyenee in eBQect vered 13 If there were no difference in effectiveness between the two medications, how many patients would we expect to ‘experience relief in the THC group? No sentence interpretation needed. 1G (Brtney E. Bailey | STAT 135 | Final Problem 2. Among Us The popular game Among Us takes place in space with 4-10 people assigned to be crewmates or impostors. Crewmates race to finish tasks while 1 to 3 impostors sabotage their progress and kill them. Crewmates win if they identify and eliminate all the impostors or finish their tasks; impostors win if they kill all the crewmates before the tasks are complete. A single game is fairly quick. Suppose we can model the length of a game of Among Us using a Normal distribution with a mean of 9.2 minutes and a standard deviation of 2.9 minutes. 2.1 What is the probability a single game is shorter than 3 minutes? No sentence interpretation needed. gz luge oF gene jg. Mee. SEMA at e(y<3) | 2.2 What is the probability the average length of 5 randomly selected games is longer than 10 minutes? No sentence interpretation needed. an 2 Ua oe 26.817? « Wager barren 23 The average weebly time spent playing Among Us on Steam for all recent active players is 107 minutes and the ‘median is 43 minutes. What might this tell you about the distribution of time spent playing Among Us? o7 & [Britney E Baley Problem 3 [STAT 135 | Final Noises make you crabby Biologists in the UK studied how noise might invoke a stress response in animals. One indicator of stress response is “enhanced metabolic rate and thus increased oxygen consumption.” Focusing on crabs, they hypothesized “the oxygen consumption of crabs would increase in response to ship-noise playback. Moreover, we predicted that the effect [of noise] might be size-dependent, with larger crabs affected more strongly.” Fourty-four (44) crabs of the same species were transferred from a local harbor to an aquarium where biologists measured the mass of each crab to the nearest 0.01 gram. The crabs were then randomized to one of two treatments: ship noise or ambient harbor noise. Each crab was placed J0-centimeters from a speaker brarT-ter airtight container completely filled with water, and exposed to an audio recording of their assigned noise for 15 minutes. The biologists measured oxygen consumption (in umoles/q/hr) for each crab over the course of the treatment. 3.1 What are the individuals in this study? Just state the answer. Tha ous 3.2 List the variables in the study, their units or levels, the type (quantitative or categorical), and ther roles in the study (explanatory or response). List your answers as demonstrated. ‘Variable (Units or Levels) ~ Type ~ Role caus Rome Cg) = dumartetie - viglanetory sade Cthie Fervor) - Quitinner = uxplent bry +e 3.3 Why were 1 ryyer Conran Camel 5/6) - Qeautite regen he crabs randomized to treatment groups? y be yeh 1 vege Wa emo gepniabou fe evperieuce (Mir FEYLENVE found, ‘Tri pemrewd le groupt repre analahis The ambi The other two were 3.4 The treatment group assignments forthe first three crabs were ship noise fist crab was smaller (29.2 g) and consumed 1256 jumoles of oxygen per gram per hour larger (61.3 g and 68.5 g) and consumed 181.9 and 157.2 sumoles of oxygen per gram per Organize this data as if you were putting it into a spreadsheet. (Start in the top left corner, here 0 rows/columns than needed ) respectively y be more 3 avabiowr s ampient 1 larger species and a smaller 3.5 Suppose the biologists wanted to include two different species of crabs in their study. 2? Describe species. They gather 44 more crabs of the second species, What type of study design would they ust how the study would be carried out. The bielacly would anseth anehhir Qandemiged contro (Ud Wel sunilar e cele would be Andomy Cre qectec, The conducted om We Fe We brik Whe Janne neces Upane A take ae hive bree neue graph pone ce lake, Tas why flay CAA obevee and determin We corre tta hen ok guly wuss @ ovyun or sound e Wyre, bur are species @ cf uy | STAT 138 | Final ray 68 Problem 4 DATA EXPLORATION 4.1 Describe the distribution of oxygen consumption (in jumoles/g/r). 4 get histogram gf _histogran(date = Crabship, ~ Oxygen, Ef Aaba(e = "Oxygen conouaption\n(eicrosoles per gran per ho binvidth + 61.19, center = 61.13/2) ih ” 15 10) count 100 20 abo ‘Oxygen consumption (micromoles per gram per hour) 4 get sumary statistics favetate(data = CrabsShip, - Oxygen) ‘> kable(digits = 2, booktabs = TRUE) 7% kable_styling() Qi median QS wne mean sed rising 70 15838 19285 2223 3280 103.147.7744 0 Ara dbs pala B oeyyer consump bon is peal noel ond uninadeth 442 What graph could we use to visualize whether there is a diference in crab mass between the two treatment groups? Just state the name of the graph here ox plo 4.3 What procedure could we use to formally test whether there isa difference in crab mass between the two treatment ‘groups? Just state the name of the procedure here. T- yee ‘Britney E. Bailey | STAT 135 | Final 4.4 Why would we be concerned if the group of crabs exposed tothe ship noise were significantly larger han the grow? ‘of crabs exposed to the ambient harbor noise? TWN would cetere Lies Jn Mae centile. The fH oF ea eset us be ANGE hing Mae res ly aed ongyen Corsimny eae a HOT Tenth ve FetMqa ye ATA yor 'able Moving 4m puck om Haw reset. ARuss becemag & onMrur any variable / ee. >) Ree on Me ory jer comaumpron, Cb yyee 11me mee, srele 4 amveur ) 4.5. What would we cal the issue in the previous question? Just state the word or phrase here Cre teundirg Qint, 4.6 Describe the relationship between crab mass and oxygen inake, # get scatterplot gf point (data = CrabShip, Oxygen ~ Mass) ‘>% Gi labs(x = "Hace (grans)", y = “Oxygen consumption\n(micronoles/g/ur)") {(micromoles/g/hr) gaeee ‘Oxygen consumption 3 ‘Mass (grams) Rene Boa WK, Goritive, Une pelubouchip behwecn ott onait aun onyjen inate, ‘Untony E Batley | STAT 135 | Finat 4.7 Now look at the relationship between crab mass and oxygen intake by treatment group (noise exposure) Comment ‘on whether It would appropriate to use correlation to describe the relationship within each group. # get scatterplot gf point (data = CrabShip, Oxygen Mase | Noise) >' x= "Mass (prams)*, y Seo : M FS a0 ; : ao _ 200: *a-* ? 88 ioe 5) wd E 100, & Le ee Ce Mass (grams) for we awnblen group ik geemy ch Urovrgn Hue ox, Wes lake Ca Htens Our Maus sUdoehng eg geabeg et naa Uncir, ave gh Hee uth Fe become He mage aime A celervely samatl, Gorvelabin wuy Le appeprinte, Foy Ure Ship gmup ik @ a srsaue g Ss Haus p elearly Uneeie peterivans hip Coree layin a pQropetare 4.8 One group has a correlation of 0.79 and the other a correlation of 0.51. Which treatment group do you expect to have a stronger correlation? Why? Vexpery Ure chip gioug le have a Shtayer correlation , beccesse F Gollma @ Une much enre cltarly Mam Wee omblent wearauar jovg Tiere nee ae Feuer outlng ge Woe ship hetermet guoug Attney E Bale | STAT 135 | Final Problem 5 MODEL FITTING AND DIAGNOSTICS. 5.1 We fit a multiple linear regression model to the data to complete the analysis. The residual plots are shown below. Discuss whether the conditions for regression are satisfied. # fat model crabla <- In(data = Crabship, Oxygen ~ Mass * Noise) # add residuals ond fated values to dataset CrabShip <- mutate(Crabship, residuals = resid(crabla), fitted = fitted(crabim)) # create restdual plote Pl <> gf_point (data * CrabShip, residuals - fitted) 1% gf_bline(yintercopt = 0) p2 << gf_qq(data = CrabShip, - residuals) ‘> gf _qqline() pS <- gf _histogran(data = CrabShip, - residuals, binvidth = 22.5, center = 22.5/2) Pl: 2s po » Ow ye 3 a a ee 3 oe Bateman 100180360280 2a oT 8 8 fitted theoretical residuals, lesen ee eee Crabs & independent Jv Urtaviry - me ceumoastig of er cesiguucly > Vinee Yo Guat Votan Une git shaus eyaal Sater wih a band abou 9 {Sey Boral Dedounen 5.2 The biologists were considering adding a third v le to the model: the width of the crab, The correlation between the width of the crab and its mass is 0.93. They ask you if they should include both mass and width of the crabs as explanatory variables in the model, What do you tell them? I would ku thew V Ree Cert lol, verweem the wid and i Ye, Hee duo have! colingevity , md 60 Ie would be bath i remove Include vie of Meese Vavialees, ov net ney E Balley | STAT 135 | Float Problem 6 MODEL SUMMARY 6.1. Assuming the conditions for regression are satisfied (or proceeding with caution), 2 summary of the model is shown below. Write the equation of the regression line # get model summary ssunsary((crablm) Estimate Std. Error t value PrO>itl) (Intercept) 62.0719 20.4156 3.040 0.0041 Mass 2.0271 0.9709 9.466 2.480-06 ++ Noiseship 64.7669 11.7484 5.513 2.130-06 s+ Residual standard error: 38.75 on 41 degrees of freedom Multiple R-squared: 0.871, Adjusted R-squared: 0.6501 Festatistic: 27.29 on 2 and 41 DF, p-value: 2.920-08 er ors + 2.0244 muss?) «6.1664 Caoits SHIP) oes 6.2 Test whether, after adjusting forthe type of noise exposure, there is 2 significant association between crab mass and ‘oxygen consumption. No need to recheck conditions Weep. ee sna To E(AT ke SMG AB, £0 hati “ay Covalee PCC E Suen ov TE -saoe | He) La rapa < o0et ret Me Thee 8 evident be conclude that Ware & sywifieanh osu tation Whueen ever gesss awd ony inate 03 tert the cote abled Mo 1s6sbsp the oetrelnnh Pek Hak bee the Crebs hater . . Inve 44 167 peonak glans ' _ aan ) {6.4 Interpeet the 90% confidence interval forthe coefficient for mass confint(crablm, level + 0.90) 5h 95% (Intercept) 27.714967 96428778 Mass 1.402951 2.651264 Noiseship 44.995720 84.637981 De ace Q07, Comdenk thay HL be vated oP HL inertan 98 ony ie penatf glee Aboy & ONE Inbneer B pass ineccert by LY Calls L403 one 2-651- 6.5 Predict the oxygen consumption ofa crab in the ship nose group with a mass of 110.3 grams (no sentence needed), ‘or explain in one sentence why we should not make that prediction NWO3 % curvy de (re Lomuls ob cbsevved mased We researches documenta Se Woe predicted outcome VIN be uneeligete 66 Predict the oxygen consumption of a crab in the ambient noise group with 2 mass of 41.9 grams (no sentence ineeded). or explain in one sentence why we should not make that prediction. ~ — endijen = 62.9714 4 aor (Mra) + 6H Teer Go> 2 aT Ot pmol /g/ bed 6.7 Thinking back to the biologists’ original questions, how would you summarize the results of this study for a broader audience? Te oryean intabe mee ia combs incised a Mae eas OF tne erate inereove! and fan evals A under siels, The petals shea Phar lod chip sounds cain Pe yer IAAY shows ceiganne , Hus incramning oryges irtele + u

You might also like