Stat 217 – Day 5Last Time – Random SamplingSlide 3Be forewarnedActivity 5-1 (p. 74)Activity 5-1Slide 7Slide 8Let’s try itThe Gold StandardSlide 11Slide 12Stat 217 – Day 5Random Assignment (Topic 5)Last Time – Random SamplingIssue #1: Do I believe the sample I have is representative of the population that I am interested in for this issue (generalizable)?Yes, if sample was selected randomly from population of interestComplete sampling frameSimple random sample (SRS) or another random sampling method to select from sampling frame eliminates sampling bias (sampling distribution of statistic centers at population parameter value)Human judgment of randomness is notoriously faultyLast Time – Random SamplingOther benefits of random samplingLater will be able to estimate how far our sample statistic might fall from the population parameterLarger samples will tend to fall closer to the population parameter than smaller samplesThe size of the population is largely irrelevantActivity 4-18, most to least sampling variabilityn=20, US Senatorsn=1000, New York residentsn=100, New York residentsn=500, Wyoming residentsBe forewarnedAre still other sources of bias “Nonsampling errors”People lie, don’t rememberWording of the question can evoke emotional responsesAppearance, tone of interviewerTiming of surveyActivity 5-1 (p. 74)Issue #2: Can I draw a cause and effect conclusion (causation)?Research question: Do “strength shoes” (modified athletic shoes with a 4-cm platform attached to the front half of the sole) increase jumping ability?Activity 5-1(a) Your friend says it worksNo, is just one case, not really based on any systematic study, can’t rely on our “impressions”Def: Anecdotal evidence = personal experiences, individual accounts, striking events, atypical outcomes can’t be trusted!Activity 5-1(b) Compare a group wearing strength shoes to those notEV = whether or not wear strength shoesRV = jumping abilityDef: Observational study (p. 38) = passively record existing information may be confounding variablesActivity 5-1(d) Assign subjects to shoesDef: An experiment actively imposes the explanatory variable on the observational units(e) How create these groups?Strength shoesRegular shoesJumping abilityObservational unitsrandomLet’s try itIn Blackboard, choose “Course Materials” and then “Stat 217 Java applets”Randomization of SubjectsObservational units and variable?Answer (d)-(i)The Gold StandardBy randomly assigning subjects to treatment (explanatory variable) groups, I have done the best I can to balance out all other factors between the two groupsEliminates potential confounding variablesIf later I see a significant difference in the explanatory variable groups, only reasonable explanation is the treatment that I imposed!Can potentially draw cause and effect conclusion!To Turn In with partnerActivity 5-5 (p. 80) (a)-(c)For next classActivities 5-4, 5-6p. 93-4, Preliminaries 1-3Lab 1 continued, HW 2 posted
View Full Document