Class Notes
(810,487)

Canada
(494,138)

University of Toronto St. George
(42,836)

Sociology
(3,207)

SOC202H1
(76)

Scott Schieman
(31)

Lecture

# lecture 4

by
OneClass5584

Unlock Document

University of Toronto St. George

Sociology

SOC202H1

Scott Schieman

Winter

Description

Lecture 4 January31121000 AMSample distribution sample statistics especially means whats the spread and the central tendencyThe sample estimates how good or bad are they Looking at the distribution of the sample statisticsSample Size matters What would happen to the man and standard deviation if we added 1 more case with 1 hour as the response the larger the sample it absorbs the amount of potential influence from the outlierThe more deviant cases especially with mitt Romney one or two cases isnt a big deal But with a small sample size and a few cases then the mean will change and pulls the averageOutliers cause problemsSampling DistributionsOften stuck with the idea that our samples are not truly representative If they were perfect precise measures itd be the entire populateHow good or bad are he estimates relative to what is happen how likely is it that if we draw another sample in the same wayit should be very close If the two are wildly different if so then we say the difference of the means if you keptdoing random samples Each of those means are a sample meanSampling error is patterned and systematic so it is predictable They cluster around the true population mean as should standard deviation Most cases cluster in the middle The bigger samples are better meaning lower standard errors Imagine in the sample it is critical to have a lower standard deviation it is a component to calculate the standard errorIf you plot the sampling means into the distribution you get a normal curve with frequencies assorted wit the sample meansSample to sample variability we use what we have to calculate it We use standard deviation to calculate It is mostly about the size having a direct implication and the size of your estimate If you inflate that the standard deviation goes up Making the spread large making it less creditable The lower the standard deviation the more credible The denominator is based on the sample size The sample size is larger that is better The sample size undermines the impact of the error The larger error but more of a cushion with your scoreWhat increases the standard error larger sample standard deviationLarger sample size the smaller the standard errorA substitute for a frequency histogram or polygon in which we replace these graphs with a smooth curve Soc202h1 Page 1 curveThe lower the frequency the lower it is in the sampleMost of the observations are in the middle of the bell shape curse For sampling distributions it isthe sample of the means As we move away from the mean they decrease50 r 50 on the middle 68 fall withinor 1 deviations95of scores fall withinor 2 deviations99of scores fall withinor 3 deviationsTo compute a Z score numerator subtract the mean from the raw scoreFrequency of work related contact The higher the standard deviations within these groups create higher standard errors He sample variability can be all over the mapIdeal scenario for a stronger case in the population work calls are associated with higher work and family conflict The best case is lower deviations All 367 said the same value but in that scenario and sample estimate was the centre and is the sameSMALLER SUBSAMPLE SIZE MEANS BIGGER STANDARD DEVIATION BUT MORE ERRORQuantitative Methods SOC202H1SLecture 4January 31 2012Professor Scott Schieman1 Soc202h1 Page 2

More
Less
Related notes for SOC202H1