0915 read ch 5 normal distribution
- z = (y- y bar)/s = (y-mu)/sigma, .682/.954/.997
- range = highest - lowest
- N(mu, sigma^2); mu & sigma do not come from population but data or sample
- p(i)=(i-0.5)/N where i is 1-N, y&p(i); to determine y is normal or not, unimodal&symmetric
0914 read ch 4 understanding & comparing distributions
- Iz-scoreI>3; unusual
- 500대기업사장의 평균연봉은 5백만불이고 반수는 2.5-7.5사이지만 15를 넘는 초고액연봉도 상당해서 최고 150까지 오른쪽으로 많이 비틀어져있고 로그값으로 하면 정규분포에 가깝다.
0913 self test1&2 by 0920 23:59
0912 read ch 3 describing quantitative data graphically & numerically
- caregorical vs. quantitative; bar/pir vs. histograms/stemNleafPlots/dotplots
- spread; range/InterQuartileRange = Q3-Q1 < range/2, Variance/StandardDeviation, use IQR > s
- skewed; one tail > the other tail, left > right tail; skewed left/negative
- mean/median/mode; ave/middleRank=Q2/mostFrequent, use median > mean
- variance s^2=sum(x-xbar)^2/(n-1) sigma^2=sum(x-mu)^2/N
- boxplot; *ourliers Q1-1.5IQR--Q1;median;Q3--Q3+1.5IQR *outliers
- 5 #; min/max q1/3 median
0910 read ch 2 displaying & describing categorical data
- marginal vs joint distribution; sum case/N vs case/N
- conditional on gender; case/sum case
0909 read ch 1 stats starts here
- Categorical/Quantitative variable; Sex or Group/Age or Height
- Nominal4Categorical, Ordinal4OrderedCategorical
- Scale4Quantitative; can add for meanning, must have units
첫댓글 https://youtu.be/WuDGps5jw7s
PLAY