PM511A - Midterm Week 3

slope interpretation format for:

log (y)
X = x centered on its mean
Click the card to flip 👆
1 / 21
Terms in this set (21)
proc corr (specify pairwise correlations)proc corr data = A; var chol tgi; with bmi; title 'Calculate pairwise correlations of chol, tg, with bmi'; run; this will tell SAS to calculate correlations of each variable in the VAR statement with each variable in the WITH statement only ex: corr of chol and bmi & tg and bmi NO corr will be shown for chol and tgoutput of Pearson Corr Coefficient orderR P-value Np (italicized) is what?population correlation coefficientproc corr null and alternative hypothesesH0: p = p0 HA: p ≠ p0 where p0 can be 0 or any other given valuewhat are the null and alternative hypos for the following: Suppose the correlation between TG and BMI in U.S. residents is believed to be 0.5 based on previous large studies. Is the correlation between TG and BMI in Louisiana residents different from 0.5?different = two-sided H0: p = 0.50 H1: p ≠ 0.50set up and hypos for: can we conclude that the correlation between X and Y is different in the two populations?can we conclude that the correlation between X and Y is different in the two populations Ho: p1 = p2 HA: p1 ≠ p2write conclusions for correlation for the following: BMI & TG (R=0.40, p<0.0001) BMI & CHOL (R=0.06, p=0.44).We found a statistically significant correlation between BMI and triglycerides (R=0.40, p<0.0001), but no significant correlation between BMI and total cholesterol (R=0.06, p=0.44).Pearsons or Spearmans?pearsons = normal distribtuions spearmans = not normal distributionsR^2 equation(SSY - SSE)/SSYwhich test is used to find if there is a difference in mean FEV among males and femalesproc ttest DATA = A; class sex; var FEV; RUN; or proc regwhich test is used to find if there is a difference in mean FEV among males and females using proc regcreate a dummy variable data b; set b; if sex = 'M' then female = 0; else if sex='F' then female = 1; run; then use proc reg model FEV=female IF b=0 then fail to reject the null hypothesis