True or false:

a) In a regression of y on x, if the p-value for the intercept is 0.01, then there is a statistically significant association between x and y.

b) Rank tests should not be used if data are not normally distributed.

c) The standard error helps us determine how far off an estimator might be from what it is trying to estimate.

d)Anoutliertends tocausemoreproblemsinaranktestthaninat-test.

e) An F-test should not be used to compare more than 4 groups

f)ifthesensitivityofascreeningtestis95%andthediseaseisveryrare,thentheprobabilityyouhavethedisease,giventhatyougotapositivetest,willalsobecloseto95%.

g)Aconfidenceintervalforasinglemeangivesarangeofplausiblevaluesforthe samplemean.

h)In logistic regression of y on x, the regression coefficient is the odds ratio for a 1 unit increase in x.

i) In survival analysis, censored observations should be eliminated before any analyses are conducted.

j) Ifa95%confidenceintervalforadifferenceinmeanbloodpressurebetweentreatmentandplacebogroupsis(-3.68,-0.93),thena2-tailedtestatalpha=.05wouldrejectthenullhypothesisthatthegroupsarethesame.