q00 |
I have complied with the Academic
Ethics Statement (type 0 or 1)
|
q01a |
What were the mean bilirubin values
(round to two decimal places)
|
q01b |
What were the median bilirubin values
(round to two decimal places)
|
q02 |
Which variable had the greater number of outliers
(type 1 or 2)
|
q03 |
What form of model between mean log
bilirubin and age is suggested by the
logbil vs ageyr ksm plot (type 1, 2, or 3)
|
q04 |
What is Pearson's correlation coefficient
between age and log bilirubin (round
to two decimal places)
|
q05 |
Within what approximate 95% limits
can log bilirubin be forecasted (round
to two decimal places)
|
|
|
q06 |
What is the estimated change in mean
log bilirubin per 10 years of age (round
to two decimal places)
|
q07 |
In the MLR model for log bilirubin
with ageyr, histologic stage, and sex as
predictors, did the studentized residuals
appear normally distributed on the q-q plot
(type 0 or 1)
|
q08 |
The residuals for the MLR model are
(type 1 or 2)
|
q09 |
The Shapiro-Wilk test for normality of
residuals shows (type 1 or 2)
|
q10 |
The case number with the highest leverage
in the MLR model fit was
|
|
|
q11 |
The largest studentized residual in the
MLR was (round to two decimal places)
|
q12 |
How many cases had extreme influence
(dfits) on MLR fit
|
q13 |
Did the case with the most influence also
have the most extreme residual
(type 0 or 1)
|
q14 |
Compared to the original MLR, did R2
increase when the influential cases were
excluded (type 0 or 1)
|
q15 |
The variance inflation factors from the
MLR show that the predictor with the
highest multiple correlation with the other
predictors was
|
|
|
q16 |
Forward and backward variable
selection procedures yielded
(type 1 or 2)
|
q17 |
Downweighting influential points
using resistant regression (rreg)
gave a P-value for ageyr of
(round to two decimal places)
|
q18 |
What were the 95% confidence limits on
ageyr using the bias corrected method
(round to two decimal places)
to
|
q19 |
In the MLR model with age centered
at the mean age (part o), what is the
interpretation of the constant term (short answer)
|
q20 |
Slope > age 60
(round to two decimal places)
|
|
|
q21 |
Slope ≤ age 60
(round to two decimal places)
|
q22 |
Difference in slopes, (a)-(b)
(round to two decimal places)
|
q23 |
What is the P-value for H0:Sex does
not modify the effect of age on mean
log bilirubin, controlling for disease stage
(round to two decimal places)
|
q24 |
In the MLR model Part 15, what is
the estimated age slope for males
(round to two decimal places)
|
q25 |
In the MLR model Part 15, what is
the estimated age slope for females
(round to two decimal places)
|
|
|
q26 |
What are the odds of stage 4 disease
for a female of average age and
average log bilirubin
(round to two decimal places)
|
q27 |
By how much is the odds of
stage 4 disease multiplied for a patient
whose bilirubin value is 2 log units higher
(round to two decimal places)
|
q28 |
What is the probability of stage 4
disease for a female of average age
and average log bilirubin
(round to two decimal places)
|
q29 |
What is the relative risk of stage 4
disease comparing otherwise
similar females vs males
(round to two decimal places)
|
q30 |
Does the binary logistic regression
model of stage 4 disease fit the data
(type 0 or 1)
|
q31 |
What type of goodness of fit test
was used (type 1, 2, or 3)
|
q32 |
Does the quadratic term for centered
log bilirubin significantly (p < 0.05)
improve the model fit (type 0 or 1)
|
q33 |
Was the goodness of fit improved
by adding the quadratic term for
centered log bilirubin (type 0 or 1)
|
q34 |
Does the odds of stage 4 disease for
females vs males vary significantly
by age (type 0 or 1)
|
q35 |
What are the odds of stage 4 disease
for a female of average age and average
log bilirubin (round to two decimal places)
|
q36 |
What are the relative cumulative
odds of a higher stage
(round to two decimal places)
|
q37 |
Is the proportional odds assumption
significantly rejected by the Brant
test (type 0 or 1)
|
q38 |
What are the conditional odds of
stage 2 vs stage 1 disease comparing
a female vs a male who both have
average age and average log bilirubin
(round to two decimal places)
|
q39 |
What is the estimate derived from the
multinomial model of the conditional
odds of stage 3 vs stage 4 disease
comparing a female with a male of
average age and average log bilirubin
(round to two decimal places)
|
q40 |
Does the multinomial model fit the data
(type 0 or 1)
|
q41 |
Which variable has the most siginificant
relationship to disease stage (type 1, 2, or 3)
|
q42 |
What is the probability under the null
hypothesis that none of the predictors
effect the odds of a seizure free 2 week
period (round to two decimal places)
|
q43 |
What is the upper 95% confidence
limit for the odds of a seizure free 2 week
period, comparing active treatment to
placebo, adjusted for the baseline
covariates (round to two decimals)
|
q44 |
According to the Hosmer-Lemeshow
goodness of fit test, does the
longitudinal binary logistic model
fit the data (type 0 or 1)
|
q45 |
What method was used to correct for the
within patient correlation in repeated
2-week seizure free periods (type 1 or 2)
|
|
|