biostats
The null value of a difference in means is...
0
Test - do women in first year at UA gain more or less weight than the avg US female college freshman?
1 sample t test
Given the equation below, what is expected SBP (Y) for a male who does not exercise? (X=hours of exercise per week). y=159.9 - 6.38x
159.9
A 95% CI for p is (14%, 22%). What is the point estimate?
18%
What is the incidence rate of cancer in the study depicted below? i. Subject #1 develops cancer and is detected at the year one follow-up ii. Subject #2 is cancer free after two years of follow-up iii. Subject #3 is cancer free after two years of follow-up iv. Subject #4 drops out after one year of follow-up v. Subject #5 develops cancer after six months of follow-up
2/6.5
A 95% CI for p is (14%, 22%). What is the margin of error?
4%
measure ht of 3 different speakers
ANOVA
Test - from a simple random sample of school children in AL, MI, and GA, which state has the higher avg BMI?
ANOVA with post hoc if necessary K-W with post hoc if non-normal
The features of a cohort study include:
Ability to assess temporal associations
The F test is determined by dividing mean squares. Which number goes in the numerator?
Between Groups Mean Square
In Framingham Heart Study, we want to assess risk factors for Diabetes. We have the following variables available to us.Outcome =Impaired Glucose (A categorical variable with 3 value options listed below). i. Diabetes (glucose > 126), ii. Impaired Fasting Glucose (glucose 100-125), iii. Normal Glucose Risk Factors i. Sex (dichotomous) ii. Age (continuous) iii. BMI (normal weight, overweight, obese) (categorical) What test would be used to assess whether sex is a RF for Impaired Glucose?
Chi-Square test for independence
What is the best design to assess the effect of a multivitamin on GPA in college freshmen?
Clinical Trial
Suppose we want to study whether individuals over 45 years old are at greater risk of diabetes than those younger than 45. What kind of variable would we want for age?
Dichotomous
T or F: If a variable is skewed then it must be either platykurtic or leptokurtic.
False
In a two factor analysis of variance, the first thing to check is the significance of the main effect?
False - check the overall test to assess whether there is a difference among the cell means
T or F: If a variable is leptokutric then it must be skewed.
False - leptokurtic means skinny tails or steep, not skewed
T or F: A Post Hoc test should be run before an F-test
False - only run a post hoc test if F is significant
Between October and March, which would you expect to have the highest incidence in the Boston Medical Center Emergency Department?
Influenza
If a variable is leptokutric and you were planning on using an ANOVA, which statistic is an appropriate alternative?
Kruskal Wallis (K-W) Test
We can compare two survival curves to test if they are identical or not. The test to do this is called what?
Log Rank Test
The outcome variable for survival analysis is often what form of analysis?
Non-Parametric
If we are doing a randomized control trial to assess the Efficacy of a new drug (compared to a placebo) for asthma in children which one of the following variables should we use to determine sigma?
Number of ER visits since taking the drug or placebo
If you were planning on using a paired t-test, but discovered that your dependent variable was not normally distributed which of the following would you select as an alternative?
Sign test
Given the results from a data analysis below, which of the type of analysis is appropriate? Skewness = .377 (se=.234) Kurtosis = -1.097 (se=.463)
Spearman's Rho
If we have a study that does not have a sufficient level of power, which is harder to interpret?
Where we failed to reject the null hypothesis
A 95% CI for the difference in means is (0.44, 2.96). Is the difference is statistically significant?
Yes - significant difference
The best graphical summary of variability in a continuous variable is
box plot
What kind of variable is race?
categorical
Test - compare 2 unpaired groups if measure is binomial
chi square
test - compare 3 or more groups if measure is binomial
chi square
Test - compare one group to a hypothetical value or ref value; measure is binomial
chi square GOF
Test - how does the percentage of children who are obese in black belt compare to percentage nationally?
chi square GOF
Test - Who has higher % of female smokers - college freshman, sophomores, juniors, or seniors?
chi square test for independence
What is the best design to assess the association between alcohol exposure and GPA in college freshmen?
cohort study
If we measure age in years, what kind of variable is it?
continuous
test - # of coping strategies vs stress level
correlation
T or F: 100 Hundred subjects is good enough for any study.
false
T or F: Confounding is a major issue in retrospective studies.
false
T or F: Survival Analysis is based on knowing when people were born.
false
Test - compare 2 unpaired groups if measure is Gaussian
independent samples t test
The best summary of a typical value for a continuous variable is the
mean
A two sided test for the equality of means produces p=0.20. Reject H0?
no
Given the results from a data analysis below, is this data normally distributed? Skewness = .377 (se=.234) Kurtosis = -1.097 (se=.463)
no
Spearman's Rho
non-parametric test used to measure the strength of association between two variables, where the value r = 1 means a perfect positive correlation and the value r = -1 means a perfect negative correlation
test motivation; measure achievement scores and compare to published norms; mean and stdev published
one sample z test
test - compare 3 or more groups if measure is Gaussian
one way ANOVA
16 rat twins put in Group 1 and Group2; measure and compare wt of cortex
paired t test
test - compare 2 paired groups if measure is Gaussian
paired t test
Test - is there a difference in % body fat from before and after for those who participate in water aerobics classes?
paired t test if normal sign or Wilcoxon signed rank test if non-normal
Test - among college freshman who consumes more alcohol - men or women?
t-test if normal Mann Whitney U test if non-normal
When deciding on how to interpret the outcome from an Oneway analysis of variance which result do you examine first?
the F test
T or F: All other things being equal it will take a larger sample size to detect a 20 lb. weight loss than a 40 lb. weight loss.
true
T or F: In a study on time to recovery from a migraine using medication or relaxation therapy has a point estimate and 95% CI for the difference in means is 11.4 minutes and (10.2 min. to 12.6 min.). The difference is statistically significant.
true
Test - how does the avg amt of fat consumed by member of sigma nu compare to the mean of 93g (s=17) for all US college students?
z score
The null Hypothesis for a one way analysis of variance is...
µ1 = µ2 = µ3 ...= µk (means are equal)