Biostat 310
Discussion Section Week 3
April 10-15, 2017
Biostat 310 Discussion Section Week 3
April 10-15, 2017
1/1
Screening: Review of Definitions
For a test to detect some disease (in some population), we define.
Sensitivity
Specificity
Positive predic
Biostat 310
Spring 2017
Homework 2 - Solutions
Total Points: 15 points
There is a total of 10 graded points (questions indicated by *) and then an additional 5 points
in total for the additional questions without an asterisk, graded according to:
Biostat 310
Spring 2017
Homework 2
Total Points: 15 points
There is a total of 10 graded points (questions indicated by *) and then an additional 5 points
in total for the additional questions without an asterisk, graded according to: 0= no additi
Data Types & Summarizing Data
(with simple numerical summaries and graphs)
Types of Data
Categorical nominal, ordinal
Quantitative continuous, discrete (counts)
Graphical Methods
Pie graph
Bar chart
Stem and Leaf
Histogram
Boxplot (and side-by
Biostatistics for the
Health Sciences
Biostatistics 310
Spring 2017
Biostat 310
1
Instructor
Lloyd Mancl, PhD (I prefer to be called Lloyd)
Research Associate Professor, Oral Health Sciences
Adjunct in Biostatistics & Pediatric Dentistry
University of
Biostat 310
Spring 2017
Homework 1 - Solutions
Total Points: 15 points
There is a total of 10 graded points (questions indicated by *) and then an additional 5 points
in total for the additional questions without an asterisk, graded according to:
Body Fat Measurement
Winter 2016
Biostat 310
1
Correlation
80
100
120
Hip circumference (cm)
140
160
A scatterplot can be used to display the relationship between
two quantitative measures
The correlation coefficient is used to measure the strength of
a
Review: Hypothesis Testing Steps
1.
Specify the significance level of the test,
2.
State the scientific question
3.
Convert the scientific question into statistical hypotheses (null and
alternative) appropriate for the study design
4.
Choose a reasonable
Prevention of Mother to Child
Transmission of HIV
Winter 2016
Biostat 310
1
HIV and PMTCT
An (untreated) HIV-infected pregnant woman has a
20-40% chance of infecting her infant in utero or
during birth
This gave rise to a generation of HIV-infected babi
HIV and Circumcision
Winter 2016
Biostat 310
0
Winter 2016
Biostat 310
1
Winter 2016
Biostat 310
2
Case Reports
Describe characteristics of a few individual patients
Often, the first description of the disease or phenomenon
Weak study design
No contro
Rachel Lee
Student ID: 1331935
BIOST 310 AA
Homework 2
1. a) Prevalence of HIV+ = 61/475 = 0.1284 or 12.84%
Prevalence of IVDU = 136/475 = 0.2863 or 28.63%
b) Prevalence of HIV+ without IVDU = 27/88 = 0.3068 or 30.68%
Prevalence of HIV+ with IVDU = 61/88
Biostatistics for the
Health Sciences
(Biostatistics 310)
Instructor: Lyndia Brumback
Cartoons and images in these notes are from
Gonick L. Cartoon Guide to Statistics. HarperPerennial, New York, 1993.
Fisher L and vanBelle G. Biostatistics: A Methodology
Rachel Lee
ID #: 1331935
BIOST 310 AA
Homework #4
1.
a) 40/318 = 0.13
b) Left: 10/48 = 0.21, Right: 30/270 = 0.11
These two probabilities are different.
c) 278/318 = 0.87
d) 30/220 = 0.14
e) 18/48 = 0.375
2.
a) sensitivity = 180/200 = 0.9
b) specificity =
Did Mendel Fake his Data?
Ronald A Fisher
Gregor Mendel
Winter 2016
Biostat 310
1
The Controversy
In 1866 Gregor Mendel wrote a paper reporting results
from experiments crossing various strains of peas
This work, although not appreciated at the time, is
Rachel Lee
Student ID: 1331935
BIOST 310 AA
Homework 1
1) Fiebig stage would be considered an ordinal categorical variable because under
the category of the Fiebig stage, the numbers I VI have a meaningful order that
signify the progression of HIV.
2) Med
Rachel Lee (131935)
BIOST 310 AA
Homework 6
1) a.
(i) statistic
(ii) 0.1 2 * (0.1*0.9)/1687) = (0.0146, 0.0853)
b.
(i) Relative Risk
(ii) Chi-square test
(iii) H0: Medication does not have an adverse response
HA: Medication does has an adverse response
2)
Rachel Lee
Student ID: 1331935
January 26, 2016
Homework #3
1. * Answer: b) A cohort study
2. * Answer: b) The comparison of outcome between treatment and control groups
is likely confounded by other differences between the groups
3. * (i) experimental &
Rachel Lee
BIOST 310 AA
Homework 7
1. Reject the null because p<0.05
2.
a) two sample unpaired t-test
b) 0.32
c) Do not reject H0 and conclude there is no strong evidence that particulate
levels are different in urban vs rural areas
3. b. Use a very small
Review for BIOST 310 Final
1
Topics covered since Midterm II
CLT; Sampling Distribution and Confidence
Interval for
Hypothesis testing (H0 and HA, , Type I and
Type II error, P-value, power, sample size)
One sample test for proportion (binomial or Z
t
Practice Problems:
1. A study enrolled 240 non-smokers with lung cancer and 260 non-smoking controls. The
investigators then determined whether each participant had been exposed to the cigarette
smoke of a spouse for at least 6 months ("exposed to passive
Biostat 310
Homework 5
Submit to the CANVAS course website
DUE:11/06/15, 11:30am
[14 points]
There are a total of 12 graded points (questions indicated by *) and then an additional 2 points in total for the
additional questions without an asterisk, graded
Cin Yu Hsieh
Biostatistics 310
Term Paper
Your Hardcore Run May Be as Dangerous as Sitting Around Doing Nothing, is a news report
aiming to summarize the potential findings of a longitudinal cohort study on the association
between different level of joggi
Biostat 310
Homework 8
Submit to the CANVAS course website
DUE: 12/11/15, 11:30am
[14 points]
There are a total of 12 graded points (questions indicated by *) and then an additional 2 points in total for the
additional questions without an asterisk, grade
Biostat 310
Homework 7
Submit to the CANVAS course website
DUE:12/04/15, 11:30am
[15 points]
There are a total of 13 graded points (questions indicated by *) and then an additional 2 points in total for the
additional questions without an asterisk, graded
Biostat 310
Homework 6
Submit to the CANVAS course website
DUE:11/13/15, 11:30am
[15 points]
There are a total of 13 graded points (questions indicated by *) and then an additional 2 points in total for the
additional questions without an asterisk, graded
Biostat 310
Homework 4
Submit to the CANVAS course website
DUE: 10/30/15, 11:30am
[14 points]
There are a total of 12 graded points (questions indicated by *) and then an additional 2 points in total for the
additional questions without an asterisk, grade
Biostat 310
Homework 3
Submit to the CANVAS course website
DUE: 10/23/15, 11:30am
[15 points]
There are a total of 13 graded points (questions indicated by *) and then an additional 2 points in total for the
additional questions without an asterisk, grade
Biostat 310
Homework 2
Submit to the CANVAS course website
DUE: 10/16/15, 11:30am
[15 points]
There are a total of 13 graded points (questions indicated by *) and then an additional 2
points in total for the additional questions without an asterisk, grade