PROTEUS

              

 

 

 

 

 

 

PAGES [BACK to PROTEUS home page]    Back to CKHS100 MAIN PAGE     Outline     Research protocol

SPECIAL TOPICS:   Histograms/Central tendency    Sensitivity/Specificity   Rates   Critical review 

               Evaluation Research   Definitions    Designs  Sample size  Sample methods  Confounding 

 

CKHS-100: MAIN PAGE:   WEEKLY LOG AND PROGRESS

These pages are using an older technology (Microsoft FrontPage). This allows me to add more material very easily, and to edit and adapt as we go along. It also remains running between courses should anyone wish to return to find some notes or slides!

    Free research dictionary! Click on this link to download the entire 5th edition, Dictionary of Epidemiology:  

 

  MAJOR TOPIC  NOTES, content details Materials Powerpoint, media and other links
SESSION

 1

Sept 12

DESIGN: OVERVIEW

 

 

IDENTIFYING VARIABLES AND

PLOTTING AND DISPLAYING DATA

 

 

CENTRAL TENDENCY

 

INTRODUCE RESEARCH DESIGN

 

Course Outline/objectives

 

* Participants' research experience - categorize

      and brief discussion on objectives, scope, process

 

 

1.  Pretest sheet 1 page

 

2.  Outline 2 pages

 

3.  Research Design summary 

 

4. Plotting practice sheet

 

 

 

 

 

 

 

 

 

CASE STUDY: onset curves for analysis

 

 DEFINITIONS

 PP: VARIABLES

 PP: STUDY DESIGN

 PP: WHAT ARE THESE DESIGNS?

 PP: PLOTTING DATA 

Variables, categories (Input/outcome,

       continuous/categorical)

Types of research:

*  Qualitative vs. Quantitative Domains

*  Experimental /intervention studies,

       randomized controlled trials (RTC)

*  Longitudinal observational studies: 

       Prospective Cohort study

       Retrospective Cohort study

       Case-Control stud

* Cross-sectional studies (surveys, one-time obs)

* Adv/Disadvantages of design types 

* Selection criteria, starting "groups" with each

     method; examples.  

Plotting Histograms, onset curves - case studies and interpretation.

METHODS REVIEW=============================

Plotting data, histograms, normal distribution, central tendency: mean, median, mode...)

SESSION

2

Sept 19

CENTRAL TENDENCY (contd)

 

 

 

 

QUIZ#1(10%) Issued

(take home): Central tendency/distributions and selecting controls  

* Sensitivity & specificity

 

* Controlling

            * during selection

            * during analysis

 CASE STUDY:  a CURIOUS OUTBREAK IN NEW YORK

 

 

 

 

 

 

*  Quiz sheet #1 issued

 

 

 

 PP: STD DEVIATION

 

 

 

 PP: CONFIDENCE

        LIMITS

 

 

 

METHODS REVIEW========================

(Standard deviation, confidence limits, "normal distribution", importance and characteristics)

SESSION

3

Sept 26

QUIZ#1 Due back

 

 DEFINING and selecting "CASES" & "CONTROLS"

 

 RATES & RATIOS

 

 WORKING WITH CONTINGENCY TABLES

 

QUIZ#2  Take home: Solving the outbreak using 2x2 tables

METHODS REVIEW============================

Risk Measures (Related Rates/true rel risk, Odds Ratio, attrib risk)

"Analytical" methods using 2x2 tables

 * Chi-Square Test,

 * Fishers Exact Test

 

* Rates, Ratios & Measures

   Incidence/prevalence/duration

   Mortality rates  (Crude MR, Age-specific MR,

        Sex-specific MR, 

        Gender-specific MR,

        Case fatality rate (CFR),

        Proportionate MR,

        Standardized MR

CASE STUDY: Glycols

 

 

 

 

 

 

 

 

* Quiz sheet #2 issued

 MEGAPLATING to show bacterial evolution of resistance

 

 PP:  ODDS RATIO AND REL RISK

 

 PP: CHI-SQUARE

 

 PP: RATES 

SESSION

4

Oct 3

 

 

 

 

DIFFERENCES BETWEEN TWO MEANS

(T-TESTS)

 

 

 

  * Appropriate use of t tests

 

  * Distinction between paired and unpaired data

 

  * Examples and practice from published research

CASE STUDY:

 

 

 

 

 

 

 

t-test practice sheet

 

 

 

PP: T-test (Unpaired)

 

PP t-test (Paired)

 

 

QUANTITATIVE METHODS REVIEW:

(comparing two means: t-tests)

Oct 10  NO CLASS      
SESSION

5

OCT 17

QUIZ#2:  Due.

 

 

     ANOVA

 

 

Further research design considerations.

Experimental and quasi-experimental study designs

 

 

 

 

 

CASE STUDY:

 

Research design: Atopic Eczema

 

 

 

 

 

 

 

 

 
PP. Quasi-Exp.Study designs

Challenge: study of patient-lifting as cohort-design

PP: ANOVA

Herd immunity

QUIZ#3: t-tests and ANOVA   

QUANTITATIVE METHODS REVIEW:  ANOVA)
SESSION

6

OCT 24

QUIZ#3  due 

 

 BIVARIATE ANALYSIS:

 

 

CORRELATION AND REGRESSION

 

 

Case studies: critical review of published research*  

* Case definition

* Sensitivity & Specificity

* Optimal selection of these 'descriptive' methods when designing investigations.

 

 

CASE STUDY:

 

 

 

 

 

 

PP: SENSITIVITY

        & SPECIFICITY 

PP: REGRESSION

pp: CORRELATION

 

SESSION

7

OCT 31

 

 

 

SURVEY RESEARCH:

sample selection

Sample size

 

 

 

 

QUIZ#4  

Sample selection and sample size  

SURVEY RESEARCH METHODS

 

*  SAMPLING METHODS:  Random sampling, Sample unit, Sample frame, Sample fraction, Sample error, Sample interval

      Random error

      Systematic error

*  Sample selection methods for survey research

      Simple random sampling

      Stratified sampling

      Systematic sampling

      Cluster sampling

      Multistage sampling

 

Sample size calculations (1) The 'n' needed for reporting parameters with acceptable confidence (survey research)

 Sample size calculations (2) The 'n' needed to find the effect if it exists: "Power Analysis"

         Use of nomogram for 'power' analysis

 

CASE STUDY:

 

 

 

 

Nomogram sheet for power analysis

 

 

PP:SAMPLING METHODS

 

 

PP: Sample size

 

 

 PP_Nomogram for power analysis

 

 

SESSION

8

NOV 7

Questionnaire design

 

 

 

 

 

 

Question writing, questionnaire design, administering the survey

 

 

 

CASE STUDY: Environmental arsenic poisoning

CASE STUDY: CRITIQUE: Dakota health department outbreak of URI among youth. published in ProMed (Handout)

 

 

 

 

 

 

 

 

 

 

PP: Questionnaire design

 

 

 

SESSION

9

NOV 14

QUIZ#4  Due 

 

HEALTH RISK ASSESSMENT

AND

RISK COMMUNICATION

 

 

CONFOUNDING

QUIZ#5   (confounding)

QUANTITATIVE METHODS REVIEW:

(CONFOUNDING VARIABLES: DETECTION AND CONTROLLING

 

CASE STUDY:

to be announced

 

 

  PP: CONFOUNDING VARIABLES  

SESSION

10

NOV 21

 

(Can submit during the coming week but returns and comments will be first-in, first-out)

 

 

 

 

 

 

 

 

INTRODUCTION TO

   Probabilistic risk assessment (PRA)

 

 

   Quantitative risk assessment (QRA)

 

CASE STUDY

 

 

 

PP: Introduction to use of TREES to calculate health risk

 

PP: Introduction to use of EPA 4-step model for Quantitative Risk Assessment (QRA)

SESSION

11

NOV 28

QUIZ#5    Due May 4

 

 

 

 

Suggested submission date for (written) Research design protocol

* CAUSATION AND BIAS

* MYTHS AND MISUNDERSTANDING

 

..............MENU for today...............

1. From data to published tables: guidelines for constructing useful tables

2. Intro to program evaluation research

3. Biases, myths, and false claims (Including cancer rates)

4. From Koch to Hill: criteria for attributing causation

5. The Baloney detection kit

6. Special look at Wakefield

7. Ethics and 'informed consent'

8. Placebo effects

9. Designing a study from end to begining

10. Seven steps for health-risk communication

* Hill's Criteria and example sheet

 

CASE STUDY: Investigation of keratoconjunctivitis

 

 

 

 

 

* CKHS100 Post-test sheet

PP: From raw data to completed table

PP: CONTROLLING

PP: HILL'S CRITERIA FOR ATTRIBUTING CAUSE 

 

 

PP: Bias & false Claims

 

PP: Evaluative research

 

 

SESSION

12

DEC 5

 

  PRESENTATIONS    

Discussion of Toxic Oil Syndrome

see also reprints of accounts

Here are three perspectives on the Spanish Toxic Oil Syndrome (1981)

 

   Kilbourne et al. NEJM Dec. 8, 1983:1406    Reprint of original investigation report 

   Woffinden The Guardian, at. 25 August, 2001    One of the "conspiracy theory articles .

   Terracini, IJE, Vol.33,issue 3, 443-444 (2004)   An epidemiologist's assessment of the "balance of Evidence"