Lectures 1 & 2

Question 1

What is biostatistics?

Answer 1

involves data collection, analysis, and answering a research question while accounting for uncertainty

Question 2

What is the purpose of statistics?

Answer 2

to use data collected from a sample to apply it to the whole population

Question 3

What are the two types of sampling?

Answer 3

probability and non probability

Question 4

What are the three types of probability samples?

Answer 4

simple (everyone has the same chance of being selected)

systematic (only people that come in w/in a certain time frame is selected)

stratified (population is divided into smaller groups (strata) and then selected from the strata in similar ratios to the population

Question 5

What are the three types of non probability samples?

Answer 5

convenience (whoever’s the closest)

purposive (only a people from a certain group is selected)

snowball (collected via an initial person selected)

Question 6

What is a parameter?

Answer 6

a numerical measure describing a characteristic of the population

Question 7

What is a statistic?

Answer 7

a numerical measure describing a characteristic of the sample

Question 8

What are the four steps in the biostatistics process?

Answer 8

1. qualification of the disease
2. estimation of newly developed cases of the disease
3. determining risk association
4. examine the effects of treatment

Question 9

What occurs in the 1st step of the biostatistics process?

Answer 9

the population/sample size/method is determined
prevalence is calculated
generalizations CANNOT be made (only region/time specific)

Question 10

What occurs in the 2nd step of the biostatistics process?

Answer 10

a follow up with those who were initially controls
observing other factors to determine causation relations

Question 11

What occurs in the 3rd step of the biostatistics process?

Answer 11

relative risk is calculated to determine if there’s a correlation between the exposure and the outcome

Question 12

What occurs in the 4th step of the biostatistics process?

Answer 12

RCTs are used to test the effectiveness of new potential treatment options and are compared to a control

Question 13

What is the incidence equation?

Answer 13

of NEW cases over a period of time / the total population that is AT RISK during the same time period (controls only)
- denominator must include a time period

Question 14

What is the prevalence equation?

Answer 14

of people in the sample with the SAME characteristic / total # of people in the same (both cases AND controls)

Question 15

What are the five types of study designs?

Answer 15

cross-sectional
case-control
case reports/series
cohort
randomized controlled trials

Question 16

Why can’t case reports/series be generalized to the population as a whole?

Answer 16

there is no comparison (control) group making all values absolute

Question 17

What is a cross-sectional study?

Answer 17

occurs at a singular point in time (no follow up), used to determine prevalence (no definitive causality connections can be made)

Question 18

What are the advantages of a cross-sectional study?

Answer 18

• lots of data all at once (many questions can be asked)
• estimates effect of a disease on a population at that time
• produces some baseline data so more detailed studies can be ran
• useful for studying diseases w slower onsets
• cost efficient

Question 19

What are the limitations of a cross-sectional study?

Answer 19

• relative risk cannot be determined (no time line, just data from a single point in time)
• inefficient for rarer diseases/exposures (no time to wait around for the cases to be present due to short time frame)
• unclear sequence of events of participants (no correlations can be made)
• can’t account for time-dependent concerns (no future predicting or generalizations)

Question 20

What is a case-control study?

Answer 20

a retrospective study that starts with the cases and controls and then asks participants about their past exposures to form connections (always equal ratio of cases:controls)

Question 21

What are the two ways of matching controls to cases?

Answer 21

individual matching
group/frequency matching

Question 22

What are the advantages of case-control studies?

Answer 22

• an efficient type of study
• suitable for rarer diseases
• multiple different exposures can be examined for the disease
• the outcome can take longer to develop (more time to trace back)
• reduces cofounding (chances of an additional unknown variable effecting the two primary variables)

Question 23

What are the limitations of case-control studies?

Answer 23

• the impacts of the matched variable cannot be analyzed (always the same number of cases/controls, RR is always 1)
• hard to match all factors between cases/controls
• over matching can occur (cases/controls are too similar)
• requires memory recall of participants (misclassification/selection bias)
• only 1 disease can be examined at a time
• no clear timeline of exposure and outcome

Question 24

What are the three types of bias in case-control studies?

Answer 24

selection (sample chosen is not representative of population)

nondifferential misclass (both cases and controls are misclassifying information at the same rate)- moves towards null (underestimates)

differential misclass (misclassification is occurring in uneven ratios between the cases/controls- can lead to either under or over estimating)

Question 25

What is a cohort study?

Answer 25

starts with the those who have or haven’t been exposed to the risk factors and then can work forwards or backwards to make links to outcomes or risks

Question 26

What are the three types of cohort studies?

Answer 26

prospective (starts in present, follows to future)

historical (starts in present, looks for past data to make correlations- NO participant recall)

historical prospective (starts in present and goes back and follows participants)

Question 27

What are the two types of sampling in cohort studies?

Answer 27

exposure based (starts with groups of EXPOSED and NOT EXPOSED and follows them prospectively to observe the outcome)

population (chooses population BEFORE any are exposed or even exposures have been determined, then follows prospectively to observe and outcome or not)

Question 28

What are the advantages of cohort studies?

Answer 28

• clear timeline present
• risk can be measured
• less susceptible to selection bias (can’t change history and cannot alter the results of the outcome)
• multiple outcomes can be studied all at once

Question 29

What are the limitations of cohort studies?

Answer 29

• expensive (can be long)
• time consuming (many follow ups needed)
• attrition of participants due to duration
• rare outcomes (lots of waiting/unknowns about/for outcomes to occur)

Question 30

What is a randomized controlled trial (RCT)?

Answer 30

a trial of an experimental treatment compared to the current standard/a control to measure the effectiveness (an intervention)

Question 31

What are the three main components of RCTs?

Answer 31

allocation concealment (assignments of groups are unknown until the moment of randomization)

blinding (limiting the knowledge of the groupings to improve validity of trial)

intention to treat analysis (participants are kept in their originally assigned groups regardless if they have stopped the trial or deviated from it)

Question 32

What are five other components of RCTs?

Answer 32

• randomization
• control groups (some ethical issues since they don’t receive the experimental treatment)
• monitoring (ensures safety of all those involved)
• data management
• reporting

Question 33

What is the Quasi-experiments Pseudo-Randomization?

Answer 33

when the “randomization” system has a discreet pattern to it so that as an investigator some selection bias of which participants end up in which group occurs (no longer truly randomized)