Modules 10-23 Onward Flashcards

Question

What is sampling bias?

Answer 1

Error that occurs due to using a nonrandom sample

Answer 2

A flaw in measuring exposure or outcome data that results in different quality or accuracy of information between comparison groups e.g. misclassification bias, recall bias, reporting bias, interviewer bias, observer bias, measurement bias

Answer 3

with bias, associations are invalid and purely error with confounding, associations are real but not causal

Answer 4

a variable which interferes with the search for causal association between exposure and outcome; masks relationship Is related both to exposure and outcome

Answer 5

“A mediating variable is a variable that explains the relationship between the independent variable and the dependent variable. So, in your example of race and knife crime, poverty being the mediating variable means that poverty correlates with both race and knife crime, and it is a possible reason why there is a link between them. A confounding variable is a variable that might account for an observation other than what is hypothesized. The example used in my research methods class was that both ice cream sales and drownings increase during the summer months. One might hypothesize that eating ice cream causes people to drown, or that more frequent drownings cause an increase in ice cream sales, but it's possible that neither of those are the true cause, there could be a third confounding variable that causes both--people eat more ice cream and go swimming more often during the summer because temperatures are warmer. Confounding variables are things you must think about when you're forming a hypothesis. They're alternative explanations for whatever you think you are observing. Stress in LGBTQIA+ people with mental health issues would be a mediating variable.”

Answer 6

randomization restricting matching standardization stratification multivariate analysis

Answer 7

Epidemiologists apply the term “effect modification” to indicate that the effect of one variable on another varies across strata of a third. There are many different measures of effect and, thus many different measures by which a variable may be an effect modifier for the relationship between a cause and an effect.

Answer 8

multivariate analysis multiple logistic regression & adjusted odds ratios

Answer 9

informed consent confidentiality respect for human rights scientific integrity

Answer 10

The Tuskegee experiment Milgram Experiment Stanford Prison Experiment Willowbrook Hepatitis Experiment

Answer 11

respect for persons beneficence justice

Answer 12

The balance between risk and benefit (maximize possible benefits and minimize possible harm)

Answer 13

person / place / time assessing the health status of a population / generating hypotheses about causal factors planning and evaluating public health programs

Answer 14

Infants: congenital malformations, low birth weight, SIDs Ages 1-14: unintentional injuries, cancer, congenital anomalies, homicide Ages 15 to 24: unintentional injuries, mainly motor vehicle accidents Ages 25 to 44: unintentional injuries, cancer, heart disease Ages 45 to 64: cancer, heart disease, unintentional injury Ages 65+: heart disease, cancer, cerebrovascular disease

Answer 15

The examination of asymptomatic people in order to classify them as likely or unlikely to have the disease.

Answer 16

While screening occurs in individuals who are asymptomatic, or not suspected of having the disease in question, Diagnosis is used to confirm whether someone actually has a disease.

Answer 17

Refers to the proportion of individuals with the target disease who have a positive test result

Answer 18

Refers to the proportion of individuals with the target disease who have a negative test result

Answer 19

The positive predictive value is the proportion of individuals who test positive who actually have the disease, a/(a+b).

Answer 20

The negative predictive value is the proportion who test negative who don’t have disease d/(c+d).

Answer 21

detection bias lead time bias length bias

Answer 22

Detection of a disease precursor by screening in someone who would not ultimately develop symptoms or die from the disease leads to detection bias, or pseudodisease.

Answer 23

In words: kappa is the quotient of the observed proportion of agreement beyond chance and the maximal proportion of agreement beyond chance. A kappa of 0.5 indicates a moderate level of agreement.

Answer 24

predisposing (create susceptibility) enabling (environmental conditions) precipitating (specific or noxious agent, e.g. pollen and asthma attack) reinforcing (aggravate already established disease)