Biostatistics Flashcards
Independent Variable
Stimuli that researchers
manipulate to create
effect
Descriptive Statistics
Refers to the differmethods applied summarize and present din a form to make them eato analyze and interpretusing methods of: • Tabulation • graphical representation • summary measures
Inferential Statistics
Methods involved in order t make generalizations and conclusions about a target population, based on result from a sample, includes: • estimation of parameters • testing of hypothesis
TYPES OF DATA
Quantitative
Qualitative
Nominal
Naming or categoric variables that are not based
on measurement scales or rank order.
• # or symbols are assigned. Lowest form of
variable: (e.g., Gender, Color, Province, occupation,skin color and blood group)
• Dichotomous (binary)- which has only two
levels (e.g., Yes or No, Normal and Abnormal, Male
and Female)
Ordinal
• Arranged in rank ordered categories
• (E.g., Social class, Likert scale, Satisfactory scale,
agree to disagree, murmur range, level of edema)
. ACCURACY/ACCURATE
o The closeness of a measured orvalue.
o Trueness of test measurements
PRECISION/PRECISE
how close measurements of the same item are to each other
Sensitivity
is the test’s ability to correctly designate a subject
with the disease as positive
Specificity
Is the test’s ability to correctly designate a subject
without the disease as negative
Positive Predictive Value
o (PPV) is the probability that a subject with a positive
(abnormal) test actually has the disease
Negative Predictive Value
o (NPV) is the post-test probability that the subject has no disease given a negative test result
Likelihood ratio
It is defined as the probability of a subject who has the disease testing positive divided by the probability of a subject who does not have the disease testing positive
SELF-SELECTION BIAS
o people presenting for screening tend to be healthier leading to
false sense of better outcomes
LEAD TIME BIAS
o refers to the phenomenon where early diagnosis of a disease
falsely makes it look like people are surviving longer. This
occurs most frequently in the context of screening.
LENGTH BIAS
o refers to the fact that screening is more likely to pick up slower-growing, less aggressive cancers, which can exist in the body longer than fast-growing cancers before symptoms develop
OVER-DIAGNOSIS BIAS
o An extreme example of length bias
o aggressive search for abnormalities might actually lead to
harm and great cost without reaping any benefits
STRATIFIED
SAMPLING
• Sampling method where we divide the population into nonoverlapping subpopulations or strata, and then select one sample from each stratum • The sample consist of all the samples in the different strata
SYSTEMATIC
SAMPLING
DESIGN
• Selection of the first element is at random
and selection of the other elements is
subsequently taking every k
• Sampling interval is represented by k
• kth element of the population is chosen
(k=N/n, where N is the total population, and
n is the sample size needed
CLUSTER
SAMPLING
• The population is first divided into sampling
units called clusters
A sample of clusters is selected
• Every element found in each cluster is
included in the study
MULTISTAGE
SAMPLING
DESIGN
• There is hierarchical configuration of sampling units and we select sample of these units in stages • The population is 1st divided into a set of primary or first stage sampling units • Each primary sampling unit included in the sample is further subdivided into secondary or second stage sampling units, from which a sample will again be taken.
MEAN,
Most commonmeasure of central tendency; “average”
MEDIAN,
• The value that falls in the middle position when the observations are ranked in order from the smallest to the largest. • If number of observations is odd, the median is the middle number • If it is even, the median is the average of the 2 middle numbers. • Useful on skewed data • For ordinal or numeric data if skewed
&
MODE
• The value that occurs with the greatest frequency in a set of observations • Used in public health statistics (top 10 mortality and morbidity) • For bimodal distribution
TYPE I (Α) ERROR LEVEL OF SIGNIFICANCE (α)
Error of rejecting the null hypothesis when it is really true Declaring a difference when none exists. Similar to false positive test
TYPE II (Β) ERROR LEVEL OF SIGNIFICANCE (α)
Error of NOT rejecting the null hypothesis when it is actually false Failing to declare a difference that does exist. Similar to false negative test
Rejecting or Accepting a Hypothesis
Low P
p
Value of sample results are far from the population parameters Unlikely events REJECT HO
High P
p>α
Value of sample results are close to population parameters Likely events DO NOT REJECT H O
. EXPANSIVE
used to describe populations that are young and growing
characterized by their typical ‘pyramid’ shape
has a broad base and narrow top
show a larger percentage of the population in the younger age
cohorts
typically representative of developing nations, whose
populations often have high fertility rates and lower than
average life expectancies
CONSTRICTIVE
used to describe populations that are elderly and shrinking
often look like beehives and typically have an inverted shape
with the graph tapering in at the bottom
have smaller percentages of people in the younger age cohorts
and are typically characteristic of countries with higher levels of
social and economic development
Base that is narrower than middle of the pyramid, usually the
result of a recent rapid decline in fertility
STATIONARY
• Narrow base and a roughly equal numbers in each age group,
tapering off at the older ages, indicating a moderate proportion
of children and a slow or zero rate of growth
• used to describe populations that are not growing
• characterized by their rectangular shape, displaying somewhat
equal percentages across age cohorts that taper off toward the
top
• characteristic of developed nations, where birth rates are low
and overall quality of life is high
SWAROOP’S INDEX:
Number of deaths among those
50 years and older in a calendar
year
CASE FATALITY RATE:
Measures killing power of disease
High CFR means a more fatal disease.
A higher CFR is expected from a hospital statistics than from
the community
Case-control (often
called retrospective)
• Define diseased subjects (cases) and non-
diseased subjects (controls); compare
proportion of cases with exposure (risk
factor) with proportion of controls with
exposure (risk factor)
Cohort (usually
prospective;
occasionally
retrospective)
• In study population, define exposed group (with risk factor) and nonexposed group (without risk factor) • Over time, compare proportion of exposed group with outcome (disease) with proportion of nonexposed group with outcome (disease)
Cross-sectional
• In study population, concurrently measure outcome (disease) and risk factor • Compare proportion of diseased group with risk factor with proportion of non- diseased group with risk factor
Clinical trial
(experiment)
• In study population, assign (randomly) subjects to receive treatment or receive no treatment • Compare rate of outcome (e.g., disease cure) between treatment and nontreatment groups
Meta-analysis
• Collates data from multiple independent
studies to maximize precision and power
in testing for statistical significance