Biostats - Exam 1 Material

Design of Experiment (DOE)

ethical design, implementation and dissemination of statistical data; design in order to eliminate as many uncontrolled variables as possible. (Vax cause autism study is an example of non ethical)

Quantitative Variables

Random data that cannot be predicted, purely by chance (ex, adult height)

Discrete Quantitative Data

Data that has gaps in the values it can assume, usually whole numbers (i.e, hospital cannot have 10.5 patients, only 10 or 11.)

Continuous Quantitative Data

Data that has no gaps in the assumed values (ex: height, weight)

Nominal Qualitative Data

Categorical (not number) values that have no inherent order; i.e, the number assigned to that category has no meaningful relation (male = 1, female = 2, its just assigning a number)

Ordinal Qualitative Data

Categorical data that DOES have an inherent order, i.e, the number assigned to each category has meaning (ex: stage of disease 1, 2, 3, 4

Population Data

A set of data that includes every single member of the group you are studying (i.e, all women in america)

Sample Data

Data collected from a smaller group within the population, usually randomly selected (i.e, 10,000 randomly selected women out of all women in america).

Stratified Random Sampling

a method for sampling a population that involves sorting subjects into groups “strata” and randomly choosing subjects from each group (i.e, group by grade, select subjects randomly from each grade)

Systematic Sampling

Picking a subject and sampling values from that point (x, x+k, x+2k)

Descriptive Measure of Tendancy

Summarize the data with one number (mean, median and mode). Conveys the typical value in a data set.

Difference between Histogram and Bar Chart

Histogram (bars touch, numerical data); Bar Chart (bars don’t touch, categorical data)

Frequency

Number of times a value appears in a set

Cumulative Frequency

Sum of frequencies per class (add them up as you go down the row of frequencies)

Relative Frequency

frequency divided by total number of observations (hint, always decimal)

Kurtosis

sharpness of a peak in frequency distribution (positive is more steep, negative is less)

Variance

describes the dispersion of the data relative to their mean (average), i.e, the difference between each value and the average value.