Biostat Methods I - Descriptive Statistics and Visualization

Descriptive Statistics

Methods for summarizing and communicating data, functions of particular variables

Inferential Statistics

Techniques and tools for making inference about a population from a sample

Measures of Central Tendency

Statistics for quantitative variables, describe ‘middle’ of a sample, mean, median, mode

Weighted Averages/Mean

Sum of the product of each x and its weight, w, where all ws sum to 1.

Measures of Spread

Describe how variable a particular data set is, range, IQR, variance, standard deviation

Sample Variance (uncorrected)

Sum of the squared differences between each x and xbar, all divided by n. Biased estimator.

Sample Variance (corrected)

Sum of the squared differences between each x and xbar, all divided by n-1. Accounts for extra variability associated with the calculation of xbar. Unbiased estimator.

Sample Standard Deviation

Square root of sample variance

Frequencies

For qualitative data, how many times a particular value is observed

Stratification

Considering a data set in a group-wise fashion

Correlation

Measures the strength of the linear association between two variables on a scale between -1 and 1

Histograms

Used for summarizing a single quantitative variable, choice of bin size is important, frequency or proportions can be shown

Boxplots

Used for quantitative variables, multiple plots can be made to compare groups

Bar Charts

Used for showing the value of a variable over the levels of a qualitative variable

Pie Chart

Evil

Scatterplots

Cartesian point plot used for visualizing the relationship between two quantitative variables