Statistical Ethics & Descriptive Statistics
The principle that ensures integrity, transparency, and fairness when reporting data
Statistical ethics
The entire dataset you wish to study
Population
The measure most affected by extreme values
Mean
The difference between the highest and lowest values
Range
The bell-shaped graphical pattern representing a balanced distribution
Normal curve
Studying the overall direction and pattern of change in data over time
Trend analysis
The process of summarizing and presenting data without making predictions or assumptions about causes
Descriptive statistics
Smaller representative portion of the population
Sample
The middle value when all numbers are ordered
Median
The average squared distance of all values from the mean
Variance
The part of a normal curve containing most data points
Center (around the mean)
A long-term, general movement in data
Secular trend
Adjusting chart scales to exaggerate results or trends
Misleading visualizations
A sampling method where each individual has an equal chance of being chosen
Random Sampling
The value that occurs most frequently in a dataset
Mode
The square root of variance
Standard deviation
A tall, narrow normal curve indicates this type of variability
Low variability
Repeating changes that follow seasonal or calendar patterns
Seasonal variation
Maintaining honesty in data so leadership can make accurate, mission-focused decisions
Data integrity
Dividing a population into meaningful subgroups and sampling within each
Stratified sampling
The only central measure that applies to nominal (category) data
Mode
Smaller values of this measure show more consistency
Standard deviation
A flatter, wider-than-normal distribution
Platykurtic curve
Fluctuations that repeat irregularly without a calendar tie
Cyclical variation
The first habit analysts develop to avoid bias in interpretation
Bias awareness
Selecting every nth record or item from an ordered list
Systematic sampling
A dataset where mean, median, and mode are equal
Normal distribution
How far sample means vary from the overall mean of means
Standard error of the mean
The middle horizontal line on a control chart
Center Line
Projecting a line beyond current data to estimate future values
Extrapolation
Reporting unfavorable results honestly and providing context
Ethical reporting
Bias caused by using only easy-to-reach participants
Convenience bias
A large gap between mean and median shows this
Skewdness
A variability measure based on only two data points, making it unreliable
Range
A data point beyond the upper or lower control limit means this
Assignable cause variation
A limitation of mechanical extrapolation due to human judgment
Subjectivity error
Following transparency and integrity so analyses support decision-making objectively
Ethical transparency
The best defense against sampling bias
True random selection
The most stable measure of central tendency when outliers exist
Median
A preferred variability statistic because it uses all data points
Standard deviation
Differentiating random vs. assignable variation helps analysts do this
Decide when to take corrective action
Why this analysis matters: it predicts future workload and resources
Forecasting maintenance needs