Flashcards
Describing Data
Cards in this deck (10)
- 1Median, IQR and centilesDepend on ranking the data
Easily interpreted in real-world terms
Robust to outliers - only cares what percentage of people are above/below a certain value, not how far above/below they are - 2Mean, standard deviationUseful for predicting total values of a group
Particularly useful in relation to the normal distribution - 3Positive skewPeak is towards bottom end
There is a natural lower boundary (often because values can't go below zero)
E.g. salaries - 4Negative skewPeak is towards the top end
There is a natural upper boundary to data values
E.g. life expectancy - 5No skew - symmetricPeak is in the middle
The boundaries are far from the middle of the distribution
E.g. heights (anything normally distributed) - 6Box plotsUsed to compare distributions at a glance
- 7HistogramsShow each value of the entire distribution
- 8Data distributions tell a story about data's originsFor example, the general health of a population at a particular time
- 9In real data sets we need to do data cleaningRemove outliers
Isolate distribution of interest - 10Purpose of statisticsUnderstand the process that generated the data
Determine whether this process was significantly different in different experimental conditions
Information
- Description
- Statistics Flashcards on Describing Data, created by Eleanor Smith on 12/01/2019.
- Created by
- Eleanor Smith
- Views
- 0
- Rating