Lecture 2+3-Descriptive Statistics and Data Presentation Flashcards
what are descriptive stats
- first step in analysis
- organize/summarize set of measurments
- id general features and trends in a set of data
- convey info a/b a group of study subjects to an audience
what are the four types of quantitative data [DONC]
- nominal (least comlex)
- ordinal
- discrete
- continuous
Define nominal data
- measurements can be placed into unordered categories or classes
- numbers can rep categories but the number is just a label
- order AND magnitude are UNIMPORTANT
Describe a type of nominal data
-if there are only two distinct categories the measurements are called: binary or dichotomous
give an example of binary or dichotomous
mortality status: 1 dead 0 alive
diagnosis of lung cancer: 1 yes 0 no
(numbers can be switched up, 1 can mean either dead or alive)
What are some examples of nominal data
- political party
- type of medial insurance
- race or ethnicity
- cause of death
Define OR-dinal data
- OR-dered categories
- ORDER is meaningful
- magnitude of numerical values still unimportant
- difference between consecutive categories is not always the same
What is an example of OR-dinal data?
severity of cardiac symptoms:
- asymptomatic
- mild symptoms
- moderate symptoms
- severe symptoms
- difference between 1 and 2 is not necessarily the same difference between 4 and 5
ranks: like ordering the states in the US based on average household income in 2016
What is the Likert scale
- an example or OR-dinal data
- used in many surveys: strongly agree, agree, neither agree nor disagree, disagree, strongly disagree
What is categorical data
-nominal and OR-dinal measurements TOGETHER
More about categorical data
- numbers used to represent categories are arbitrary (just labels, distance between them means nothing)
- most math operations (like taking average) do not make sense
Define Discrete Data
- step up from categorical data
- measured quantities that take on MEANINGFUL numerical values, but are restricted to integers or counts
- BOTH the order and the magnitude of the numbers are IMPORTANT
- **differences between values mean the same thing
- numbers are not merely labels, math rules can be applied
What are some examples of discrete data
- number of medication errors in a hospital over a one week period (diff b/w 1 and 2 errors is the same as 7 and 8; cant have 2.3 errors)
- number of homicides in MA in 2017
- number of cases of Zika virus reported in Miami in July 2017
- Number of times a female has given birth
- number of new cases of diabetes diagnosed in Canada over a one year period
Define continuous data?
- measured quantities that are not restricted to specific numerical values
- fractional values (and decimal places) are possible-but not required