2|VARIABLES

Overview

Making statistics

  • Where do statistics come from?

Measuring things

  • How many books are red

Measuring things

  • How much grain is in a bushel?

Making statistics

In the United States today half of all children (35.6 million) live in a household where a parent or other adult uses tobacco, drinks heavily or uses illicit drugs1

  • Other questions…
    • How many students are smokers?
    • Young people, narcissism, anxiety, depression 2

Measuring things

Constructs & operational definitions

  • Construct: Extroversion
  • Operational definition: Big 5 questions
  • Construct: Intelligence
  • Operational definition: IQ test
  • Construct: Height
  • Operational definition: How far the top of your head is from the floor according to a tape measure

Operationalizing variables

  • Usually more than one way we could measure & record data
  • Result in different types of data, and potentially different applicable analyses
  • How to decide on operational definition?
    • Aspects to consider:
      • Type of variable (discrete / continuous)
      • Scale of measurement (nominal / ordinal / interval / ratio)

Types of variable

  • Discrete
    • Count as whole numbers
    • Separate, indivisible categories
    • No values exist between neighboring categories
    • E.g. number of children/cats/tvs; positive cases; hospital admissions
  • Continuous
    • Can be measured with decimals
    • Has infinite number of possible values
    • Every interval is divisible into infinite number of parts
    • E.g. height, time, temperature

Scales of measurement

Scale Characteristics Examples
Nominal Named categories
No quantitative distinctions
Gender
Eye color
Experimental condition
Ordinal Ordered categories
Indicates direction, but not size of difference
Rank
Clothing sizes
Olympic medals
Interval Ordered categories
Equal intervals between categories
Arbitrary or absent zero point
Temperature (Celcius/Fahrenheit)
Golf scores
Ratio Ordered categories
Equal interval between categories
Absolute zero point
Temperature (Kelvin)
Number of correct answers
Response time

Likert scales

What is your current level of happiness?

  1. A lot less than usual
  2. A little less than usual
  3. About average
  4. A little more than usual
  5. A lot more than usual

Barry, D. (2017). Do not use averages with Likert scale data. https://bookdown.org/Rmadillo/likert/

Populations & samples

Draw sample, make inference

Terminology

  • Populations
    • Population parameters
    • Usually Greek symbols
    • e.g. \(\mu\); \(N\)
    • Inferential statistics
  • Samples
    • Sample statistics
    • Usually letters
    • e.g. \(M\); \(n\)
    • Descriptive statistics

Learning checks

  • A tax form asks people to identify their
    • Age
    • Annual income
    • Number of dependents
    • Social security number
  • For each variable…
    1. Identify the scale of measurement that probably is used
    2. Explain whether the variable is continuous or discrete