Choosing appropriate statistical test

• How to choose appropriate statistical test?

  1. Difference or correlation
  2. In case of difference, what experimental design
  3. Level of measurement.
• What designs are related?

  Matched pairs and repeated measures.
• What designs are unrelated?

  Independent groups
• Analogy for statistical tests table

  Can’t, sing, cant, meow, will, say, unrelated, rant, please
• Level of measurements

  • Nominal data
  • Ordinal data
  • Interval data
• Nominal data

  • Data represented in the forms of categories.
  • E.g asking people if they like psychology?
  • One category= yes
  • One category= no
  • Data is discrete, can only appear in one category
• Ordinal data

  • Data is ordered in some way.
  • E.g asking everyone on a scale of 1-10 how much they like psychology
  • Does not have to follow equal intervals.
  • E.g someone who rated psychology as 8 loves psychology twice as much as someone who rated psychology 4.
  • Lacks precision because it is based on subjective opinion rather than objective measures.
• Interval data

  • Based on numerical scales that include equal units in size.
  • More detail is preserved in interval rather than ordinal.
  • E.g data from weighing scale, and thermometer.
• What measure of central tendency is used for each level of measurement?

  • Nominal: Mode
  • Ordinal: Median
  • Interval: Mean
• What measures of dispersion are used for the levels of measurement?

  • Ordinal: Range
  • Interval: Standard deviation