Factors Affecting Measurement

Level of Measurement

  • Nominal: Name only, just words - numbers have no ranking, weight, meaning.

    • Gender
    • Yes or No
    • Ethnicity (African American, Native American, Asian American, Caucasian… etc)
  • Ordinal: Ranking or weighting of responses, but not real mathematical properties

    • Attitude surveys (Strongly Agree | Agree | Disagree | Strongly Disagree)
    • Questions that ask for relative answers and are formatted by least to most
  • Interval: Mathematical properties, even spacing between categories & no true zero

  • Ratio: Raw numbers, true zero

    • precise, exact - "How many times did you miss class?"

Statistically Viable, Areas to Explore

Reliability

  • Reliability focuses on consistency
  • The scale being off, so appears to be valid

Validity

  • Validity focuses on the accuracy of the statistics
  • Having the scale be the correct number
  • Does the measure the measure?

Methods of Data Collection

Qualitative: More ideographic method of data collections

  • Open ended questions & answers
  • Depth of human experience
  • Instead of stats we analyze themes & patters of quotes
  • More in respondent's own words
  • Focus groups or ethnographic interview
  • Inductive (not testing a hypotheses)

Quantitative: A non-mothetic & deductive method of data collection

  • Statistical & closed ended
  • Helps make generalizations / describes large groups
  • Superficial - not in-depth