Variables and Measurement in Research

Types of IVs or Manipulations

  • Presence/Absence (bivalent)

    • You have a level of IV that involves the treatment and a level that does not (control group)

  • Type

    • Each level of IV involves a different type or version of factor being manipulated

  • Amount

    • Each level of IV had a different amount of the treatment

Other Research Designs/Variables

  • Quasi-Experiment

    • Type of research design where a comparison is made, as in an experiment, but no random assignment of participants to groups occurs

    • Quasi-Independent Variable/Subject (attribute) variable

      • Measurable characteristic of participants that cannot be manipulated (gender, height, weight)

  • Between subjects

    • Different participants in each experimental group

      • Experiences only 1 level of IV

      • Random assignment into conditions

  • Within subjects

    • Same participants in each of the experimental groups

      • Each participant experiences all levels of IV

Measuring Variables

  • Qualitative: Nonnumerical participant responses

  • Quantitative: Numerical data

  • Measurement: Assigning numbers to indicate the level of a variable

  • Scaling: Specifies the relationship between the measured variable and the conceptual variable

Nominal/Categorical Scale

  • Non-ordered category responses, organizing your data into groups

  • Scores represent a particular characteristic/category, but have no actual value

    • Gender, hair color, place of birth, political affiliation, diagnostic category, ethnicity

  • Does not address direction or magnitude of difference between groups/values

Ordinal Scale

  • Ordered category responses (ranking groups)

  • Scores indicate whether there is more or less of the variable but not how much

    • Taste test, order of finish in a race

  • Ranking addresses the direction of difference (middle value vs highest and lowest) but not the magnitude of difference between values

Interval Scale

  • Ranked with equal distances between scores corresponding to equal size changes

    • Temperature scale, IQ scores

  • Addresses both the direction of difference (rank orderings, like an ordinal scale) and the magnitude of difference between values

    • The difference between 50/60 degrees is the same as the difference between 60/70 degrees

    • But no true “0” 0 degrees Fahrenheit isn’t a lack of temperature. Also scores are not ratio of each other

Ratio Scale

  • Measures differences between groups and the absence of the variable, scores are ratios of each other

  • Values are anchored by a non-arbitrary zero point: can be measured as having no value

    • Kelvin scale, reaction time, salary, distance

  • Called ratio because dividing one point on the scale by another gives a meaningful value

Discrete vs. Continuous Variables

  • Discrete Variables (nominal/categorical, ordinal date)

    • Put into groups or groups in ranked order (whole numbers)

      • Student’s major, enrollment status, gender, ethnicity, etc

  • Continuous Variables (interval, ratio date)

    • Each measurement gets a distinct score

      • GPA, test scores, reaction time

  • The distinction?

    • Affects whether a particular statistic can be used