Making sense of psychometrics to post

Measurement involves interaction with a concrete system to represent aspects of it in abstract terms (e.g. numbers, vectors).
Characterization by Eran Tal, Stanford Encyclopedia of Philosophy highlights that measurement is an interaction with a system, which can also include perceptual and linguistic activities.
Measurement can involve ideal systems in addition to real systems.

In 1946, S.S. Stevens discussed the formal properties of measurement scales.
The topic was debated by the British Association for the Advancement of Science regarding the possibility of quantitative estimates of sensory events.
Conflicting ideas arose about the definition of measurement.
The committee’s final report in 1940 aimed to clarify the assignment of numerals to objects and their corresponding rules, noting different types of scales.

Nominal:
- Classifies data into distinct categories.
- Example: Employed vs. Unemployed.
Ordinal:
- Establishes a rank order.
- Example: Marathon rankings based on performance, does not imply equal spacing.
Interval:
- Measures quantities with arbitrary zero and assumes equal intervals.
- Common example: Temperature.
Ratio:
- Measures quantities with an absolute zero.
- Allows for additive and multiplicative operations, e.g., reaction time.

Psychometrics focuses on measuring psychological attributes logically dependent on cognition (e.g., personality or emotions).
Psychological attributes are not directly observable, so behaviors are used to infer these attributes.
The relationship between observed behavior and the underlying psychological construct is crucial.

Different scales create rules for number assignment in research.
For instance:
- Observing categories (Nominal).
- Observing ranks (Ordinal).
- Observing measured quantities (Interval and Ratio).

Operational Definitions: Specify how psychological constructs are measured.
Construct Definitions: Explain variables that are not directly observable.
Psychometrics connects theoretical properties with observable outcomes (e.g., test results).

Scales of measurement defined by their empirical operations and how they enable the assignment of numerals based on defined rules.
Each scale type has a specific statistical operation permitted (e.g., mean for interval, median for ordinal).

Example scenario for measuring lateness in school/work:
- Types of scales to operationalize lateness can be Nominal, Ordinal, Interval, or Ratio.