• The science concerned with evaluating the attributes of psychological tests
  • Relevant to all aspects of psychology- research and practice
  • There are two major attributes of psychological tests- reliability and validity.

Measurements and good decisions:

  • Psychological measurement plays three critical roles:

o  Decide which test to use in a given situation o Interpret the scores from a test in a given situation o Possibly create your own test for a given situation

  • If you misapply a test or misinterpret scores, you run the risk of harming people (results in lawsuits!)
  • Using a published test score is not a defence, you can still get into trouble

Creating a test:

  • In some cases, the tests that have been published for a particular purpose may not be of good quality
  • rn other occasions, no test may exist that measures the psychological attribute you’re interested in
  • Therefore you may be tasked with the responsibility of creating a test for your own needs- people in research do this often, people in industry often too do it for money.
  • Testing consequences can be a matter of life and death, for example death penalties in America can’t be applied to people who are mentally retarded. So how can this be measured in a reliable and valid manner?

Unobservable psychological attributes:

  • rften the attribute of interest isn’t directly observable- e.g. self-esteem, working memory
  • Psychologists rely on a collection of overt behaviours as indirect indicators of the attribute of interest.
  • g. measure working memory using a variety of verbal and spatial tests

Psychological constructs:

  • An unobservable hypothetical entity that is used to represent a pattern of psychologically related phenomena.
  • The nature of constructs can only be determined through inference of behaviours theorized to be indirect representations of the psychological construct of interest.
  • No single test can establish the existence of constructs; instead researchers rely on observations to support their existence.

The role of theory:

  • When the construct is defined clearly, one can dismiss obviously irrelevant indicators of the construct of interest. E.g. running speed is obviously irrelevant to working memory capacity

Psychological tests:

  • A systematic procedure for comparing the behaviour of two or more people  Three important components:
    • Tests involve behavioural samples of some kind
    • The behavioural samples must be collected in some systematic way o The purpose of tests is to compare behaviours of two or more people

Purpose of psychological testing:

  • Purpose of psychological testing is to compare the cognition/behaviour of different people, known as measuring ‘inter-individual differences’.
  • Also use tests to measure changes in the behaviour of the same people across time, known as ‘intra-individual differences’

Types of tests:

  • Content (e.g. skill, personality, attitude)
  • Types of responses (e.g. multiple-choice, open ended)
  • Administration procedure (e.g. individual vs. group)
  • Intended purpose (criterion vs. norm referenced)
  • Time constraints (speed vs. power)

Criterion referenced texts:

  • Used in contexts where a decision must be made about someone’s skill level
  • Involve a cut off score that specifies whether someone has achieved a sufficient level of the skill or capacity.
  • Typically used in personnel selection contexts (e.g. demonstrate at minimum capacity to type 60 words a minute).

Norm referenced texts:

  • Involve comparing a person’s score with that of a reference sample- a representative of the population of interest
  • A person’s score is compared against the average of what you would expect from all people in the population of interest.
  • g. may need to score at the 75th percentile of a test in order to get accepted into a program.

Scaling:

  • Concerns the way in which numerical values are assigned to psychological attributes
  • With respect to psychological measurement, there are three key properties to understand:

o Properties of identity o Properties of order o Properties of quantity Properties of identity:

  • Simplest measurements are those that differentiate categories of people who share a psychological feature
  • Teachers can categorise students into behavioural problems and non-behavioural problems
  • Students in each category should be similar to each with respect to this attribute, differences are considered in kind rather than amount.
  • Rules of sorting behaviours into categories- categories must be mutually exclusive (e.g.

introvert can’t also be extravert)

  • Secondly, categories should be exhaustive
  • Thirdly, all people classified within a given category must be identical with respect to the attribute of interest.

Properties of order:

  • More informative than the property of identity which is measurement. E.g. ranking is a property of order.
  • Numbers serve as labels, type of label doesn’t matter as long as the connection between label and order is established.

Properties of Quantity:

  • Provides more info than the property of order
  • Reflects the ability of numerals to provide information about the magnitude of differences between people.
  • The number 1 is used to define the size of the basic unit on any particular scale, all other values on the scale are multiples of 1 or fractions of 1.

The number zero:

  • Can imply or not imply the total absence of an attribute, some zeros are absolute others are arbitrary.
  • If people score zero on a percentile test, often it just means that only 1% of the population scored less than they did, not that they have zero skill levels.

Additivity:

  • Implies that the unit size of a measurement does not change as the units are being counted.
  • Additivity requires unit size to remain constant, when this is perfectly satisfied we refer to the condition as ‘conjoint measurement’.
  • This is rarely satisfied, e.g. scoring 50/100 doesn’t imply that a person possesses 50 units of knowledge.
  • Not satisfying this assumption doesn’t mean psychological measurement isn’t possible- just need to acknowledge the limitation of measurement and move on. Measurement:
  • The assignment of numerals to objects or events according to rules  Four scales of measurement:
    • Nominal o rrdinal
    • Interval o Ratio
  • There is a linear increase in the amount of information associated with levels of measurement, going from nominal to ratio.

Nominal scales:

  • Most fundamental level of measurement, symbols or numerals that have a property of identity are used to label observations in which behaviours have been sorted into categories according to some attribute
  • g. sex, state living in. Very few psychological nominal scales however.

Ordinal scales:

  • Links observations of behaviour thought to reflect qualitative differences in amounts of an attribute to symbols or numerals that have the property of order.
  • Key difference in comparison to nominal scales is the ordering in the attribute of interest
  • Not all ordinal scales are equal, as ranks are the least informative. Some ordinal scales do not satisfy the conditions of interval scales, but you do get some sense of differences in magnitude between the cases.

Interval scales:

  • More informative than ordinal scales, there is a constant distance between each of the units. However there is no meaningful zero point in interval scales.
  • Absence of a meaningful zero point prevents the multiplication or division of values

Ratio Scales:

  • Share the same properties of interval scales, have the additional condition of a meaningful zero point.
  • There are probably no psychological tests which satisfy the conditions of a ratio scale, none have a natural and meaningful zero point