Psychometrics

Psychological Test

An instrument that measures a psychological construct by sampling behavior under standardized conditions with rules for quantifying the construct

Measuring Instrument

Tool used to quantify physical properties

The best measures are…

…the ones where participants can not tell what you are measuring because it eliminates bias

Psychological Construct

Abstract psychological variables (thought of as concepts or topics - multiple ways to measure)

Sampling Behavior

All psychological tests require people to do something (answering a question is still considered a behavior)

• It’s a sample of behavior because we can not measure all aspects of the construct

Standardization

Ensure conditions under which test is administered is the same across all respondents - way of ensuring that we can trust the data presented

Quantifying

Identifying one number that represents the construct - when we get our responses, what do we do with them (add, take average, etc.)

Quantifying requires an __________ __________ of a construct

Operational Definition (what requires this)

Operational Definition

Taking an abstract concept and turning it into a concrete, measurable item

What are the three general categories of psychological tests?

1. Performance

2. Observations

3. Self-Report

Performance Test

Participants are asked to engage in a task, and are measured on their ability to engage in a task

Observational Test

Directly observable actions are quantified

Self-Report Test

Participants are asked to report his or her thoughts, feelings, or behaviors

What are the three purposes of testing?

1. Basic Research

2. Applied Psychology

3. Clinical Practice

Basic Research

Test causal psychological hypotheses (usually done in a lab)

Applied Psychology

Needs assessment (is there a need for ______), process/outcome evaluation (research in REAL WORLD settings)

Clinical Practice

Assessment of psychopathology, treatment targets

The purpose of an operational definition is to…

…establish rules for identifying when a construct is present or absent

In terms of using ttests to operationally define a construct, the importance is to…

Establish reliable and valid measures

Reliable

Gives the same measurement each time

Valid

True and accurate measure of the construct

Operational Definition

a definition of a variable in terms of the operations (activities) a researcher used to measure or manipulate it

What are operational definitions dependent on?

on what aspect of the construct you want to measure (there are three measures)

What are the three measures?

Performance, observation, and self-report

Performance Measures

Designed to measure “ability” so it usually requires them to engage in a task or series of task

Observational Measures

Careful observation and recording of behavior - requires specific coding rules

Direct Observational Measures

Observe while it’s happening

Indirect Observational Measures

Recorded in archives, video tape, etc

What are some problems with observational measures?

Reactivity (presence of researcher might influence behavior), human error

Self Report Measures

Used to assess psychological constructs that may not be directly observable (personality, attitudes, beliefs, opinions, values)

Personality

Relatively consistent pattern of thought, feeling, and behavior that characterizes each person as a unique individual.

Five Distinct Personality Types (Big 5)

Openness, conscientiousness, extraversion, agreeable, neuroticism

Scores can vary in one of four ways…

Nominal - discrete (categorical) ex. major

Ordinal - discrete (categorical) ex. grade level

Interval - continuous (equidistant) ex. temperature

Ratio - continuous (equidistant) ex. weight

Decision Making

Applied uses of testing comes in the form of making decisions

• Individual vs. Institutional

  • individual: psych assessment

  • institutional: admission, personnel selection (job application)

• Comparative vs. Absolute

  • comparative: job applications

  • absolute: diagnosis

Ethical use of tests can be understood in 3 contexts

1. Ability testing’s impact on society

2. Invasion of privacy

3. Fair use of tests

What are the ethical concerns about using tests when groups score differently?

Essentially it is systematic bias - Think about sex differences. Two groups. Only on AVERAGE do men score higher not always, there are still some females that score higher than men

Ability Testing

• Are differences in test scores real or due to an extraneous variable?

• Are the differences large?

• Do tests help or hurt?

Invasion of Privacy

Many tests involve questions that are typically considered sensitive and personal

• ensure responsible use of these tests

  • confidentiality

  • informed consent

Fair use of Tests

Avoid unfair advantage

All stakeholders in the testing process

• test developer

• test user

• test taker

• test sponser

• test administrator

• test reviewer