The Big Picture: What is Psychological Measurement and Why is it Important

Psychological measurement

The systematic application of basic principles to measure personal attributes which are thought to be important in describing, understanding and predicting human behaviour

Assumptions in psychological measurement

relatively stable differences among people - people differ in traits like intelligence and personality → consistent over time - variability - different levels of these traits, measure and compare

hypothetical constructs vs. directly observable characteristics - height and weight can be measured directly while psych traits (personality intelligence) cant be directly observed - behaviour as an indirect index

Measurement of constructs is fundamental to science…

because conclusions from research are dependent on the quality of measures

Research Hourglass

top (broad): start with general research question

Narrow down: focus on a specific question and operationalize the variables (define how they will be measured)

Neck: measurement is critical because you need to accurately measure the variables being studied

Bottom (broadening): analyze the data, draw conclusions, and generalize the findings back to the original research question

Application

Psychiatric Hospitals or Clinics

Family or Child Guidance Centers

Elementary and High Schools

Colleges or University

Business or Industry

Elementary and High School

Learning disorders

ADD/ADHD

Autism

Giftedness

Vocational interest

Hospitals, Psychology and Psychiatric Clinics

Alzheimers

Aphasia

Bipolar disorder

Dissociative disorder

Psychopathology

Construct to item

We believe that specific test items can measure an underlying psychological construct

construct influences behaviour

extraversion influences how people behave, think, and feel

items

questions on a test assess these behaviours, thoughts, and feelings to help measure the construct of extraversion

main question for construct

do the items form a reliable and valid measure of construct?

Reliability

are scores on the test/measure repeatable?

Validity

do scores asses the construct?

test scores

the average of all item scores, influenced by the construct of interest and measurement errors

Random measurement error

unpredictable factors that can raise or lower scores, such as fatigue, we want to minimize this because it reduces reliability

Systematic measurement error

consistent influences on scores that are not the construct being measured, such as text anxiety, want to minimizes this because it reduces validity