Psychological tests and measurement scales

what are the components of ethical research

Informed consent, deception, protection from physical and psychological harm, right to withdraw, confidentiality and ensure anonymity, freedom from coercion (persuading someone using threats). Study followed by debriefs

define coercion

persuading someone using threats

define ‘ construct ‘

a variable that you want to measure e.g. personality

Aim of measurement in psychology

test theoretical hypothesis, and measurement/ quantification of psychological constructs allows us to do this

Psychometrics

an area of psych that focuses on scientific measurement of individual differences within psychological constructs. Psychometrics is at the intersection between statistics and psych

What are psychometric tests?

standardised instruments for assessing an aspect of an individual, like ability, aptitude, attitude and personality. Same or statistically similar questions are administered and scored in a consistent way every time - compare individuals

Concerns of psychometrics

addressing the quality of scales and items that were designed to measure psychological constructs

Define ‘ discriminatory power’

should be able to differentiate individuals across a range of profile, and not just pick of extreme cases

define ‘population comparison’

Allow its results to be applicable to a population and not only the individuals that were tested (e.g. through standardisation)

What 2 metrics are important for screeners/ assessments of disorders?

Sensitivity and specificity

define ‘ sensitivity’

TRUE POSITIVE rate. Proportion of actual cases correctly identified by the test

what does a high sensitivity ensure?

most people with the trait/ condition are detected (low false negatives)

define ‘ specificity’

TRUE NEGATIVE rate. Proportion of non-cases correctly classified by the test

What does a high specificity ensure?

most people without the trait/condition are correctly ruled out (low false positives)

define reliability

consistency and stability of a measurement tool over time, across items, and between raters

when is a test reliable?

when it produces the same results under consistent conditions

what is the rule of reliability and validity?

An instrument can’t be valid unless it’s reliable. An instrument can be reliable and not valid

Why is reliability important?

• ensures psychological tests measure constructs consistently

• reduces measurement error, improving the accuracy of research findings and assessments

• necessary for validity - a test cannot be valid if it is not reliable

What is external reliability?

assesses the degree that a measure varies from one use to another

What is internal reliability?

assesses the degree to which the individual items within a measure is consistent within itself

What are the different types of reliability?

• Test-retest reliability

• Inter-rater reliability

• Internal consistency

• Parallel-forms reliability

define ‘test-retest’ reliability

the stability (consistency) of scores over time.

measured when giving the same test to same group at different time points and compute the correlation

define inter-rater reliability

consistency of scores across different raters/observers

measured when assessing degree of agreement between multiple raters

define internal consistency

the degree to which items on a test measure the same construct

measured using Cronbach’s alpha to assess how well test items correlate with one another

define ‘parallel-forms reliability’

consistency of 2 equivalent versions of a test

measured by correlating scores from 2 different forms of a test measuring the same construct

1. what should a good retest reliability correlation be?

2. What does the correlation depend on?

1. .75 - .80

2. The length of time between administrations of the test, especially with children (developmental stage). The type of construct measured - stable (IQ) vs malleable construct (e.g. mood)

which ways are there to calculate internal reliability?

Cronbach’s alpha, split half reliability, Kuder Richardson - the one you choose depends on the types of items you have and format of responses

Split half reliability

split the test in half (either randomly or assign all odd items to one set and all even to another), then calculate the correlation between the scores on the 2 halves of the test. Higher correlations = better internal reliability

What are the 5 factors affecting reliability?

1. test length - longer tests = more reliable

2. item quality - poorly written or ambiguous questions reduce reliability

3. testing conditions - environmental factors

4. score variability

5. participant variables

define validity

the extent to which a test measures what it is intended to measure. ensures test scores are meaningful and useful for decision making

Why is validity important?

• ensures accurate interpretation of test results

• increases confidence in using test scores for research, diagnosis and education

• essential for fairness - valid tests reduce bias in psychological and educational assessment