Testing and Measurement Flashcards
Personality Assessment
the measurement and evaluation of psychological traits, states, values, interests, attitudes, worldview acculturation, sense of humor, cognitive and behavioural styles, and/or related individual characteristics
Personality Type
A constellation of traits that is similar in pattern to one identified category of personality within a taxonomy of personalities
Personality State:
The transitory exhibition of some personality trait; a relatively temporary disposition
Personality Profile:
A narritive description of the extent to which a person has demonstrated certain personality traits, states, or types
Parameters of measuring personality
- Reliability
- internal consistency (usually measured by Cronbach’s alpha)
- test-retest
- alternate form (one test, two different forms)
- Validity
- Construct Validity
- extent to which a test truly measures a theoretical construct
- convergent validity, which involves showing that the test relates to the variables it should theoretically relate to.
- discriminant validity, which means showing that the test does not relate to the variables it should not relate to.
- criterion-related validation The ongoing process of validation also involves demonstrating that a test predicts some relevant real-world outcome
- extent to which a test truly measures a theoretical construct
- Content Validity
- measures the domain that it is meant to be measuring, and not some other domain.
- Construct Validity
- Not a TRUE type of validity but important
- Face Validity (Does it look like its going to measure what it wants to measure)
Major sources of Bias in personality measurement
- Bias
- Ethic
- Most tests are built in Western educated countries, so lots of items don’t translate, and
- Gender
- Some tests won’t matter, in some tests it matters a lot, can be history and culturally dependant (gender roles, in terms of normal or non normal behaviour)
- Response sets
- Unrelated to the question we are asking, (naysaying response sets and such)
- Social desirability responses will be people trying to inflate or deflate socially un/desireable sets.
- People rushing through the test, people faking good, faking bad
- Good tests will design scales to account for and measure these response sets
- Ethic
Different Types of measures of personality
- Self Report
- Most widely used test around today (MMPI-2/RF, 16PF, NEO-PI, CPI, PAI, Millon Clinical Multiaxial Inventory)
- Most reliable and valid, can be measured quantitively
- Projective Tests
- Rorschach, thematic apperception Test
- Lots of subjectivity in how its responded to, and scored
- Not very reliable or valid
Different Types of measures of personality 2 - less common items
- Q-Sort
- Sorting self reports into particular piles of groups to represent a more accurate report, used to get people to try and describe themselves
- Ratings/Judgements by others
- Used particularly with parents/teachers of young children, as they will tend to have a very good understanding of their child when they are too young to reliably self-report
- Biological Measures
- Phrenology (debunked)
- EG scans to determine brainwave response to certain stimuli, not used too much in extent to personality
- MRI changes of oxygenated blood flow when mapping cognitive abilities, also looking at emotional reactions.
- Encephalitis, lead poisoning to see how they affect personality and behaviour
- Behavioural observations
- Experience sampling method popular method now to collection information, participant is sent a text at a certain time or moment, and asked to report what they are doing, thinking, feeling.
- Interviews
- Lots of information on personality especially if its structured, but still relies on the person giving accurate information. Open ended questions still leave a large part open to subjectivity
- Expressive Behavior
- Non-verbal behaviour that can be representative of certain traits, can have bias (cultural differences like eye contact in some is respectful, in some its not)
- Demographics and Lifestyle
- Online Internet Analysis of Social Media and Big Data
- Controversial, also big business
- Raises issues about ethics, privacy and abuse
Limitations of Personality Tests
- Look at the purpose of the research question
- Clinically oriented?
- Is it a good test? Valid/reliable? Culturally appropriate? Is there Gender Bias?
- Can we combine tests to cover these pitfalls?
- Is the test too long?
- Is it an actual personality test?
- Some are not true measures such as Myers-Briggs, Astrology, DISC, Card Readings (all these are not very evidence based, and not reliable/valid)
Basic Types of designs in studying personality
- Case Studies
- In-depth study of a person, usually in cases of very famous people. E.g a very charismatic leader, what elements of personality made them so successful?
- Require further systematic study before we can generalize findings, cause its only focused on one person
- Correlation Studies
- Measuring degree of relationship between variables
- Understand relationships between factors/items/personality traits, can use factor analysis to find commonality in data sets
- Non directional and no causation
- Experimental Studies
- Can infer causation
- harder to implement, more strict control and requires randomisation
- Very hard to randomly assign people to groups in personality design, limited mostly by random assignment. We tend to use quasi-experimental
Ethical Context of Psychological Testing
- Bias
- Misuse
- Abuse
This is why there is quite heavy restriction around psychological tests, administration and use generally restricted to qualified people.
Are we taking into account concerns of the public when administering and inferring finding for tests.
Participants need to know how their data will be used, and will require informed consent.
