WEEK 4 - 2250A PSYCH

Vocabulary flashcards reviewing concepts of response biases, reliability types, validity, and generalizability based on the lecture transcript.

Social Desirability

Giving responses that make you seem more positive/desirable or socially acceptable than you really are. Can involve exaggerating positive traits/feelings or denying negative ones.

Example: Agreeing with “I’m incredibly happy,”. Can be conscious (e.g., trying to look good in a job interview) or unconscious (e.g., personality-related). can be Influenced by cultural/social norms.

Faking

Consciously altering responses (100% conscious) { verbal intelligence good at faking}. Fake good: Making yourself seem better than you are. Fake bad: Making yourself seem worse than you are.

Reliability

Consistency in measurement.

Test-retest Reliability

Consistency over time. Give the same test to the same people at two different times; similar results = high reliability. Used for things that stay relatively stable, like personality or intelligence, not things that naturally change like growth in children.

Inter-rater Reliability

Consistency between observers. Used with observational data. Two or more people observe the same thing; similar ratings = high reliability. E.g., one person interprets an expression as a smile while another sees it as a smirk - low reliability

Internal Consistency Reliability

Measures consistency between questions/items on the same test. If different questions measure the same construct, the answers should be related.

• Example: A questionnaire asks several different questions that measure the same construct. If responses to those questions are consistent → high internal reliability.

Internal consistency cont

Questions that measure the same construct should have similar/related scores.

Example — Sentience: If someone scores high (4–5) on one sentience question, they should generally score high on the other sentience questions too. Same construct → similar responses = high internal consistency.

Validity

Accuracy; the degree to which a test actually measures what it claims to measure. Example: If a test claims to measure anxiety, it should actually measure anxiety, not something else.

{ types: face validity, Predictive/criterion validity, Convergent validity, Discriminant validity, Construct validity}

Construct Validity

The umbrella form of validity used for things we can’t directly see or physically measure. A construct is something like intelligence, anxiety, happiness, or personality. We gather evidence to see if our test is actually measuring it properly.

Generalizability

The degree to which a measure stays valid in different contexts, across different groups of people, conditions, and situations. Do the results still apply in other situations and to other people?

remember

Remember: Reliability = consistency. Validity = accuracy

cannot have validity without reliability, but can have reliability without validity.