click below
click below
Normal Size Small Size show me how
RMA 4
RMA Lecture 4
Question | Answer |
---|---|
Why is validity important? | Psychological tests guide policies and decision making. Without it, these decisions can be misinformed |
What is content validity? | The content of a test measures the content that *should be* included in the test e.g. Baratt Impulsiveness Scale (BIS-11) |
What are the 2 threats to content validity? | 1. Construct-irrelevant content, and 2. Construct under-representation |
What is construct-irrelevant content? | Content that is irrelevant to the core construct e.g. I like to watch the walking dead etc |
What is construct under-representation? | Content fails to represent the full scope of the content |
Why should there be a balance between construct-irrelevant and construct under-representation? | Unable to include every facet of a construct due to test-taking duration and participant fatigue |
What is face validity? | The degree to which a measure appears to be related to a construct in the judgement of non-exerts |
Why is face validity dangerous to validity? | Superficial meaning and relevance of content might influence participant to respond in a less-than-honest manner |
What are the 4 types of evidence for validity? | 1. Test content 2. Internal structure 3. Response Processes 4. Associations with other variables |
What is internal structure? | The way the parts of a test are related to each other. |
How should internal structure reflect validity? | There should be a match between the actual internal structure of a test and the structure that the test should possess. I.e. we should expect a single tight cluster of items if, SAY, all the items in a test are correlated with each other. |
How do response processes reflect validity? | There should be a match between the psychological processes that respondents actually use when completing a measure and the process they should use. |
How does the test's association with other variables reflect validity? | There should be a match betwen a measure's actual associations with other measures and the asssociations that the test should have with other measures. |
What are the two types of evidence reflecting validity in the test's associations with other variables? | Convergent and discriminant evidence |
What is concurrent validity evidence? | The degree to which test scores are correlated with other relevant variables that are measured *at the same time* |
What is predictive validity evidence? | The degree to which test scores are correlated with relevant variables that are measured at a future point in time. |
What is a nomological network? | The interconnections between a construct and other related constructs. |
What are the four methods used to evaluate the degree to which measures show convergent and discriminant associations? | 1. Sets of correlations 2. Multi-trait multi-method metrices (MTMMM) 3. |
How do sets of correlations demonstrate convergent and discriminant associations? | "Eyeballing" correlations and making judgments about the degree to which the correlations match what would be expected. |
How is a MTMMM carried out? | Researchers obtain measures of several traits (Multi-trait), each of which are measured through several methods (Multi-method) |
How does shared method variance undermine validity? | Correlation of traits in the method is a by-product of bias inherent in method, and not commonality between the traits being measured. |
How do correlations in the MTMMM reflect convergent validity? | High correlations among measures of the same construct (Mono-trait), using different methods (Hetero-method) provide evidence for convergent validity. |
How do correlations in the MTMMM reflect convergent validity? | Low correlations among measures of different constructs (Multi-method), using the same methods (Hetero-method) provide evidence for discriminant validity |
What are the 4 requirements to conclude that a measure is valid in MTMMM? | 1. High cor. in 1trait-manymethod 2. Cor. in 1trait-manymethod more than cor. in manytrait-1method 3. Cor. in the 1trait-manymethod more than cor. in manytrait-manymethod 4. Patterns of trait interrelations same in both 1method and manymethod |