chapter 5

• what is validity?

  agreement between a test score or measure and the quantity it is believed to measure. (basis for debates, lawsuits, etc. / not a characteristic of tests).
• what is face validity?

  the appearance that a measure has validity, often said that a test has face validity if the items seem to be reasonable related to the perceived purpose of the test. (an item looks valid; you know what it is measuring) e.g., optimism; I am a glass half full person.
• what is content related validity?

  the content reflects what it is supposed to reflect.
• content related validity; what is domain representation?

  domain (what you're trying to measure) representation; e.g., exam on chapter 4, if the domain reflects material from chapter 4 then it is content valid. [educational or achievement tests.]
• logic and expert judgement?

  person who knows the material says that is what is on the test.
• what is construct underrepresentation?

  the test (construct) is underrepresented; e.g., exam on chapters 1, 2, 3 but there are mostly items from chapter 3 over 1 or 2; chapters 1 and 2 are underrepresented.
• what is construct irrelevant variance?

  error; something outside of the domain. / e.g., a question on an exam being confusing; word problems (sorting through the irrelevant information to get the relevant information).
• what is criterion related validity?

  no domain; correlation with criterion (outcome); e.g., SAT predicting GPA, test scores to predict GPA; x axis will always be testing scores and y axis will always be the criteria, / criterion is what you're trying to predict.
• what is predictive (future outcome)?

  we predict things that haven't yet happened.
• what is concurrent (immediate outcome)?

  "concurrent means at the same time." e.g., going to jail for two 20 year sentencing; you serve both 20 year sentences at the same time.
• what is validity coefficients (a number) (correlations)?

  a number that gives you can idea of how valid your tests are; relationship between a test and criterion. e.g., validity coefficient of SAT would be if their score and GPA are correlated; then there are validity coefficients.
• construct validity; what is a construct?

  some underlying thing that can't be observed. e.g., knowledge of chapter 5, preliminary construct paper
• what is construct validity?

  relationship between test and other measures (not just other tests); two tests that are measuring the same thing better correlate!
• what is convergent validity?

  coming together, intersect (measures of similar things should converge); "correlates with stuff it should correlate with." e.g., happiness scale and optimism scale, reading test and a language test.
• what is discriminant validity?

  diverge, separate. (measures of things that are not similar should not correlate). "does not correlate with stuff it should not correlate with." e.g., happiness scale does not correlate with depression scale.