* Content Validity: The test items accurately reflect the content domain being assessed. For example, a math test designed to assess multiplication skills should primarily contain multiplication problems, not division or addition.
* Construct Validity: The test accurately measures the underlying theoretical construct it's designed to assess. This is more complex and often involves looking at correlations with other measures and exploring whether the test behaves as expected based on theoretical understanding of the construct. For example, a test measuring intelligence should correlate with other measures of intelligence (like IQ tests) and show expected differences between groups known to have different levels of intelligence. Different types of construct validity exist, including convergent validity (correlates with similar measures), discriminant validity (doesn't correlate with dissimilar measures), and predictive validity (predicts future behavior related to the construct).
In simpler terms, a valid test gives you a true and accurate picture of what you're trying to measure. An invalid test, on the other hand, might produce results that are misleading or meaningless. For example, using shoe size to measure intelligence would be an invalid test because shoe size doesn't relate to intelligence.