CRIM 220 Lecture Notes - Lecture 3: Discriminant Validity, Physical Disorder, Convergent Validity
Document Summary
Whether statements about cause/measures are correct (valid) or invalid. Validity threats are possible sources of false conclusions. Statistical conclusion: whether 2 variables are statistically related to each other. Would results be the same in a different sample. Understand how a measure relates to the conceptual entity that we wish to study. Conceptions: mental images we have that represents our thoughts about something we routinely encounter. Constructs: words/symbols used to represent the mental images. Direct observations: qualities that can be observed directly. Constructs: theoretical creations, cannot be observed directly/indirectly (iq) Specifying precisely what we mean when we use particular terms. Results in a set of indicators of what we have in mind. Indicates a presence/absence of the concept we are studying. Ex. violent crime = offender uses force/threatens to use force against a victim. Requires determining what might work as a data-collection method. Able to classify every observation in terms of 1 of the attributes.