Ch. 3 exercise: ratings and reliability

systematic errors

As you read through the checklists, you find that, in creating the forms used in your school system, a clerical error was made and both the means (having the means to follow through on the threat) and focus (the identification of specific individuals, locations or events as the target of the threat) ratings for the threat were omitted. In terms of classical test theory, this will add to _____.

coefficient alpha

In order to compute a measure of internal consistency, you consider which specific measure to use. Because the scoring is not dichotomous, you would be well advised to use _____.

split-half

Since there are multiple items for each of the components (the threat itself; the student's current behavior; personal factors; and home/community factors), you decide to look at a measure of internal consistency: _____.

random errors

When you review the situations in which the protocol has been implemented, you find that the checklist is often initially completed by a teacher or administrator who has recently been the focus of a student's verbal aggression and includes at least threats of and sometimes actual physical aggression. As a result, the person completing the checklist may be very stressed at the time it is completed. Moreover, the extent to which the stress impacts their ratings is likely to vary from person to person, depending on their individual training and personal history. In terms of classical test theory, you would expect this to affect _____.

interrater

You are pleased to find that the school system has a requirement that, when more than one staff member witnesses an incident, each staff member must independently complete a checklist. Although you recognize that the conditions for doing so are not ideal, you notice that this will give you a chance to estimate _____ reliability.