When developing a scale to measure a construct that is not based in any currently published/known theory, what should the developer do?
They should conceptualize their own theory through either a definition or a more formal set of relationships between the construct and other related constructs.
The level of specificity used within a scale should be determined by what?
The level of specificity necessary to answer the developer’s research questions.
What are four common goals for creating or using a measurement scale?
What are three major considerations that should be made when considering the content of a measurement scale?
What are the three major considerations, in terms of form, that should be made when developing a measurement scale?
How many items should a scale developer start with when creating a pool of items for a new scale?
There is no set rule of thumb, however, DeVellis suggests trying to create 3 to 4 times as many as items as you expect the final scale to have. If however, there are way too many items to test on one occasion, then it might be a good idea to eliminate several items that do not follow best item generation practices (e.g., clarity, asking about one thing, etc.).
What is one good strategy for beginning to write items for a scale?
Do NOT worry about the quality and instead focus on simply creating as many derivations as you can come up with.
What are 7 central questions when it comes to scale development?