Professional Documents
Culture Documents
Chapter 2 Research Concept, Hypotheses, and Tests
Chapter 2 Research Concept, Hypotheses, and Tests
not have to. Made explicit as a first step in research projects, preconceptions can be helpful. 2. Formulating hypotheses Investigators initially formulate exploratory hypotheses based on theory and previous empirical data, then they use preliminary, unfocused investigation to decide what specific data they will use to confront these hypotheses. Complex and sophisticated possible solutions to a problem that is hypotheses can be thought of as conceptual models analogous to the physical models that designers use. Models represent the intended resolution of the problems in mathematical, symbolic, physical, or some other form. A. Classifying Hypotheses a) List of types: The list enabled the researchers to ascertain that demonstration, filmstrips with story line. Further explanation of why these types work the way they could be useful in deciding how to convey public information effectively b) Substructures: Investigators organize information by structuring it within a descriptive ordering system. This organizing system may be constructed by juxtaposing several attributes that describe the research problem. B. Explanatory Hypotheses a) Manifest explanations: refer to actors intent b) Latent explanations: refer to unforeseen effects. They are testable. But they harder to formulate because they are unexpected by participants. 3. Empirical Testing A. Observing a) Single observations. One observation is the simplest research datum a smile, a park, a movement, an answer to a question, an event, etc. Single observation that surprise the observer tend to indicate interesting research avenues because such observations conflict with exploratory hypotheses formulated from theory, from other empirical research, or from common sense. b) Regularities: Developing from a single observation is to look for other like it. Looking for a regularity and not finding it makes visible another regularity-Its absence. c) Contexts. To test an explanatory hypothesis of an observed event, researchers use its context. C. Sampling
a) Randomizing: E-B researchers often study diverse groups of people, places, and environments they know little about relative to their hypothesis being tested. Randomizing is not only a useful idea, but also a surprising one. If you draw a random sample from a large group, you can generalize or project results from the sample onto the group within statistically definable limits. A common way 0to select a random sample is to put names or numbers of elements on pieces of paper. b) Matching: We want to observe a sample from a larger population with which we are familiar, we think that a characteristic of the larger group will affect what you observe. We can match the sample to the larger group on that characteristic. c) Combining randomizing and matching procedure: when researchers want to reduce generalization error from both known and unknown causes, they use mixture of randomizing and matching procedures to select their sample.