However, more information is needed about the contextual determinants of excessive drinking. Identification of the interpersonal and intrapersonal situations associated with excessive drinking would facilitate prevention efforts and would guide educators and counselors in the development of focused interventions to reduce abusive drinking. This study focuses on college students who drink alcohol regularly but have not been identified as problem drinkers; the situational contexts that differentiate heavy drinkers from light and moderate drinkers are evaluated. ( Carey, 1993, p. 217)
In the fourth example, the author begins by explaining that he has generated a theory out of other researchers' work (which he cites in brackets) and that the study reported in his article was designed to test hypotheses derived from that theory.
Many writers have suggested that the effects of high school tracking [separating students into college-preparatory and vocational streams] on student achievement vary among schools, but none has offered a compelling theory for why this may occur [ Heyns, 1974; Hauser, Sewell, and Alwin, 1976; Rosenbaum, 1984]. I use existing knowledge about tracking to develop hypotheses for between-school differences in tracking's effects. Building on the work of Sorensen [ 1970], I argue that the impact of tracking varies according to the structural characteristics of school tracking systems. I also consider claims that tracking has different effects in public and Catholic schools [ Gamoran and Berends, 1987; Page and Valli, 1990]. I test these hypotheses by applying methods of multi-level contextual analysis to data on tracking and achievement in a national sample of high schools. ( Gamoran, 1992, p. 812)
The three aims of this chapter have been to (a) describe ways people can choose to state their research problem, (b) illustrate methods of defining key terms, and (c) show how a rationale offered at the outset of a project enables researchers to locate their problem in an appropriate domain of knowledge and to indicate what their intended study can contribute to that domain. Performing these three activities before selecting ways to collect data will greatly simplify the task of choosing the data collection methods and will enhance the likelihood that the methods will be well suited to the needs of the study.
In planning their investigations, researchers can profit from answering the following queries: