Research Hypothesis

To conduct research is to collect, analyze, and interpret data systematically so as to answer specific questions about a phenomenon of interest. These questions may be derived from conjectures about (a) an efficacious cause that brings about the phenomenon, (b) the intrinsic nature of the phenomenon, or (c) how the phenomenon is related to other phenomena. Tentative answers to these questions are research hypotheses if they are (a) consistent with the to-be-explained phenomenon, (b) specific enough to serve as guidelines for conducting research, and (c) testable (i.e., there are well-defined criteria of rejection).

Depending on the level of abstraction adopted, underlying an empirical research are three hypotheses at different levels of theoretical sophistication or specificity, namely, substantive, program, and individual research hypotheses (see Table 1).

A practical problem or a new or intriguing phenomenon invites speculations about its cause or nature or relation to other phenomena. For example, it is commonly accepted that people in small townships are friendlier than their counterparts in bigger cities. To investigate whether or not this is the case, as well as the reason, researchers would first offer a speculation (e.g., environmental effects) that, if substantiated empirically, would explain the phenomenon (see row 1 of Table 1). Such a conjecture is a substantive hypothesis because it explains a real-life (substantive) phenomenon (see row 2).

Substantive hypotheses are typically too general or vague to give directions to empirical research. Part of the reason is that the phenomenon is multi-faceted. For example, environmental factors are air quality, noise level, amenities of various sorts, traffic volume, and the like. By itself, any one of these environmental factors also has multiple components. Hence, an investigation of a substantive hypothesis implicates a program of related hypotheses. Such a related set of hypotheses may be characterized as program hypotheses (see R1, R2, and R3 in row 3).

To the extent a program hypothesis is well defined, researchers can conduct an experiment by specifying the to-be-used research method, materials, procedure, and to-be-measured behavior (see [i] in row 4a). Specifically, the independent variable envisaged is Ion M, whose two levels are presence and absence. A hypothesis of such specificity is an individual research hypothesis, which may be in the form of an experimental hypothesis (see E1 in row 4b).

Ion M is the independent variable in the present example (a) because Implication R1 (namely, E1) in [i] of row 3 is adopted, and (b) because of the methodological assumption that Ion M cleanses the air. Had R2 in row 3 been used instead, a different independent variable (e.g., food supplement H) would be used (e.g., see [ii] in rows 4a and 4b).

In short, research hypothesis may refer to any one of three hypotheses underlying an empirical research study, namely, the substantive, program, and individual study hypotheses. Although the three hypotheses are literally different and may ostensibly be about different things, they are theoretically bounded. Specifically, the substantive hypothesis implies any one of the program hypotheses, which, in turn, implies one or more individual study hypotheses. For these implicative relations to be possible, the substantive or research hypothesis must be well defined. An individual study hypothesis of sufficient specificity becomes an experimental hypothesis. At the same time, [Page 1260]even when one is collecting data to test a specific experimental hypothesis, one's ultimate conclusion is about the substantive hypothesis via the research hypothesis.

...