The Research Experience. Ann Sloan Devlin
we would not appreciate how the variables under examination are related to each other; and we would have a difficult time developing a comprehensive understanding of human behavior.
In the social and behavioral sciences, theories use interrelated concepts to describe and predict behaviors (events) by making clear the relationships among variables. In Ulrich’s (1991) Theory of Supportive Design, his explanatory framework, he offers three dimensions predicted to enhance patients’ well-being in health care settings. These dimensions are (1) positive distraction, (2) social support, and (3) perceived control. Having (1) positive aspects of the environment (e.g., artwork, a view to nature, or music) to distract us from our worries; (2) the social contact of others, either in person (accommodated by seating) or by phone, e-mail, or Skype; and (3) the ability to control aspects of the environment around us, for example, by using the remote control or adjusting the temperature, is theorized to lead to greater well-being. These three constructs have in common their focus on the physical environment, on the one hand, and their predicted effect on human well-being, on the other. Could we imagine generating hypotheses within this theoretical framework? Before answering that question, let’s revisit the concept of a hypothesis, which is based on a theory, is testable, and states an expected empirical outcome based on observable data.
Try This Now 1.3
Come up with two hypotheses based on Ulrich’s (1991) Theory of Supportive Design.
Theories are important because they help to organize and structure information and provide a way to think about ideas. Theories provide a structured foundation that should support the generation of hypotheses. At the same time, in returning to one of the themes of this chapter, it is important to recognize that biases may be embedded in theories. Shermer (1997, p. 46) quoted the physicist and Nobel laureate Werner Heisenberg, who stated, “What we observe is not nature itself but nature exposed to our method of questioning.” In the case of Ulrich’s (1991) theory, for example, we may stop thinking of other dimensions of supportive design if we accept his three-factor model. You can see that research questions are shaped by theory, and it is always a good idea to question whether the theory is limiting how you might think about the topic. In the case of Ulrich’s model, you might ask whether other aspects of design could be supportive beyond the three dimensions he identifies. For example, perhaps maintenance and upkeep need to be considered.
Ways in Which Theories May Differ: Scope and Parsimony
Theories differ in scope, that is in how much territory they cover, or the range of behaviors to which they apply. An example of one environmental theory broad in scope is Urie Bronfenbrenner’s Ecological Systems Theory (Bronfenbrenner, 1977, 1979), which explains the various environmental contexts of child development. There are five contexts in this theory: microsystem (immediate environment), mesosystem (connections), exosystem (indirect environment), macrosystem (social and cultural values), and chronosystem (changes over time). Thus, this theory essentially covers the entire “landscape” of child development! In contrast, Ulrich’s theory previously described is more limited in scope, applying more narrowly to health facility design.
Theories may also differ in their parsimony, that is the extent to which few rather than more explanations are needed to explain the phenomenon. Scope and parsimony are distinct concepts. One could, for example, have broad scope but need few explanations for the behavior(s). An example of a parsimonious theory with broad scope is Lewin’s postulation (Lewin, 1933) that behavior is a function of the person and the person’s environment (psychological life space) as experienced. This theory was initially represented as B = f(P E) and sometimes later as B = f(P x E).
In the history of science, parsimony appears frequently as a valued characteristic of theories. What is known as Occam’s Razor, named after 14th century English philosopher William of Occam, posits that, given competing explanations for a phenomenon, the simplest one is best. The concept of the razor comes from the idea of reducing (i.e., shaving) the explanations to the simplest. One advantage of such simplicity is that it makes the theory easier to support (or falsify). This emphasis on simple explanations does not originate (or end) with Occam, as Aristotle claimed superiority of using fewer postulates or hypotheses. One can also point to Newton, Einstein, Hawking, and others in support of this general concept (see, for example, Baker’s entry on “Simplicity,” 2016, in the Stanford Encyclopedia of Philosophy).
Another aspect of theories to appreciate is that they are powerful; in some instances, theories may become self-fulfilling (Ferraro et al., 2009). In their commentary about why theories matter, Fabrizio Ferraro et al. cited the work of Carol Dweck (2006), who showed that people’s beliefs about intelligence (whether fixed vs. mutable/changeable) can shape their behavior. In particular, those who believed that intelligence was fixed behaved differently (e.g., avoided tasks where they thought they would fail) than did those who believed that intelligence was mutable. Dweck’s research also showed that these beliefs about intelligence can be changed through social influence. In every situation, we need to ask research questions about the tenets or principles that theories propose.
Making a Connection Between a Theory and a Good Research Question
Often the way research questions are posed limits their scope and potential generalizability; that is, questions are asked in a way that limits them to a particular situation. As an example, you might be interested in the size of artwork displayed in a doctor’s waiting room and the effect of that displayed artwork on patients’ satisfaction with their time spent in the waiting room. Notice that the statement did not say patients’ satisfaction with the entire visit. Here is an important observation. When we think about variables we might manipulate or vary, such as the size of artwork, and the outcomes that might be affected, we need to think about the “distance” or “gap” between these manipulated variables (such as artwork) and the outcome variables (such as satisfaction). Why would we expect the size of artwork in the waiting room to affect satisfaction with the entire visit? Wouldn’t we want to limit our test to a more reasonable relationship—satisfaction with the time spent in the waiting room?
Now when we return to the idea of the size of the artwork, we need to pose this question in a way that avoids a restrictive and narrow focus. If we try to answer this question only for one particular size of art at two different distances, won’t we have to repeat the study with many other sizes and distances? Thus, it would help to see our research question in terms of a larger framework, something like the psychophysics of size, where we could answer the question in terms of the ratio of the size of the art displayed relative to the viewing distance in terms of the effect on satisfaction. For an example of this kind of study, see the work of Jack Nasar and Arthur Stamps (2009) on what are called infill McMansions (“too big” houses constructed in existing smaller-scale neighborhoods). Basing their research on the Weber–Fechner law (proposing a relationship between the magnitude of a physical stimulus and the intensity of people’s responses), Nasar and Stamps showed that what bothers people is not the actual size of the infill house but the relative size of the house (i.e., how the house fits into the neighborhood). Furthermore, in terms of style, large discrepancies in the height of the infill houses relative to neighbors’ houses were more disliked than were large discrepancies in width. This study uses computer-generated houses as stimuli (a very effective means of experimental control) and demonstrates how the order of presentation can be counterbalanced (Chapter 10) to make sure it is the stimuli themselves, and not the order in which they are seen, that influences our responses. In our example of the research on size and distance of art, we would need to test our hypothesis with carefully selected sizes and distances, following the Weber–Fechner law, to reach a conclusion about the ratio of size to distance that produces the most positive outcome. We want to take a specific research question and ask it in a way that has more generalizability (i.e., greater reach)—but not so much that we wouldn’t expect to see any impact (see Chapter 2 for more discussion of the research “gap”).
Weber–Fechner law: When the magnitude of a physical