Theme 3: Research and theory (Before)

What theory is and what theory is not

Theory is the argumentation or the logical part of something which you are trying to prove. A causality or a link between two events may be shown to be likely, or very likely, by data or other observation. Then theory is the logical explanation to that event. Theory is not reference to the works of others, data in any form nor the hypothesis of a scientific work. Theory is the attempt to explain why something is happening and gives the possibility to at least to some extent predict what will happen under certain circumstances. The latter is especially true for strong theories. The results of a study constitute, in a broad sense, data and should therefore not be seen as theory. Instead, theory arguments about the connection between artifacts of that data and explains how and why they are connected.

The Journal of the Acoustical Society of America (JASA)

I have chosen both to write a short description of the above journal. I have also chosen to write about an article published in that journal.

The JASA publishes a range of articles within the area of acoustics. Acoustics is a broad research area and therefore the articles published in JASA come from many different disciplines with acoustics in common. Acoustics are relevant for construction, different kinds of sonar (audio radar) and their applications, and transduction (the field of research concerned with detection and production of sound waves eg. microphones and speakers). Acoustics are also relevant for areas more closely connected to humans and human perception such as the acoustics of speech, psychology and physiology of hearing and music. The latter fields are of course of great relevance for media technology.

The article of my choice:

Suied et al. (2010): JASA 2010 Mar;127(3):EL105-10. doi: 10.1121/1.3310196. "Why are natural sounds detected faster than pips?"

Suied et al. investigate which of artificial and natural warning sounds give the shortest reaction time. The study is divided into two parts with the second part being a follow up investigation to further investigate the result of the first part. In the first experiment study subjects (12 people) responded to eight different warning sounds, four artificial sounds in the form of bursts of 1kHz beeps and four natural sounds of predator cats.

In the second experiment study subjects (12 other people) responded to the same natural sounds as in experiment 1 and to artificial versions of these sounds. The artificial versions were modified white noise (random amplitude noise ranging the full spectrum of human perception) with the temporal envelope (amplitude variation over time) of the four different kinds of animal sounds.

The result in the first experiment was that animal sounds had faster reaction times. The second experiment was conducted due to the result of the first experiment. The researchers wanted to test if the effect was due to acoustic properties of the animal sounds. The result of the second experiment was that there was almost no difference in reaction time between the artificial/modified sounds and the natural animal sounds, leaning towards slightly faster reaction time for the artificial/modified sounds.

The researchers' main theory is that the higher spectral content of the natural sounds will induce more neural reactions in our hearing apparatus and brain and therefore gives a shorter reaction time. Since the second experiment shows no difference in responses a "reflex" to predator sounds is more or less excluded, given that the artificial sounds of the second experiment do not sound like predator sounds. Very little is said to prove the neural theory further, and they do conclude that further investigation is necessary. My observation is that a control should have been done to test modified white noise against just white noise. 

The main theory would fall under "Theory for Explaining and Predicting", I would say. Although the researchers only make an attempt at explaining how the effect of neural reaction would work the theory is logically sound. As stated in the text by Sutton & Staw, a theory is not compromised by not proving every underlying process to it. The theory, though incomplete, also gives the possibility of prediction, not only for the exact same conditions but also for similar conditions. And the theory has testable propositions. However since further research can and should be done, and the underlying causality is not precisely explained, the theory is balancing between Type 3 (Predicting) and Type 4 (EP).

The benefits of using this theory is that further research can be done upon it without fully explaining the premises. The acoustical characteristics that gave a certain response can be subjected to variation to further specify what the important characteristics are. But since the underlying causes are not proven coming up with new experiments is harder. The theory is limited by assumptions of the acoustical characteristics. If the reason for the observed effect is not due to the "neural" explanation then further research might steer in the wrong direction.