Section 2. Philosophical Perspectives

A useful perspective on what quantitative, positivist research methods are, conceptually, can be gained by seeing them in the context of Myers' framing in the ISWorld Qualitative Research Section (reprinted here with permission).

Figure 1 shows how, for qualitative research, the basic epistemological positions to choose from are threefold: positivist, interpretive, or critical. In the case of quantitative research, however, the interpretive and critical positions are not meaningful; only the positivist one is. The positivist epistemology (discussed in details next) relies on a host of scientific methods that produce numerical and alphanumeric data. Accordingly, epistemological assumptions for both quantitative and qualitative research can be represented as in Figure 2.

2.1 Positivism
At the heart of positivism is Karl Popper's dichotomous differentiation between "scientific" theories and "myth". A scientific theory is a theory whose predictions can be empirically falsified, i.e., shown to be wrong. Therefore, a scientific theory is by necessity a risky endeavor, i.e., it may be thrown out if not supported by the data. Einstein's Theory of Relativity is a prime example, according to Karl Popper, of a scientific theory. When Einstein proposed it, the theory may have ended up in the junk pile of history had its empirical test not supported it. This, despite the enormous amount of work put into it and despite its mathematical appeal. The reason Einstein's theory was accepted was because it was put to the test: Eddington's eclipse observation in 1919 confirmed its predictions, predictions that were in contrast to what should have been seen according to Newtonian psychics. Eddington's eclipse observation was a test-or-break event for Einstein's theory. The theory would have been discredited had the stars not appeared to move during the eclipse because of the Sun's gravity. In contrast, according to Popper, is Freud's theory of psychoanalysis which can never be disproven because the theory is sufficiently imprecise to allow for convenient "explanations" and the addition of ad hoc hypotheses to explain observations that contradict the theory. The ability to explain any observation as an apparent verification of Psychoanalysis is no proof of the theory, because it can never be proven wrong to those who believe in it. A scientific theory, in contrast to Psychoanalysis, is one that can be empirically falsified. This is the Falsification Principle and the core of positivism. Basically, experience can show theories to be wrong, but can never prove them right. It is an underlying principle that theories can never be shown to be correct.

This demarcation of science from the myths of non-science also assumes that building a theory based on observation, otherwise known as induction, does not make it scientific. Science, according to positivism, is about solving problems. It is not about fitting theory to observations. Central to understanding this principle is the recognition that there is no such thing as a pure observation. Every observation is based on some preexisting theory or understanding. Furthermore, it is almost always possible to choose and select data that will support almost any theory if the researcher just looks for confirming examples. Accordingly, scientific theory, in the positivist view, is about trying to falsify the predictions of the theory.

In theory, it is enough, therefore, for one observation that contradicts the prediction of a theory to falsify it and render it incorrect. Furthermore, even after being tested, a scientific theory is never verified because it can never be shown to be true, as some future observation may yet contradict it. Accordingly, a scientific theory is, at most, extensively corroborated, which makes it accepted until proven otherwise. Of course, in reality, measurement is never perfect and is always based on theory. Hence, positivism differentiates between falsification as a principle, where one negating observation is all that is needed to cast out a theory, and its application in the real world through methodology, where it is recognized that observations may themselves be erroneous and hence where more than one observation is usually needed to falsify a theory.

The viewpoint of this presentation is that positivism should be regarded as one of the tools in the arsenal of a researcher. Arguably, recognizing science as a problem-solving endeavor, positivism itself (if it could be separated from the people who articulate or follow this position) would probably endorse this position.


2.2 Post-Popperian Perspectives
As noted above in Section 1, Cook and Campbell (1979) present a less stringent and less severe perspective on the epistemology of science. While the positivist epistemology deals only with observed and measured knowledge, the post-positivist epistemology recognizes that such an approach would result in making many important aspects of psychology irrelevant because feelings and perceptions cannot be readily measured. In post-positivist understanding, pure empiricism, i.e., deriving knowledge only through observation and measurement, is understood to be too demanding. Instead, post-positivism is based on the concept of critical realism, that there is a real world out there independent of our perception of it and that the objective of science is to try and understand it, combined with triangulation, i.e., the recognition that observations and measurements are inherently imperfect and hence the need to measure phenomena in many ways. The post-positivist epistemology regards the acquisition of knowledge as a process that is more than mere deduction. Knowledge is acquired through both deduction and induction.

Another such commentator was Imre Lakatos (1970, 1978).  Lakatos argues that there are a core set of propositions in a scientific theory, surrounded by a "protective belt" of hypotheses that are related to the core set, but not requisite for establishing the "truth" of the core.  If the hypotheses in the belt are proven to be untrue, then the core propositions may still be inviolate.  The spirit behind Lakatos' scientific epistemology is similar to Cook and Campbell (1979) [which superceded a predecessor work, Campbell and Stanley (1966)] in that there is greater latitude in allowing for exceptions.


2.3 Underlying Positivist Assumptions
In all the QPR research methods the underlying statistics (mainly T, F, and Chi-square statistics) deal with rejecting the null hypothesis of no effect. (The Chi-Square statistic in LISREL is an exception, although this method too applies the T statistic.) Viewed from a positivist point of view, the objective of statistics employed by the QPR methods is to falsify the null hypothesis, which is the assumption that the data in the dependent variable are not affected by the data in the independent variable or variables. Since each theoretical hypothesis (the hypothesis as stated in the theory) should be the exact opposite of its null hypothesis by predicting a difference in the dependent variable, it follows logically that if the null hypothesis is rejected, then presumably the theoretical hypothesis is supported. The theoretical hypothesis is supported in this case but not proven, because theory in the positivist philosophy cannot be proven, strictly speaking. The essence of the statistics also takes into account the positivist recognition of imperfect measurement; hence, statistics test the probability that the results could have been obtained due to randomness in the data given the nature of the sample. It is based on this probability that the null hypothesis is rejected and by implication that the theoretical hypothesis is supported.