Empiricism uses observation to judge the tenability of arguments.
The scientific method, in which findings are based on objective, systematic observation and verified through public inspection of methods and results, is the dominant methodological approach in political science. The ultimate goal of science, which is not always attained, is to use verifiable results to construct causal theories that explain why phenomena behave the way they do.
Scientific knowledge exhibits several characteristics, ten of which are particularly noteworthy:
- Science depends on empirical verification to confirm that statements are true through objective observation.
- Statements or hypotheses must be falsifiable, meaning that the statements or hypotheses can be refuted through contravening empirical evidence.
- While political scientists sometimes produce normative knowledge, which is concerned with evaluation or prescription about what should be, most scientists would agree that the goal of science is nonnormative knowledge, or, the factual or objective determinations of what is.
- Scientific knowledge must be transmissible—the methods used in making scientific discoveries must be made explicit so that others can analyze and replicate findings.
- Scientific knowledge is cumulative because scientists build upon the research techniques and results of previous work in advancing the scientific enterprise.
- Science summarizes relationships between two or more individual facts through the use of empirical generalization.
- Scientific knowledge is explanatory because it answers “why” and “how” questions through a logically derived set of propositions about the relationship between two or more components. Causal relationships, more so than correlation, are especially important in establishing informative and useful explanations of political phenomena.
- Science seeks to explain through the power of prediction by offering systematic, reasoned anticipation of future events, which once confirmed, provide evidence that the scientific knowledge responsible for generating the prediction is correct.
- Most scientists accept probabilistic explanation—that 100-percent accuracy in prediction is not necessary to understand a phenomenon.
- Science relies on parsimony, or simplicity and elegance, to choose between alternate explanations. The explanation that explains the most about a phenomenon with the fewest parameters will be preferred.
A theory is a body of statements that synthesize knowledge of and explain phenomena. It leads to specific and testable predictions about empirical reality—the more observations support these predictions, the more the theory is confirmed.
There is no single prescription for finding scientific truth—research taking many different approaches can reach the same goal of being labeled “scientific.” The ideal construction of a research project would begin with a well-defined research question. Next the researcher will explicitly specify hypotheses, or the terms to be tested through the collection and analysis of empirical data. The final steps include making conclusion about whether the data support the hypotheses or whether the hypotheses must be abandoned or modified.
Theories are sometimes described by their explanatory range, or the breadth of the phenomena they purport to explain. Narrow theories only apply to limited events or behaviors, while broad theories apply to an entire body of human behavior. The broader the range of the things to be explained, the more valuable the theory.
The development of a political science project begins with developing an idea from a number of sources including general observation and the scientific literature. It also includes forming testable hypotheses, collecting data to test the hypotheses, analyzing the data, and interpreting the results.
There are two general objections to classifying political science as a science:
- Practical objections: political behavior is very complex, people can intentionally mislead researchers, and data can be difficult or impossible to attain.
- Philosophical objections: human reasoning cannot be objectively measured and “facts” are conditioned by the observer’s perceptions and opinions.
The increasing reliance on statistical methods has been somewhat controversial within the discipline. There are many alternative approaches to research including constructionism. According to constructionism, humans do not simply discover knowledge of the real world through neutral processes, such as experimentation or unbiased observation, but rather create the reality they analyze.