A choice of one is not a choice.
A hypothesis is a proposed explanation for a collection of data. A rival hypothesis is an alternative explanation for the same set of data - another way of explaining the same results.
Problems with Data Analysis:
In the realm of hypothesis, the data do not speak for themselves; they must be interpreted. The act of interpretation involves many difficulties including:
a) Experimenter bias.
b) Confusing cause with correlation.
c) Improper sampling.
d) Perceptual biases.
e) Cognitive biases.
Dangers of Having Only One Hypothesis:
a) Some evidence might be ignored. If we are focused on a single hypothesis, we might overlook some information as not being relevant. However, this 'irrelevant' information might be supporting evidence for another hypothesis.
b) We may become emotionally committed to 'our' hypothesis. Falling in love with a pet theory can be disastrous in problem-solving. When this happens we tend to search for and select out only the evidence that supports our hypothesis, ignoring or subconsciously filtering out information that argues against our pet theory.
To avoid these problems, we should attempt to generate as many hypotheses as possible, and then test each of them against the known facts. An excellent way to test rival hypotheses is to use the Hypothesis Testing Matrix. You can also use the Analysis of Competing Hypotheses software program.
Rules for Generating and Testing Hypotheses:
a) The hypothesis should account for all possibly relevant data. An explanation that covers only part of the data or that is in conflict with a major fact is not a good explanation.
b) Simpler explanations are usually preferred over more complex explanations. This is the principle of Occam's razor.
c) More probable explanations are usually preferred over less probable ones.
d) When you first read how facts match a theory, you might be tempted to think, "Why, yes, that must be it." Always ask yourself, "What else might also account for the same result?”.
Psychology of Intelligence Analysis, Richards J. Heuer, Jr., New York: Novinka Books, 2006.