Establish a firm span of concepts that will be included in the concept map. For example, if you are creating a concept map of statistical research, you may want to limit the types of statistical processes and analyses involved to a predefined set. Such a set may be only the statistical methods dealing with categorical data, for instance.
List down the important concepts related to the topic that will be included in the concept map.
Make notes as to which concepts fall above, below or parallel other concepts, creating a hierarchy of concepts. For example, you may find a specific concept is actually a special case of another concept; this would be a above-below relationship. In addition, you may find that two concepts are the same except for one small difference; this would be a parallel relationship.
Write the concepts out on a piece of paper or poster board. Keep in mind the relationships between concepts while doing so. Place the concepts in boxes, circles or other shapes that act as barriers between your words. You may consider having different shapes for different meanings. For example, you may consider placing discrete statistical distributions in boxes and continuous statistical distributions in circles.
Place arrows between the concepts in a meaningful manner. You can decide yourself as to what the arrows’ directions infer. For example, for a statistical distribution concept map, you may show arrows leading to special cases of more general distributions. For instance, the Gaussian distribution may to the standard normal distribution, indicating that the standard normal distribution is a special form of the Gaussian distribution.