There are many possible sources of error; the major sources are listed below:

1) Factors Considered

The factors considered in this study were mainly landuse, and land type related. There are many other factors which would help to determine the location of a golf course. Factors such as income of the local population, age, or education would all be considered when decisions such as this are made, however they are not considered here, for the sake of simplicity and clarity of results. Also not all physical factors are considered, large amounts of variation in relief should have been considered however due to data collection and conversion constraints this factor was eliminated from the study.

If all possible factors had been considered results would have been unclear and difficult to interpret. For this reason the possible scope of the study was focused on which land uses and land types affected the location of courses, and even these choices were subjective.

2) Subjectivity

Performing an MCE requires the user or modeler to perform many subjective decisions. One area that subjective decisions are made is in the functions, and limits chosen for use in the fuzzy module for the standardization of data. Although the different options and limits were tried in order to find the one that is the most suitable the decision as to which option this was still a subjective one.

In addition to this the weights assigned in the MCE have an element of subjectivity. Although sensitivity analysis was preformed in order to minamize the level of subjectivity it can never be fully eliminated.

3) Sensitivity Analysis

The sensitivity analysis relates the factors chosen to the present locations of golf courses. However the GVRD is a dynamic ever changing area, there for the factors that place a course in its location 30 years ago may have little relevance today. Therefore the sensitivity analysis loses some of its credibility. The factors affecting location will change over time.

4) Validation

No official validation was preformed. The sensitivity analysis did this to a point however no official validation techniques were employed, even ground truthing, or Taylor uncertainty analysis would have been beneficial to the study however none was preformed.

5) Limits of Knowledge

Although the user had a base knowledge of IDRISI and its functions a more extensive knowledge could have resulted in the employment of more appropriate analysis techniques.