RECOMMENDATIONS

The statistical value of the Fisher’s exact test (p < 0.03) evaluating the accuracy of The Story Template to predict classics and non-classics based on the presence or absence of one component ([proprietary]) is intriguing. Unfortunately, since this study tested for 16 variables, the statistical result is not as strong as it could have been. For further studies, the research protocol might be designed to test only for the one variable that changed between the classics and non-classics. Also, a greater number of study stories would be desirable, although the time and effort needed to analyze each story is not trivial.