Skip to main content

Table 1 Research questions and Evaluation Criteria

From: Predicting epidemics using search engine data: a comparative study on measles in the largest countries of Europe

no Research Question (RQ) Evaluation Criteria (C)
1 The predicted cases show a strong correlation with the real cases c1-Correlation: The correlation between measles cases and predicted cases must exceed 0.650 and must also be statistically significant. The significance level (p) was set to be < 0.01 in two-tailed significance
2 The prediction shows the right time of measles’ outbreaks c2-Time: The prediction point of measles outbreaks must not exceed one month in relation to the real cases and must not be observed after the outbreak
3 The predicted value of outbreaks is close to the real cases c3-Volume of predicted cases during outbreak periods: must not exceed 28% of the real cases during outbreaks
4 The prediction includes all periods with excessive activity of measles (outbreaks) c4-Outbreak predicted periods: the distributions of each prediction must include all outbreak periods within the 5-year examined period
5 The error of the estimate (MSE%) is smaller than 28% c5-Mean Standard Error (MSE): the MSE of all predicted cases must not exceed 28% of the real cases mean